Browse the corpus

Introduction Ischemic stroke is a common disease that is one of the major causes of death and disability worldwide, as well as being a significant economic burden. Given an aging population, ischemic stroke is projected to become even more important in the future [1]. Animal stroke models have shed light on the pathophysiology of ischemic stroke [2] and numerous potential targets for stroke therapy have been identified. As of 2006, a total of 7,554 results of 1,082 putative neuroprotective interventions in experimental stroke have been published [3]. The translation of these results from bench to bedside, however, has been overall disappointing. To date, thrombolytic therapy using recombinant tissue plasminogen activator is still the only effective pharmacological treatment in acute ischemic stroke [4]. Many factors have been discussed that may have contributed to the lack of concordance of preclinical and clinical study results. Inappropriate conduct of preclinical research (especially insufficient quality) as well as suboptimal design of clinical trials (especially patient selection criteria and clinical evaluation) have been identified [5]. Lessons from translational failures led to the Stroke Therapy Academic Industry Roundtable (STAIR) conferences of academic and industrial experts from which recommendations and standard operating procedures for preclinical and clinical stroke research have been devised [6-10]. Ongoing deficits in experimental stroke research have been intensely discussed and analyzed, and additional useful recommendations for quality improvement have been published [11-15]. In this review article, the commonest rodent models of focal ischemic stroke and their respective advantages and disadvantages are presented and critically discussed with a focus on quality issues and translational problems.

zed, and additional useful recommendations for quality improvement have been published [11-15]. In this review article, the commonest rodent models of focal ischemic stroke and their respective advantages and disadvantages are presented and critically discussed with a focus on quality issues and translational problems. Ischemic stroke models The classification of animal models of ischemic stroke can be based on a variety of factors such as species, occlusion mechanisms and stroke etiology, presence or absence of reperfusion (temporary/transient or permanent ischemia), the involved vascular territory and infarct distribution, or a combination of these factors. A classification of experimental ischemic stroke according to infarct distribution seems reasonable, and with this approach global, focal and multifocal ischemia models can be differentiated [16]. Global cerebral ischemia, a model of circulatory failure, cannot be regarded as a stroke model in the strict sense and is beyond the scope of this review. Focal cerebral ischemia occurs when blood flow in a specific brain region is critically limited. It is usually the result of occlusion of a major cerebral artery, such as the middle cerebral artery. Multifocal ischemia represents a reduction of cerebral blood flow in multiple regions and has been described as patchy reduction of cerebral blood flow [16]. Multifocal cerebral ischemia can be caused by injection of embolus material into a brain-supplying artery.

or cerebral artery, such as the middle cerebral artery. Multifocal ischemia represents a reduction of cerebral blood flow in multiple regions and has been described as patchy reduction of cerebral blood flow [16]. Multifocal cerebral ischemia can be caused by injection of embolus material into a brain-supplying artery. The etiology of ischemia can be categorized as extravascular or intravascular [17]. Extravascular mechanisms are vessel ligation, electrocauterization, or clipping. An innovative method employing an extravascular mechanism is local application of endothelin-1 adjacent to a brain-supplying artery [18,19] or on the brain surface [20]. Intravascular mechanisms include an intravascular occluding suture in focal cerebral ischemia and injection of blood clots and other embolus material in multifocal cerebral ischemia models. An overview of cerebral ischemia models is given in Table 1. For ethical and practical reasons, rats and - especially in transgenic studies - mice are the most widely used laboratory animals in preclinical stroke research. In the following, we focus on major models of focal stroke in these rodents. Table 1 Overview of experimental stroke models Cerebral Ischemia Etiology Reperfusion (transient ischemia) or permanent ischemia Examples Global (complete or incomplete) = model of circulatory arrest or severe hypotension Intravascular Transient or permanent Cardiac arrest with or without cardiopulmonary resuscitation Extravascular Transient or permanent Cervical compression by neck cuff; ligation of several brain-supplying arteries

Cerebral Ischemia Etiology Reperfusion (transient ischemia) or permanent ischemia Examples Global (complete or incomplete) = model of circulatory arrest or severe hypotension Intravascular Transient or permanent Cardiac arrest with or without cardiopulmonary resuscitation Extravascular Transient or permanent Cervical compression by neck cuff; ligation of several brain-supplying arteries Focal Intravascular Transient or permanent Intraluminal thread middle cerebral artery occlusion model Extravascular Transient or permanent Surgical middle cerebral artery occlusion models using ligation, clipping, electrocauterization etc.; endothelin-1-induced middle cerebral artery occlusion Multifocal Intravascular Transient (spontaneous lysis or thrombolytic therapy possible in blood clots) or permanent Embolization models using blood clots, microspheres or other embolus material Major rodent models of focal stroke Intraluminal thread middle cerebral artery occlusion model The middle cerebral artery occlusion (MCAo) model using an intraluminal thread was developed by Koizumi et al. in rats [21], and has subsequently been modified [22,23] and adapted to mice [24]. Briefly, an occluding suture or monofilament is advanced via the common carotid artery into the internal carotid artery towards the junction of the anterior and middle cerebral arteries. Thus the middle cerebral artery vascular territory is subjected to ischemia. The occluding suture may be removed after various time periods, allowing for reperfusion (transient MCAo) or it may be left in place permanently (permanent MCAo).

arotid artery towards the junction of the anterior and middle cerebral arteries. Thus the middle cerebral artery vascular territory is subjected to ischemia. The occluding suture may be removed after various time periods, allowing for reperfusion (transient MCAo) or it may be left in place permanently (permanent MCAo). There are several advantages of this technique: it models focal infarction in a large vascular territory and does not require craniotomy. Thus it can be regarded as relatively simple, though microsurgical skills are required.

arotid artery towards the junction of the anterior and middle cerebral arteries. Thus the middle cerebral artery vascular territory is subjected to ischemia. The occluding suture may be removed after various time periods, allowing for reperfusion (transient MCAo) or it may be left in place permanently (permanent MCAo). There are several advantages of this technique: it models focal infarction in a large vascular territory and does not require craniotomy. Thus it can be regarded as relatively simple, though microsurgical skills are required. Specific complications of the intraluminal thread MCAo model have been reported. MCAo may be complicated by subarachnoid hemorrhage in a substantial number of animals [25-28]. Other adverse events include ipsilateral retinal injury and consequent visual dysfunction [29], ischemia in the external carotid artery territory [30,31], intraluminal thrombus formation [32,33], and inadequate MCAo and premature reperfusion [26,27,34]. If the infarct involves the hypothalamus - as seems to be frequently the case in animals subjected to permanent MCAo or to longer times of transient MCAo - hyperthermia may develop as a consequence of thermal dysregulation, leading to altered stroke outcome [35-37]. Technical modifications and the use of standardized monofilaments can reduce the incidence of these complications [28,38,39]. In line with differences in the technical approach, researchers may report tremendously different incidences or may not produce certain complications in their hands at all [27]. For example, external carotid artery territory ischemia, such as temporal muscle necrosis resulting in impaired outcome (decreased body weight and delayed recovery), has been reported as frequently as in 50% of animals [30,31], but has not been confirmed by others and seems to be related to the use of electrocauterization for vessel dissection [40]. In mice [25,41-43], but also in rats [44-46], infarct volumes vary considerably between different strains and ages, and show a steep relationship to occlusion time [47]. For example, C57Bl/6 mice have significantly larger infarcts than SV129 mice in the permanent MCAo model [25,42], most likely due to the absence of one or both posterior communicating arteries in many C57Bl/6 mice [43,47,48]. These are the most widely used parent strains for the production of transgenic mice. Moreover, infarct volumes found in different murine MCAo studies using the same mouse strain and same MCAo duration can range over a fivefold difference [49]. The different absolute and relative incidences of complications, as well as minor technical differences, may contribute to the considerable variability observed with the intraluminal thread MCAo model.

rent murine MCAo studies using the same mouse strain and same MCAo duration can range over a fivefold difference [49]. The different absolute and relative incidences of complications, as well as minor technical differences, may contribute to the considerable variability observed with the intraluminal thread MCAo model. Surgical middle cerebral artery occlusion models There are also models of MCAo that do not use an intraluminal occluding suture, but employ an extravascular mechanism (for example, ligation, clipping, or electrocauterization). These surgical MCAo models require craniotomy and incision of the dura. For example, the proximal middle cerebral artery can be identified and occluded after removal of the coronoid process of the mandible and zygoma and opening a burr hole lateral to the foramen ovale [50]. A more refined method that has been reported to cause reproducible infarctions is tandem occlusion of the distal middle cerebral and ipsilateral common carotid arteries [51]. Disadvantages of the surgical approaches, however, include technical intricacy, possible impairment of cerebral blood flow auto-regulation by damage to autonomic nerves, and craniotomy. The latter has been shown to alter brain temperature, intracranial pressure, and blood-brain barrier permeability [52,53]. For these reasons, most researchers prefer the intraluminal thread MCAo method.

intricacy, possible impairment of cerebral blood flow auto-regulation by damage to autonomic nerves, and craniotomy. The latter has been shown to alter brain temperature, intracranial pressure, and blood-brain barrier permeability [52,53]. For these reasons, most researchers prefer the intraluminal thread MCAo method. Photothrombosis model In the photothrombosis model originally described by Watson et al. in rats [54] and later modified for mice [55], a cortical brain lesion is induced by systemic injection of a photosensitive dye, such as Rose Bengal, and subsequent focal irradiation of the skull. Though some researchers expose the dura or brain surface prior to illumination, the method does not require craniotomy if the light penetrates the skull, and it can in this case be regarded as minimally invasive. A special advantage of photothrombosis is the high reproducibility of lesion size and location. It is even possible to determine the region of irradiation stereotactically and thus selectively induce infarcts in cortical areas representing specific functions [56]. The technique is relatively simple and allows high throughput.

al advantage of photothrombosis is the high reproducibility of lesion size and location. It is even possible to determine the region of irradiation stereotactically and thus selectively induce infarcts in cortical areas representing specific functions [56]. The technique is relatively simple and allows high throughput. The penumbra, however, an important target of many putative stroke therapeutics, is lacking in photothrombosis, where blood-brain barrier disruption and vasogenic edema develop within minutes [57,58], though modifications (ring models) seem to be able to produce a penumbra-like lesion edge [59,60]. The mechanism by which photochemically induced brain lesions develop is thought to involve vascular endothelial damage, platelet activation and subsequent thrombotic vessel occlusion [57,58]. Recently, however, we have shown that the photothrombotic lesion develops independently of the presence of functional platelets or plasmatic coagulation [61]. Thus photothrombosis does not reflect vascular-ischemic brain injury or stroke, but focal brain necrosis seems to be caused directly and irrespectively of additional endothelial damage leading to local thrombosis. These findings reveal significant differences in mechanisms of tissue injury induced by photothrombosis as compared to focal ischemia induced by MCAo, where platelet activation and the intrinsic coagulation pathway are instrumental [62,63]. When evaluating antithrombotic pharmaceuticals to limit tissue injury using the photothrombosis paradigm, one should therefore be cautious because ensuing platelet-containing thrombosis may be present, but may not be mandatory to induce tissue damage.

tion and the intrinsic coagulation pathway are instrumental [62,63]. When evaluating antithrombotic pharmaceuticals to limit tissue injury using the photothrombosis paradigm, one should therefore be cautious because ensuing platelet-containing thrombosis may be present, but may not be mandatory to induce tissue damage. Models of embolic cerebral ischemia The first embolic stroke model was developed in dogs [64] and has subsequently been adapted to rats [65] and mice [66]. After extensive methodological modifications and novel technical developments, various experimental models of embolic cerebral ischemia exist today. Endothelial injury of large vessels, for example by irradiation of the common carotid artery after administration of a photosensitive dye [67], leads to local thrombus formation and embolic stroke in the distal vascular territory. Other stroke models use embolic materials, such as polyvinyl compounds, latex microspheres, or blood thrombi that are prepared ex vivo and injected into the common carotid artery [68-70] or into the proximal middle cerebral artery [66,71-73]. Thrombus formation can also be induced at the origin of the middle cerebral artery by local thrombin injection [74].

ls, such as polyvinyl compounds, latex microspheres, or blood thrombi that are prepared ex vivo and injected into the common carotid artery [68-70] or into the proximal middle cerebral artery [66,71-73]. Thrombus formation can also be induced at the origin of the middle cerebral artery by local thrombin injection [74]. The main advantages of embolic stroke models are their pathophysiological relevance - embolic vessel occlusion is the most frequent cause of ischemic stroke in humans - and, if blood-borne thrombi are used, the possibility to test thrombolytic compounds. Those models not placing emboli directly in the proximal middle cerebral artery, however, often show high variability of infarct location, distribution, and size (multifocal stroke model). Spontaneous lysis of thrombi can occur and is difficult to assess. Moreover, placement or induction of emboli in the proximal middle cerebral artery is technically demanding. A common complication, especially in mice, is subarachnoid hemorrhage, reported in about 40% of mice [66]. In emboligenic endothelial lesions of the common carotid artery, lesion type (endothelial denudation versus photothrombosis) may determine the fibrin content of the thrombi, and thus their embolic potential [75].

A common complication, especially in mice, is subarachnoid hemorrhage, reported in about 40% of mice [66]. In emboligenic endothelial lesions of the common carotid artery, lesion type (endothelial denudation versus photothrombosis) may determine the fibrin content of the thrombi, and thus their embolic potential [75]. As most experimental stroke models address the anterior circulation, models of vertebrobasilar occlusion are sparse and have been developed in larger animals such as rabbits, cats, and dogs. Recently, however, the autologeous thromboembolic stroke model [68] has been adapted to the posterior (vertebrobasilar) circulation in rats [76] allowing the study of stroke pathophysiology in the brainstem and cerebellum that may differ from the situation in the anterior circulation. Potential reasons for translational failures and proposed actions Species differences There are, of course, significant physiological, neuroanatomical and metabolic differences between humans and small rodents, which are the most widely used experimental animals in preclinical stroke research. For example, small rodents usually require higher drug doses on a mg/kg body weight basis for a similar effect than larger mammals [77]. Thus, effective doses derived from preclinical stroke studies in small rodents cannot simply be transferred to the situation in humans, even if adjusted for body weight. Dose-response curves in laboratory animals and in humans can be helpful to address this problem.

basis for a similar effect than larger mammals [77]. Thus, effective doses derived from preclinical stroke studies in small rodents cannot simply be transferred to the situation in humans, even if adjusted for body weight. Dose-response curves in laboratory animals and in humans can be helpful to address this problem. Moreover, it should be kept in mind that mice and rats are lissencephalic. Therefore, if a drug has been effective in preclinical stroke studies in small rodents, it is recommended to reproduce the result in higher species (in particular non-human primates; for review, see Fukuda and del Zoppo [78]) prior to the initiation of clinical trials [6]. These animals are gyrencephalic and more closely resemble the situation in humans. Strain differences Strain differences in mice [25,41-43] and rats [44-46] must not be underestimated and imply that results in one strain may not necessarily be reproducible in another strain. The profound genetic differences and phenotypic variance between mouse strains can explain that certain isogenic strains poorly mimic human disease or even that the effects of a targeted mutation are overshadowed, which has raised concern about the common practice of using a single mouse strain, mostly C57BL/6, to address many questions [79]. In transgenic studies in particular, not only the single gene mutation but also the genetic background must be taken into account [79]. Thus, preclinical studies of interventions in stroke should include more than one rat or mouse strain.

e of using a single mouse strain, mostly C57BL/6, to address many questions [79]. In transgenic studies in particular, not only the single gene mutation but also the genetic background must be taken into account [79]. Thus, preclinical studies of interventions in stroke should include more than one rat or mouse strain. Sex, age, and comorbidities Preclinical stroke experiments are often restricted to juvenile male animals to avoid the variability caused by female hormone cycling. It was assumed that pathophysiology and response to therapy seen in male animals would also apply to the other sex. In various experimental stroke models, however, young female adult rodents had smaller infarcts than their male counterparts [80-82]. In rats, there seems to be a correlation to the estrous cycle, with high endogenous estradiol levels in proestrus being correlated with smaller infarcts than low endogenous estradiol levels in metestrus [83]. In human observational studies, premenopausal women have a smaller incidence of ischemic stroke than men; and stroke risk in both genders increases with age [84].

rous cycle, with high endogenous estradiol levels in proestrus being correlated with smaller infarcts than low endogenous estradiol levels in metestrus [83]. In human observational studies, premenopausal women have a smaller incidence of ischemic stroke than men; and stroke risk in both genders increases with age [84]. There is a complex and significant influence of age on stroke outcome. Most studies have reported more severe consequences of cerebral ischemia in aged as compared to young animals. For example, older rats have been shown to develop larger infarcts [85,86]. Others, however, have reported contrary results with a more favorable outcome in aged animals [45,87]. The correlation between age and brain damage may not be linear [88]. Aging is associated with microvascular changes in laboratory rodents [89], and a small decline of cerebral blood flow has been observed [90]. Rats suffering from streptozotocin-induced diabetes and spontaneously hypertensive rat strains have been shown to develop larger infarcts [45], illustrating the impact of comorbidities. Taken together, the influence of species, strain, sex and age (and, if applicable, comorbidities) on experimental stroke must be taken into account. Whereas preclinical stroke research often uses healthy male juvenile rodents, a considerable number of stroke patients in clinical studies will be female, elderly, or suffer from comorbidities such as diabetes, hypertension and atherosclerosis.

plicable, comorbidities) on experimental stroke must be taken into account. Whereas preclinical stroke research often uses healthy male juvenile rodents, a considerable number of stroke patients in clinical studies will be female, elderly, or suffer from comorbidities such as diabetes, hypertension and atherosclerosis. Stroke model Selection of the most appropriate stroke paradigm is critical. Models requiring craniotomy, such as proximal middle cerebral artery ligation, are traumatic and do not mimic human stroke very closely. MCAo by an intraluminal thread and models using blood clot emboli are more similar to the clinical situation. Because human stroke is heterogeneous, however, there can be no single ideal stroke model. The different experimental stroke paradigms vary with respect to the underlying pathophysiological mechanisms and the size or location of the lesion. It is thus prudent to choose an approach that matches the anticipated situation in stroke patients as closely as possible. For example, thromboembolic models using blood clot emboli are helpful when the efficacy of thrombolytic drugs is to be examined. Ideally, the results should be reproducible in different models of focal ischemic stroke and a novel substance should be tested both with and without reperfusion [6].

as closely as possible. For example, thromboembolic models using blood clot emboli are helpful when the efficacy of thrombolytic drugs is to be examined. Ideally, the results should be reproducible in different models of focal ischemic stroke and a novel substance should be tested both with and without reperfusion [6]. Not only should the stroke model reflect the clinical situation, but the mechanism of action of a putative stroke drug or intervention should also be relevant for human stroke pathophysiology and, thus, has to be elucidated prior to the initiation of clinical studies. It is, for example, important that potential neuroprotective drugs are able to penetrate the blood-brain barrier. Cerebral ischemia causes disruption of the blood-brain barrier, but this occurs only several hours after stroke onset [91].

ology and, thus, has to be elucidated prior to the initiation of clinical studies. It is, for example, important that potential neuroprotective drugs are able to penetrate the blood-brain barrier. Cerebral ischemia causes disruption of the blood-brain barrier, but this occurs only several hours after stroke onset [91]. Anesthesia For practical and ethical reasons, experimental focal stroke has to be induced under appropriate anesthesia and analgesia. Any type of anesthesia, including inhalation anesthesia, can, however, alter stroke outcome [92]. Barbiturates have various side effects including hypothermia and may mediate neuroprotection [93,94], thus interfering with putatively neuroprotective interventions. Concern has also been raised about the use of ketamine that could overestimate the neuroprotective effect of an intervention [95]. Inhalation anesthesia is superior to intraperitoneal or intravenous administration of anesthetics concerning the control of the depth of anesthesia [96]. Even if inhalation anesthesia is used, however, animals under general anesthesia that breathe spontaneously have been shown to exhibit larger infarct volumes and a higher variability of physiological parameters than those endotracheally intubated and mechanically ventilated [96].

ontrol of the depth of anesthesia [96]. Even if inhalation anesthesia is used, however, animals under general anesthesia that breathe spontaneously have been shown to exhibit larger infarct volumes and a higher variability of physiological parameters than those endotracheally intubated and mechanically ventilated [96]. Potential pitfalls may thus be avoided by selection of appropriate anesthesia. Mechanical ventilation seems to reduce possible side effects of anesthesia on stroke outcome best, but may be too laborious to be applied routinely. Inhalation anesthesia should be preferred to intravenous or intraperitoneal administration of anesthetics, and it seems prudent to avoid barbiturates and ketamine. Flaws in basic study design Only a minority of published experimental stroke studies have reported randomized treatment allocation and blinding of investigators [15]. In a metaepidemiologic approach examining 13 meta-analyses of experimental stroke studies, those studies with unblinded induction of ischemia reported considerably greater effects than blinded studies demonstrating bias in the design of preclinical stroke studies [97]. Thus, basic universal standards of scientific research such as blinding of investigators should be strictly followed in modeling cerebral ischemia. This should also include confirmation of results by different independent laboratories prior to the initiation of clinical studies.

sign of preclinical stroke studies [97]. Thus, basic universal standards of scientific research such as blinding of investigators should be strictly followed in modeling cerebral ischemia. This should also include confirmation of results by different independent laboratories prior to the initiation of clinical studies. Therapeutic time window There is a therapeutic time window in ischemic stroke [98]. It is defined as the time interval during which an intervention can lead to partial or total recovery by restoration of perfusion (reperfusion window) or by protection of penumbra tissue from necrosis (neuroprotective window). It has also been noted that a rigid and universal time window for acute stroke therapy does not exist and that the specific pathophysiological state must be taken into account to assess the individual therapeutic potential [99].

rfusion window) or by protection of penumbra tissue from necrosis (neuroprotective window). It has also been noted that a rigid and universal time window for acute stroke therapy does not exist and that the specific pathophysiological state must be taken into account to assess the individual therapeutic potential [99]. The therapeutic time window in ischemic stroke may differ critically between species, and, in addition, depends on the mechanism of drug action and is influenced by factors such as temperature or collateral circulation. The time point at which a certain drug is administered during stroke development should thus be carefully selected and has to balance maximum efficacy on the one hand and clinical applicability on the other hand. Early drug administration, or even pretreatment, is usually the most promising approach regarding efficacy [100], whereas for routine clinical practice it is desirable to develop drugs that are still effective several hours, or even days, after stroke onset. To address these problems, in-depth pharmacokinetics in experimental animals and, if possible, in humans should be determined and compared. Physiological variables Physiological parameters can profoundly influence infarct development and outcome after ischemic stroke. Severe hyperglycemia has been shown to be detrimental and is associated with larger infarcts in animal stroke models [45,101]. A study using insulin after cerebral ischemia in rats found optimal blood glucose levels at approximately 6 to 7 mM and adverse effects of hypoglycemia below 3 mM [102].

after ischemic stroke. Severe hyperglycemia has been shown to be detrimental and is associated with larger infarcts in animal stroke models [45,101]. A study using insulin after cerebral ischemia in rats found optimal blood glucose levels at approximately 6 to 7 mM and adverse effects of hypoglycemia below 3 mM [102]. Brain temperature can also have a pronounced influence in experimental stroke [103]. An interesting example to illustrate the importance of temperature is the story of MK-801, an N-methyl-D-aspartate receptor antagonist. Several studies had shown neuroprotective efficacy of N-methyl-D-aspartate receptor antagonists in experimental stroke [104-107]. For MK-801 in particular, promising results from studies in the gerbil had been published [108,109]. It was, however, observed that animals developed prolonged hypothermia following treatment with MK-801. In 1990, researchers studied the effects of (a) MK-801 administration, (b) MK-801 administration in animals kept normothermic, and (c) hypothermia without drug administration [110,111]. Significant and similar effects were only found in (a) and (c). Thus it was demonstrated that the beneficial effects of MK-801 in experimental stroke, previously thought to be mediated by a specific mechanism involving N-methyl-D-aspartate receptor antagonism, could in fact be explained by hypothermia. Hypothermia can not only reduce infarct size, but may also alter infarct distribution [112].

demonstrated that the beneficial effects of MK-801 in experimental stroke, previously thought to be mediated by a specific mechanism involving N-methyl-D-aspartate receptor antagonism, could in fact be explained by hypothermia. Hypothermia can not only reduce infarct size, but may also alter infarct distribution [112]. Since arterial blood pressure determines the size of the area where cerebral blood flow is reduced, induced hypotension in animals subjected to focal cerebral ischemia may cause larger infarcts [113]. In line, induced hypertension during [114] and after [115] an episode of focal cerebral ischemia has been shown to ameliorate infarct size. In the rat intraluminal thread MCAo model, dependency of damage on blood pressure and on MCAo duration has been demonstrated [116]. Of course, the level of arterial blood gases (pO2, pCO2) and pH [96,117,118] can also influence the development of experimental stroke, but though these parameters are often monitored, it seems that their impact has not been studied systematically and extensively in rodent focal cerebral ischemia models.

strated [116]. Of course, the level of arterial blood gases (pO2, pCO2) and pH [96,117,118] can also influence the development of experimental stroke, but though these parameters are often monitored, it seems that their impact has not been studied systematically and extensively in rodent focal cerebral ischemia models. Given the impact of physiological variables such as body and in particular brain temperature, blood pressure, pH, blood gases, and glucose levels on stroke initiation, development and outcome, it can be helpful to monitor these parameters. To avoid prolonged operation times, however, recording of all relevant physiological variables in each individual animal may not be feasible. Except for experiments modeling comorbidities such as hypertension or diabetes, physiological variables should usually be kept within normal limits. By any means, special care must be taken to prevent hypothermia during and after surgery. For this purpose, heating pads, heated water baths, closed chambers with heating fans, and heating lamps are available and should be used routinely [119].

betes, physiological variables should usually be kept within normal limits. By any means, special care must be taken to prevent hypothermia during and after surgery. For this purpose, heating pads, heated water baths, closed chambers with heating fans, and heating lamps are available and should be used routinely [119]. Outcome analysis The outcome analysis should include infarct size, histology, mortality rate and frequency of complications such as subarachnoid hemorrhage, and functional (behavioral, motor and cognitive) scores. Since morphologically intact tissue does not always imply intact function and does not exclude delayed lesion development [120,121], endpoints should not be restricted to infarct size, but functional and neurological evaluation should also be performed. Moreover, histological, biochemical, and molecular endpoints should be studied to address the multiple spatially and temporally distributed molecular and cellular changes in the infarct core and penumbra that occur after focal cerebral ischemia [122]. Especially if tissue size varies (for example, due to extensive brain edema), infarct volumes or long-term atrophy should not be determined as absolute values (mm3), but as relative measures (percentage of the contralateral hemisphere) [116]. The extent of the lesion and functional deficits may change over prolonged time periods [123], whereas the duration of preclinical experiments is usually limited. It can thus be helpful to extend the time frame and study long-term outcome. An elegant way to study the kinetics of infarct development and to scan for possible hemorrhagic transformation, for example after therapeutic interventions, is the serial use of magnetic resonance imaging in individual animals [62,63]. Magnetic resonance imaging in laboratory animals subjected to experimental stroke can even detect diffusion/perfusion mismatch [124] or provide information on the metabolic state [125]. Mortality rate and frequency of complications should not only be recorded, but information should also be given on excluded animals. The exclusion of certain animals from the outcome analysis may completely alter results [11]. Deletion of outlier data can be reasonable, but exclusion and inclusion criteria have to be defined a priori.

nd frequency of complications should not only be recorded, but information should also be given on excluded animals. The exclusion of certain animals from the outcome analysis may completely alter results [11]. Deletion of outlier data can be reasonable, but exclusion and inclusion criteria have to be defined a priori. Publication of preclinical stroke studies It has been demonstrated that publication bias is evident in clinical stroke trials [126]. To address this problem, it has been suggested that all clinical trials should be registered and published irrespective of outcome. For a candidate neuroprotective drug, nicotinamide, it has been shown that publication bias also plays a role in preclinical stroke research: Those investigations published in abstract form showed a smaller effect of nicotinamide compared to those published as full articles [127]. This illustrates that results of preclinical stroke trials, whether positive or negative, should be completely reported and presented to the scientific community. If electronic journals are used, journal space must not be a limiting factor precluding the publication of negative results [128]. To reduce the risk of translational failure, all relevant information on preclinical studies should be available in peer-reviewed print or electronic journals when planning the design of clinical trials.

journals are used, journal space must not be a limiting factor precluding the publication of negative results [128]. To reduce the risk of translational failure, all relevant information on preclinical studies should be available in peer-reviewed print or electronic journals when planning the design of clinical trials. Coordination of preclinical and clinical research Evidently, animal models of human diseases cannot be more than an approximation to the situation in humans and are always limited by nature. In experimental stroke models, variables that influence the development of the ischemic brain lesion are controlled to ensure homogeneity, reliability and reproducibility of the results. These factors include age and sex of the animals studied, physiological parameters, and the experimental protocol. In contrast, patient selection criteria in clinical studies cannot achieve similar homogeneity as in preclinical experiments. Perhaps extreme homogeneity would not even be desirable, given that study results must be meaningful for everyday clinical practice. Study patients may differ in type, location and development of the ischemic stroke lesion, age, sex, comedication, comorbidities, physiological parameters, and time of application of the study medication. To account for the heterogeneity, large numbers of patients have to be included in clinical trials [129].

practice. Study patients may differ in type, location and development of the ischemic stroke lesion, age, sex, comedication, comorbidities, physiological parameters, and time of application of the study medication. To account for the heterogeneity, large numbers of patients have to be included in clinical trials [129]. Although these problems cannot be completely circumvented, they may be ameliorated by appropriate coordination of preclinical and clinical research. The design of basic investigations should consider future translation to the clinic from the very beginning. Conversely, if experimental treatment has been effective, clinical trials should adhere as closely as possible to the conditions under which the preclinical data were collected. There is evidence that this has been neglected in the past: of more than 1,000 potential neuroprotective drugs studied in animals as of 2006, at least 114 had been tested in acute stroke patients, yet, interestingly, those drugs studied in humans were not those that had shown superior efficacy in animal studies [3].

d. There is evidence that this has been neglected in the past: of more than 1,000 potential neuroprotective drugs studied in animals as of 2006, at least 114 had been tested in acute stroke patients, yet, interestingly, those drugs studied in humans were not those that had shown superior efficacy in animal studies [3]. Heterogeneity of patients in clinical stroke studies can be taken into account at the level of preclinical stroke research. Protocols solely using juvenile male animals may produce more homogeneous and robust results. Prior to the initiation of clinical studies, however, it seems to be prudent to perform additional experiments in female and aged animals or in strains bearing vascular risk factors such as obesity, hypertension, or diabetes. Indeed, it has been demonstrated that the use of healthy animals instead of animals with comorbidities overstates the effects of interventions in experimental stroke research [97]. Considerations and proposed actions for the design of preclinical stroke studies from this review are summarized in Appendix 1. They overlap and partly extend the STAIR recommendations for preclinical stroke drug development [6,13,14] and are intended to help avoiding procedural pitfalls and translational problems. It seems unlikely that there is one "ideal" preclinical stroke study evaluating a stroke drug or intervention comprehensively. Usually, several different experiments will address different issues.

linical stroke drug development [6,13,14] and are intended to help avoiding procedural pitfalls and translational problems. It seems unlikely that there is one "ideal" preclinical stroke study evaluating a stroke drug or intervention comprehensively. Usually, several different experiments will address different issues. Summary The presence of intact blood-perfused and then temporarily or permanently occluded vasculature is essential for the study of ischemic stroke. Research in healthy human subjects or in stroke patients is very limited. Moreover human stroke may be very heterogeneous; there may be many confounding variables, and it may not be possible to assess physiological and outcome parameters rigorously enough. Therefore, animal models, and in particular rodent models, play a pivotal role in stroke research in addition to in vitro techniques such as brain slices, tissue and cell culture. Rodent models of focal cerebral ischemia have added tremendously to our understanding of the etiology and pathophysiology of ischemic stroke. The translation of therapeutic approaches from the animal model to patient care, however, has been less fruitful. With the notable exception of intravenous thrombolysis with tissue plasminogen activator [4,130], the overwhelming majority of drugs showing promise in preclinical research have failed in subsequent clinical trials. All available stroke models have inevitable shortcomings and problems. Lessons to improve preclinical research methodology and to better coordinate preclinical and clinical stroke research have been derived and should be followed to reduce the risk of future translational failures.

failed in subsequent clinical trials. All available stroke models have inevitable shortcomings and problems. Lessons to improve preclinical research methodology and to better coordinate preclinical and clinical stroke research have been derived and should be followed to reduce the risk of future translational failures. Competing interests The authors declare that they have no competing interests. Authors' contributions SB and CK contributed equally. Appendix 1 - Synopsis of considerations and proposed actions to improve the design of preclinical stroke studies 1) Select species. Small rodents such as mice and rats should be used initially (advantages: ethics, costs, practicability). After promising results in other species, studies in non-human primates (advantages: gyrencephalic, closer to human situation) should be considered. 2) Select strain. Ideally, results (both positive and negative) should be confirmed in other strains. 3) Consider animal studies in juvenile male animals (advantages: homogeneity, reproducibility). Perform additional studies in female, aged or comorbid animals (advantage: closer to clinical situation).

Appendix 1 - Synopsis of considerations and proposed actions to improve the design of preclinical stroke studies 1) Select species. Small rodents such as mice and rats should be used initially (advantages: ethics, costs, practicability). After promising results in other species, studies in non-human primates (advantages: gyrencephalic, closer to human situation) should be considered. 2) Select strain. Ideally, results (both positive and negative) should be confirmed in other strains. 3) Consider animal studies in juvenile male animals (advantages: homogeneity, reproducibility). Perform additional studies in female, aged or comorbid animals (advantage: closer to clinical situation). 4) Select stroke model. The selected stroke model should be pathophysiologically relevant for the anticipated clinical situation and match the putative mechanism of the drug or intervention studied (for example, embolic stroke model using blood-borne thrombi for evaluation of thrombolytics). Stroke models with reperfusion (advantage: more common in human stroke) should be preferred over permanent ischemia models, or, ideally, a putative drug or intervention should be evaluated in both transient and permanent stroke models. 5) Select anesthetic and anesthesia method. Inhalation anesthesia should be preferred to intravenous or intraperitoneal drug administration. The potential interference of anesthetics with the stroke drug or intervention studied should be considered. In particular, barbiturates should be avoided.

4) Select stroke model. The selected stroke model should be pathophysiologically relevant for the anticipated clinical situation and match the putative mechanism of the drug or intervention studied (for example, embolic stroke model using blood-borne thrombi for evaluation of thrombolytics). Stroke models with reperfusion (advantage: more common in human stroke) should be preferred over permanent ischemia models, or, ideally, a putative drug or intervention should be evaluated in both transient and permanent stroke models. 5) Select anesthetic and anesthesia method. Inhalation anesthesia should be preferred to intravenous or intraperitoneal drug administration. The potential interference of anesthetics with the stroke drug or intervention studied should be considered. In particular, barbiturates should be avoided. 6) Include basic scientific standards such as randomization and blinding of investigators; define inclusion (for example, postoperative neurological score indicating clinical deficit) and exclusion criteria a priori. 7) Determine pharmacokinetics and dose-response curves. From these, information on dosage and therapeutic time window is derived. 8) Monitor and control relevant physiological parameters, at least account for the profound effects of brain temperature. 9) Consider infarct size, functional/neurological scores, histological, biochemical and molecular evaluation for outcome analysis. The outcome analysis should at least not be restricted to only infarct size or histology.

8) Monitor and control relevant physiological parameters, at least account for the profound effects of brain temperature. 9) Consider infarct size, functional/neurological scores, histological, biochemical and molecular evaluation for outcome analysis. The outcome analysis should at least not be restricted to only infarct size or histology. 10) Ideally, study both short-term and long-term outcome. Serial magnetic resonance imaging examinations may be helpful. 11) Elucidate mechanism of stroke drug or intervention studied. The mechanism should be relevant for human stroke. 12) Consider combination therapy (for example, putative neuroprotective drug plus thrombolysis). 13) Publish both positive and negative results. 14) Confirm results in other laboratories. 15) Coordinate preclinical and clinical research. The anticipated clinical situation should be considered in preclinical study design; clinical studies should adhere as closely as possible to the conditions in preclinical studies that showed efficacy.

Editorial It is with great pleasure and enthusiasm that we take the opportunity to introduce the new open access, peer-reviewed journal Experimental & Translational Stroke Medicine published by BioMed Central. As its title suggests, the journal is intended principally to gather and disseminate new knowledge in the field of experimental stroke in order to facilitate future translation and development of new clinical stroke treatments. There are many examples in modern medicine of translational research opening new perspectives for the improved diagnosis or treatment of human diseases and enabling findings from basic bench top discoveries to be transferred into clinical applications. A good example for this approach in the field of cerebrovascular disease is diffusion weighted magnetic resonance imaging. Introduced in the last decade of the 20th century in experimental stroke studies, this novel diagnostic tool was rapidly and successfully translated to humans, revolutionized early diagnosis of stroke and represents now the gold standard of modern stroke imaging worldwide. Contrary to this success in stroke diagnosis, the translation of novel experimental therapies into an effective treatment for stroke patients has so far not been successful for a number of reasons, with one central problem certainly rooted in insufficient stringency of the translational process from bench to bedside. A clear goal of the journal is therefore to improve the quality of experimental stroke research and reduce translational failures from bench to bedside.

r not been successful for a number of reasons, with one central problem certainly rooted in insufficient stringency of the translational process from bench to bedside. A clear goal of the journal is therefore to improve the quality of experimental stroke research and reduce translational failures from bench to bedside. The new journal particularly focuses on translational aspects of pathophysiological, diagnostic and therapeutic issues of cerebrovascular diseases. We encourage authors to submit full-length original articles, short reports or reviews related to the topics of emerging therapies, diagnostics, models, animal behaviour testing, pathophysiology, and clinical trial data. Moreover, guest articles summarizing translational aspects of other neurological diseases apart from stroke may occasionally be featured. Following evaluation by the Editorial Board [1] and peer-review by two independent experts in the field, all accepted articles will be published online immediately and soon after listed in PubMed. We believe that the open access model featured by BioMed Central is ideally suited to rapidly and broadly disseminating novel findings, and promoting fruitful discussion among active stroke researchers, with the ultimate goal of providing recommendations and guidelines aimed at improving experimental and translational stroke research.

the open access model featured by BioMed Central is ideally suited to rapidly and broadly disseminating novel findings, and promoting fruitful discussion among active stroke researchers, with the ultimate goal of providing recommendations and guidelines aimed at improving experimental and translational stroke research. Fittingly, the launch articles provide detailed reviews on promising stroke drug candidates which are currently in late clinical development, such as the hematopoietic factors G-CSF [2] and EPO [3]. Gee and colleagues address long term immunologic consequences of experimental stroke and mucosal tolerance. They provide clear-cut evidence that although induction of immunological tolerance to MBP improves outcome after stroke, mucosal administration of antigen could also account for detrimental autoimmunity in the long run [4]. The guest article by R. Linker provides an in depth review on multiple sclerosis models away from the old experimental autoimmune encephalitis (EAE) concept towards the clinically more relevant conditional knockout models tailored to study the pathology of B cells, CD 8 cells, and inflammation as a pathological trigger for primary degeneration [5].

provides an in depth review on multiple sclerosis models away from the old experimental autoimmune encephalitis (EAE) concept towards the clinically more relevant conditional knockout models tailored to study the pathology of B cells, CD 8 cells, and inflammation as a pathological trigger for primary degeneration [5]. We strongly believe that the open access format of the new journal - a true novelty in the field of cerebrovascular research - has clear benefits for science, medicine and the general public: First, all articles are freely and universally accessible online, and so an author's work can be read by anyone at no cost. The easy and widespread availability of articles significantly enhances reading and citation of the results. Second, all accepted articles are immediately published with no delay and therefore allow particularly rapid dissemination of new results. Third, the BioMed Central format allows interactive discussion and anotation of articles providing an online tool for open discussion of data. Fourth, in contrast to other stroke journals, there is no size restriction for articles, and the type of data that can be attached to an article (movies, large image data etc.). Authors hold copyright for their work and grant anyone the right to reproduce and disseminate the article, provided that it is correctly cited. Finally, the journal's articles are archived in PubMed Central, the US National Library of Medicine's full-text repository of life science literature, and also in repositories at the University of Potsdam in Germany, at INIST in France and in e-Depot, the National Library of the Netherlands' digital archive of all electronic publications. This complies with the policies of a number of funding bodies including the Wellcome Trust, NIH and Howard Hughes Medical Institute.

d also in repositories at the University of Potsdam in Germany, at INIST in France and in e-Depot, the National Library of the Netherlands' digital archive of all electronic publications. This complies with the policies of a number of funding bodies including the Wellcome Trust, NIH and Howard Hughes Medical Institute. We hope that unlimited access to the latest information on stroke research published in Experimental & Translational Stroke Medicine will create new ideas and opportunities to improve the understanding of stroke and to find new treatment opportunities. We encourage the stroke research community to make full use of this very welcome resource. Competing interests The authors declare that they have no competing interests.

Introduction Granulocyte-colony-stimulating factor (G-CSF) was identified among a set of humoral factors on which the survival, proliferation, and differentiation of hematopoietic cells in cell culture assays is dependent [1,2]. After purification of the murine G-CSF more than 25 years ago its human analogue was discovered [1]. The complete species cross-reactivity [3] of the human and the murine G-CSF molecule exhibits a strong evolutionary conservation and emphasizes its importance for white blood cell regulation. A decade after its identification, G-CSF was approved by the FDA for prevention and treatment of chemotherapy-induced neutropenia and apheresis for hematopoietic transplantation [4,5]. Much interest focused on the use of G-CSF as a neuroprotective candidate when its infarct size reducing capabilities in animal stroke models were discovered in the year 2003 [6-8]. Beyond its initially as key protective mechanism assumed capability to mobilize bone marrow stem cells, a deeper understanding of G-CSF's action in stroke pathophysiology has been developed. This review focuses on the neuroprotective and neuroregenerative properties of G-CSF in animal models of focal cerebral ischemia. In addition, the evidence and efficacy from preclinical studies as the basis for current clinical trials is reviewed.

standing of G-CSF's action in stroke pathophysiology has been developed. This review focuses on the neuroprotective and neuroregenerative properties of G-CSF in animal models of focal cerebral ischemia. In addition, the evidence and efficacy from preclinical studies as the basis for current clinical trials is reviewed. Mechanisms of action of G-CSF in ischemic injury Mobilization of stem cells G-CSF's natural function of mobilizing stem cells from the bone marrow triggered initial explorations of its potential usefulness in stroke with the idea that mobilized stem cells may home into the injured brain [9]. A series of preclinical investigations in animals using G-CSF for the therapy of ischemic stroke was initiated to answer the question whether mobilized bone marrow cells contribute to improved outcome [9-11]. The capacity of bone-marrow derived cells to restore function in the injured brain has indeed been demonstrated (for review see [12]), but the mechanism of their advantageous action remains unclear. The proposed transdifferentiation of bone marrow derived cells into neural cells that induce functional and structural recovery poststroke was recently doubted by several studies (e.g. [13,14]). The assumption that G-CSF mobilized bone marrow cells might have caused the observed functional improvements was also propagated by Shyu et al [11]. However, dividing cells in the ischemic hemisphere, mainly seen in the subventricular zone, were presumably originated from adult neural stem cells concordantly with results from other groups [7,10]. Komine-Kobayashi and colleagues [15] subjected chimeric mice with EGFP-expressing bone marrow-derived cells to transient occlusion of the middle cerebral artery. The authors report that indeed migration of bone-marrow derived monocytes was not increased at all after G-CSF treatment, but rather decreased. So far not evidence proven is another bone marrow cell mediated mechanism. G-CSF may induce invasion of bone-marrow-derived stem cells into the infarcted brain which could contribute to enhanced neuro- and angiogenesis by secretion of neurotrophines and other trophic factors [12]. In conclusion the recent evidence from animal experiments cast doubt on the perception that mobilization of stem cells is the sole or even most important mechanism of action for functional recovery after G-CSF treatment.

to enhanced neuro- and angiogenesis by secretion of neurotrophines and other trophic factors [12]. In conclusion the recent evidence from animal experiments cast doubt on the perception that mobilization of stem cells is the sole or even most important mechanism of action for functional recovery after G-CSF treatment. Anti-apoptotic activity A first indication of a potential direct effect on cells of the brain came from the observation that G-CSF had a direct protective effect in cultured neurons against glutamate-induced cell death [6]. After cerebral ischemia, endogenously released G-CSF is presumably active on the upregulated G-CSF receptor in periischemic regions at risk, the so called penumbra, and may provide protection against apoptotic cell death in neurons (figure 1). Schneider showed that after interaction with its receptor, G-CSF activates through JAK signalling, three independent anti-apoptotic pathways: The signal transducer and activation of transcription (STAT)-3, the extracellular-signal-regulated kinase (ERK) and the phophatidylinositol 3-kinase (PI3K)-Akt pathway [7]. Komine-Kobayashi also found antiapoptotic effects of G-CSF on neurons after cerebral ischemia through the JAK/STAT signaling pathway and subsequent activation of Bcl-2 [15]. Moreover, G-CSF increased cIAP2 levels in the ischemic cortex and thereby decreased the activation of caspase 3, an important trigger of apoptotic processes [16]. In a rat model of intracerebral hemorrhage G-CSF's antiapoptotic activity in cells in the perihematomal area was revealed by a TUNEL assay, which detects less DNA fragments as a result from apoptotic signaling cascades after G-CSF treatment [17].

ctivation of caspase 3, an important trigger of apoptotic processes [16]. In a rat model of intracerebral hemorrhage G-CSF's antiapoptotic activity in cells in the perihematomal area was revealed by a TUNEL assay, which detects less DNA fragments as a result from apoptotic signaling cascades after G-CSF treatment [17]. Figure 1 G-CSF reduces infarct sizes and enhances functional recovery in stroke models by several mechanisms of action, such as the induction of anti-apoptotic pathways, neurogenesis and angiogenesis. Thereby G-CSF acts as a direct protectant for neurons expressing its receptor. G-CSF's influences on immunocompetence and inflammation parameters are potential additional effects.

recovery in stroke models by several mechanisms of action, such as the induction of anti-apoptotic pathways, neurogenesis and angiogenesis. Thereby G-CSF acts as a direct protectant for neurons expressing its receptor. G-CSF's influences on immunocompetence and inflammation parameters are potential additional effects. Neurogenesis Neural progenitor cells reside for a lifetime in certain areas of the brain, particularly the subventricular zone (SVZ), the olfactory bulb and the hippocampus. Certain conditions such as stroke induce the generation of new neurons from precursor cells, a phenomenon which may potentially be utilized to restore brain function. G-CSF's most striking effect regarding neurogenesis was seen in the dentate gyrus, where the number of newly generated neurons under ischemic conditions [7,10,11] but also in nonischemic, sham-operated animals was increased. In the striatum there was only a trend toward an enhanced neurogenesis after G-CSF treatment which was not statistically significant and the number of newly generated cells was rather small [7]. This finding is not surprising, since the striatum is known to habor only a low number of neuronal precursors [18,19]. Generation of new differentiated cells from endogenous stem cells is an intricate interplay among different components such as proliferation, differentiation, and selective survival. In vivo experiments revealed that G-CSF promotes neurogenesis in all of these components. The number of newly generated cells was increased, the cells differentiate towards a neuronal fate and anti-apoptotic pathways are activated [7]. The in vivo findings of an increased neurogenesis after G-CSF treatment were confirmed by in vitro experiments. It was shown that adult neural stem cells isolated from the rat SVZ or hippocampal region that grow as neurospheres in culture express the G-CSF receptor [7,20]. G-CSF dose-dependently induced maturation of cultured progenitor cells towards a neuronal phenotype and increased the population of the differentiated cells [7].

was shown that adult neural stem cells isolated from the rat SVZ or hippocampal region that grow as neurospheres in culture express the G-CSF receptor [7,20]. G-CSF dose-dependently induced maturation of cultured progenitor cells towards a neuronal phenotype and increased the population of the differentiated cells [7]. Angiogenesis Angiogenesis is a process where new vessels arise from pre-existing ones [21]. Future treatment strategies in stroke focus on optimisation of this process in the ischemic boundary zone [22]. However, the contribution of angiogenesis to functional recovery after stroke is still unclear [23-25]. Lee and colleagues showed that G-CSF enhanced angiogenesis in a rat stroke model measured by endothelial cell proliferation, the vascular surface area, the number of branch points, and the vascular length [26]. The G-CSF effect was more pronounced when treatment was initiated earlier. But even when treatment was delayed up to seven days after the induction of ischemia an increased angiogenesis accompanied by an enhanced long-term functional recovery could be observed [26]. Expression of the vascular marker von Willebrand factor in BrdU positive cells after G-CSF treatment demonstrated the generation of new endothelial cells [11]. Taguchi found an accelerated angiogenesis measured by an angiographic score without an enhanced functional outcome [27]. However, the results of this study have to be interpreted with caution since immunodeficient mice were used in which G-CSF may not exert its immunomodulative properties. The immunomodulative effects are presumably important for post-stroke functional recovery, as described below.

ut an enhanced functional outcome [27]. However, the results of this study have to be interpreted with caution since immunodeficient mice were used in which G-CSF may not exert its immunomodulative properties. The immunomodulative effects are presumably important for post-stroke functional recovery, as described below. Immunomodulation Recent research revealed that interactions between cerebral ischemia and the immune system are exceptionally relevant for the functional outcome of stroke patients [28]. Stroke induced immunodepression can cause infections, such as pneumonia, a frequent complication in stroke patients. However, immunodepression may potentially improve stroke outcome by alleviating the autoaggressive responses due to ischemia-induced exposure of central nervous system-specific antigens to the immune system [29-31]. Thus, immunomodulation and an increase of in immunocompetence may also be responsible for the rather acute effects of G-CSF[32] Indeed, a reduced infiltration of neutrophils and microglia in the ischemic hemisphere after G-CSF treatment was observed [15,26], whereas our group could not detect such a difference between the placebo group and the G-CSF group [6]. A further analysis of inflammatory cells in the ischemic hemisphere revealed a decreased activation of inducible nitric oxide synthase (iNOS)-positve microglia in animals treated with G-CSF [15]. As a consequence of the iNOS inhibition, a reduced nitrotyrosine production as a marker for nitrosadative stress was detected in NeuN positive cells [15]. In contrast to these immunohistochemistry and western blot findings there was no G-CSF induced reduction of iNOS on the mRNA level [33]. Administration of interleukin-1 beta is known to deteriorate cerebral ischemia and an interleukin-1 beta receptor antagonist may neutralize this effect [34]. Thus, the reduction of the ischemia induced interleukin-1 beta upregulation by G-CSF may contribute to infarct size reduction [16,33]. Recently, our group showed that G-CSF suppresses MMP-9, which is known to mediate inflammation, blood-brain barrier breakdown with subsequent edema formation and tissue injury in acute stroke [35].

duction of the ischemia induced interleukin-1 beta upregulation by G-CSF may contribute to infarct size reduction [16,33]. Recently, our group showed that G-CSF suppresses MMP-9, which is known to mediate inflammation, blood-brain barrier breakdown with subsequent edema formation and tissue injury in acute stroke [35]. Efficacy of G-CSF in stroke models G-CSF in experimental stroke and the STAIR criteria Successful testing of a candidate stroke drug in animal models does not firmly predict efficacy in clinical studies [36]. As a result of many failed clinical stroke trials the Stroke Academic Therapy Industry Roundtable (STAIR) established recommendations for the preclinical evaluation of stroke drugs [37]. The STAIR criteria postulate that the efficacy of a new drug should be demonstrated in a variety of stroke models performed in different species and by different laboratories. Indeed, G-CSF showed efficacy in different species and different stroke models such as transient ischemia in mouse [9,15,38,39] and rat [6,7,11] as well as permanent ischemia in mouse [10,38] and rat [7,8]. Moreover, G-CSF's efficacy was investigated in animals with comorbitiy, such as diabetes and hypertension, which is important, since stroke patients usually exhibit those conditions [40,41]. As recommended by the STAIR functional outcome in animal should be tested besides measuring infarct size reduction. G-CSF demonstrated an improvement in short-term [7,15,38] and long-term [7,8,10,11,39] functional neurological deficits. Systemic parameters relevant for stroke pathophysiology such as blood pressure or oxygen saturation were not influenced by G-CSF as measured by physiological monitoring of animals subjected to stroke [6,7,15]. The overall high methodological quality of preclinical G-CSF stroke studies was corroborated by a recent systemic analysis [42]. Philip and colleagues found that animal experimental stroke studies of G-CSF had the highest quality in a STAIR guideline derived quality score compared to all other neuroprotective agents that are currently investigated in clinical phase II or III trial [42].

tudies was corroborated by a recent systemic analysis [42]. Philip and colleagues found that animal experimental stroke studies of G-CSF had the highest quality in a STAIR guideline derived quality score compared to all other neuroprotective agents that are currently investigated in clinical phase II or III trial [42]. Meta-analysis and meta-regression analysis of G-CSF in experimental stroke To enhance the chance of a successful transfer of preclinical data in clinical trials beyond the application of the STAIR criteria, systematic meta-analyses of candidate neuroprotectants in animal experiments were conducted [43-46]. To get an overall impression of G-CSF's efficacy in the recently published preclinical studies and for potential guidance of further clinical studies, we have performed a meta-analysis and meta-regression analysis of G-CSF in animal models of focal cerebral ischemia [47]. The meta-analysis showed that G-CSF effectively reduced both infarct volumes and sensorimotor deficits. Infarct sizes were reduced by 42%. The reduction of infarct volumes in G-CSF-treated animals was proportional to the infarct volumes of placebo-treated animals as indicated by the L'abbé plot [47]. This proportional infarct size reduction demonstrates G-CSF's efficacy in milder stroke models as well as in severe hemispheric stroke models. Sensorimotor deficits which were categorized in three subgroups (Rotarod running, neuroscore, limb function) were improved between 24% and 40%. Our meta-regression, which was the first meta-regression analysis of a neuroprotective drug in animal stroke models, identified higher doses of G-CSF to be associated with significantly smaller infarct volumes for doses between 10 and 60 μg/kg body weight (infarct size reduction 0.8% per one μg/kg body weight increase in dose when applied within the first 6 hours and 2.1% per one μg/kg body weight increase in dose when applied later than 6 hours after induction of ischemia). Time on Rotarod was significantly extended by 2.1% and 2.2% per one μg/kg body weight increase in dose for early and late treatment initiation, respectively. Also, limb function and neuroscore improved significantly when G-CSF dose was increased. This dose-response relationship is particularly important finding since conclusive experimental dose finding data deriving from a singular stroke study are currently not available. Also a critical aspect of stroke drug development is the therapeutic time window.

ore improved significantly when G-CSF dose was increased. This dose-response relationship is particularly important finding since conclusive experimental dose finding data deriving from a singular stroke study are currently not available. Also a critical aspect of stroke drug development is the therapeutic time window. For G-CSF effects on infarct size the time window is at least 24 hours in the transient suture occlusion model in rodents [9,11]. Regarding functional outcome Zhao [48] reported a beneficial effect of G-CSF when administered more than three month after the onset of ischemia. Using a meta-regression technique we found that a delayed treatment was as effective as an early treatment initiation and may even lead to smaller infarct sizes [47]. This result is particularly interesting since the time window for most candidate neuroprotectants closes early after symptom onset [49]. The potential of a much longer time-window of G-CSF compared to other stroke drugs might be explained by the above mentioned multimodal actions consisting of neuroprotective and particularly proregenerative properties.

ting since the time window for most candidate neuroprotectants closes early after symptom onset [49]. The potential of a much longer time-window of G-CSF compared to other stroke drugs might be explained by the above mentioned multimodal actions consisting of neuroprotective and particularly proregenerative properties. Conclusion Hematopoietic factors as candidate drugs for stroke treatment were intensely studied in stroke models over the last years. However, efficacy of candidate neuroprotectants in animal experiments may not necessarily predict efficacy in stroke patients, particularly when the preclinical experiments are insufficient and incomplete. New candidate drugs should therefore be tested in stroke models threefold to enhance the chance of a successful bench-to-bedside progress: 1. Meaningful interaction in stroke pathophysiology, 2. Integrity regarding fulfilled STAIR criteria, and 3. Efficacy analyzed in meta-analysis of animal studies. G-CSF, as novel candidate stroke drug, widely addresses these issues due to its multimodal mode-of-action in combination with a broad spectrum of efficacy in animal stroke models. Aside from this, G-CSF's has a comprehensive safety profile as demonstrated by its clinical use for many years.

-analysis of animal studies. G-CSF, as novel candidate stroke drug, widely addresses these issues due to its multimodal mode-of-action in combination with a broad spectrum of efficacy in animal stroke models. Aside from this, G-CSF's has a comprehensive safety profile as demonstrated by its clinical use for many years. Abbreviations G-CSF: Granulocyte-colony stimulating factor; STAIR: Stroke Therapy Academic Industry Roundtable; EGFP: enhanced green fluorescent protein; STAT: signal transducer and activation of transcription; ERK: extracellular-signal-regulated kinase; PI3K: phophatidylinositol 3-kinasel; Bcl-2: B-cell lymphoma 2; TUNEL: terminal uridine deoxynucleotidyl transferase dUTP nick and labeling; SVZ: subventricular zone; BrdU: bromdeoxyuridine; iNOS: inducible nitric oxide synthase. Competing interests WRS is inventor on a patent application regarding the neuroprotective effects of G-CSF. Authors' contributions JM wrote the manuscript. SS wrote the manuscript. WRS supervised manuscript preparation.

The cytokine erythropoetin (EPO) The cytokine erythropoietin (EPO) is a 34 kD glycoprotein which was originally described to stimulate erythropoiesis. EPO supports the proliferation and differentiation of erythroid progenitor cells and is critical for their survival [1]. The main site of EPO production is fetal liver and adult kidney [1]. Mice deficient for EPO or EPO receptor (EPOR) genes die at embryonic day 13 (E13) because of severe anemia caused by deficiency in definitive erythropoiesis [2-4]. Over the last decade it has become clear that EPO acts as growth and survival factors for multiple tissues expressing the EPO receptor [1]. The number of described targets of EPO action continues to grow.

nes die at embryonic day 13 (E13) because of severe anemia caused by deficiency in definitive erythropoiesis [2-4]. Over the last decade it has become clear that EPO acts as growth and survival factors for multiple tissues expressing the EPO receptor [1]. The number of described targets of EPO action continues to grow. EPO Receptor (EPOR) EPO acts by binding to its specific transmembrane receptor (EPOR). EPOR belongs to the single-chain cytokine type I receptor family [5]. These receptors are characterized by an extracellular N-terminal domain with conserved cysteines and a WSXWS-motif, a single hydrophobic transmembrane segment and a cytosolic domain with no intrinsic kinase activity [5]. Two transmembrane EPOR molecules form a homodimer that binds one EPO molecule leading to a conformational change and tight bonding of the two EPOR monomers which in turn activate two Janus family tyrosine kinase 2 (JAK2) molecules which associate with cytoplasmic domain of the EPOR. Once activated, JAK2 phosphorylates distal parts of receptors which subsequently serve as docking sites for downstream signaling molecules. Multiple signal transduction pathways are activated downstream of EPOR/JAK2 [1,5]. In neurons these include signal transducers and activators of transcription (Stat), phosphatidylinositol 3-kinase (PI3K)/Akt, Ras/extracellular signal regulated kinase (ERK1/2), nuclear factor-kappa-B (NF-κB), and calcium [6-8]. Best investigated from these are PI3K/Akt and Ras-MAPK pathways, both of which are important for the antiapoptotic and trophic effects of EPO [8-18]. NFκB pathway plays a role in antiapoptotic activity of EPO in neurons [7,19,20] as well as in EPO-mediated propagation of neural stem cells [21]. In addition EPO by activating phospholipase Cγ modulates intracellular calcium concentration, electrical activity and neurotransmitter release in PC12 cells [22-24] as well as in hippocampal neurons and brain slices [25,26]. The role of the Stat transcriptional factors in regard to EPO effects in the neural cells remains unclear; EPO induces phosphorylation of Stat5 in neurons [8,19,27], neural stem cells [21] and neuroblastoma cells [17,28]. Using primary hippocampal neurons from Stat5a/b knock-out mice we have recently shown that the presence of functional Stat5 is required for the trophic but not for the protective effect of EPO against glutamate-induced toxicity [9].

of Stat5 in neurons [8,19,27], neural stem cells [21] and neuroblastoma cells [17,28]. Using primary hippocampal neurons from Stat5a/b knock-out mice we have recently shown that the presence of functional Stat5 is required for the trophic but not for the protective effect of EPO against glutamate-induced toxicity [9]. EPO signaling is terminated by activation of phosphatases which dephosphorylate JAK2. The ligand-receptor complex is then internalized and degraded by the proteasome [1,5]. In hematopoietic cell lines 60% of the internalized EPO is re-secreted [29].

of Stat5 in neurons [8,19,27], neural stem cells [21] and neuroblastoma cells [17,28]. Using primary hippocampal neurons from Stat5a/b knock-out mice we have recently shown that the presence of functional Stat5 is required for the trophic but not for the protective effect of EPO against glutamate-induced toxicity [9]. EPO signaling is terminated by activation of phosphatases which dephosphorylate JAK2. The ligand-receptor complex is then internalized and degraded by the proteasome [1,5]. In hematopoietic cell lines 60% of the internalized EPO is re-secreted [29]. The brain EPO/EPOR system mRNA and protein of EPO and EPOR are detected in brain (hippocampus, internal capsule, cortex, midbrain), as well as in vitro in neurons, astrocytes, oligodendrocytes, microglia and cerebral endothelial cells [24,30-44] suggesting that this factor can function in the brain in a paracrine and/or autocrine manner. In the developing mouse brain expression of EPO and EPOR peaks during midgestation and decreases to adult levels in late gestation [43]. Brain specific ablation of EPOR leads to deficits in neural cell proliferation and neuronal survival in the embryonic brain and in post-stroke neurogenesis in the adult brain [45,46]. Expression of EPO and EPOR in the adult brain is stress-responsive and is regulated by oxygen supply: Both receptor and ligand expression is upregulated after hypoxia or ischemia [36,42,43,47-50]. Other stimuli such as hypoglycemia, insulin release, reactive oxygen species and insulin-like growth factor activate hypoxia-inducible factor and lead to increased expression of EPO [51,52]. Proinflammatory cytokines downregulate expression of EPO mRNA but increase that of EPOR in astrocytes [34]

emia [36,42,43,47-50]. Other stimuli such as hypoglycemia, insulin release, reactive oxygen species and insulin-like growth factor activate hypoxia-inducible factor and lead to increased expression of EPO [51,52]. Proinflammatory cytokines downregulate expression of EPO mRNA but increase that of EPOR in astrocytes [34] Based on the loss of some of the tissue protective effects of EPO and its non-hematopoietic derivative, the carbamylated EPO (CEPO) in mice lacking the common β chain shared by the members of the IL-3 receptor family, Brines and Cerami have proposed that the cytoprotective effects of EPO and CEPO are mediated by a heteromeric receptor complex comprised of one EPOR subunit and a dimer of the common β chain [6,53]. Immunoreactivity of the common β chain has been detected in brain tissue with a pattern reminiscent to that of the classical EPOR [53]. Furthermore, the common β chain can be coimmunoprecipitated with EPOR antibodies from the P19 embryonal carcinoma cells [53], but the existence of the proposed heteromeric receptor structure in primary cells or tissues has yet to be directly proven. In a recent study no expression of the common β subunit was detected in neuronal PC-12 cells even if EPO rescued these cells from staurosporine-induced apoptosis [28]. Interestingly, the classical EPOR is required for EPO-stimulated neuronal differentiation and survival but not for neurogenesis induced by CEPO suggesting that differential receptor interactions mediate the effects of EPO and CEPO in brain cells [28,54].

PO rescued these cells from staurosporine-induced apoptosis [28]. Interestingly, the classical EPOR is required for EPO-stimulated neuronal differentiation and survival but not for neurogenesis induced by CEPO suggesting that differential receptor interactions mediate the effects of EPO and CEPO in brain cells [28,54]. Neuroprotection by EPO is independent of hematopoesis The neuroprotective actions of EPO can be separated from its hematopoietic actions, a fact that is of value for therapeutic applications where the increase in red cell mass is not desired. EPO and EPO derivatives are directly neuroprotective in cell culture models (see below) and after direct application into the brain [36,55,56]. Moreover, CEPO and other EPO derivatives which do not bind to EPOR in myeloid cells and thus lack hematopoietic activity display potent tissue protective activities [57-59]. Expression of EPO and the classical EPOR in brain cells is induced by hypoxic-ischemic stress and contributes to ischemic tolerance [14,24,32,37,41,49,50,60-66] while neutralization of the brain endogenous EPO augments ischemic damage [56]. Brain-specific genetic ablation of the classical EPOR impairs post-stroke neurogenesis and neuronal survival [45,46] whereas transgenic brain specific overexpression of human EPO is associated with reductions in postischemic infarct volume, brain swelling and functional deficits in a transient stroke model [16].

[56]. Brain-specific genetic ablation of the classical EPOR impairs post-stroke neurogenesis and neuronal survival [45,46] whereas transgenic brain specific overexpression of human EPO is associated with reductions in postischemic infarct volume, brain swelling and functional deficits in a transient stroke model [16]. Multimodal neuroprotective profile EPO has been reported to induce a broad range of cellular responses in the brain directed to protect and repair tissue damage (Figure 1). EPO is neuroprotective in a variety of hypoxic, hypoglycemic, and excitotoxic in vitro models [7,8,10,14,17,18,20,21,42,45,50,57,67-72]. A fundamental mechanism of EPO-induced neuroprotection in cultured neurons is its ability to inhibit apoptosis reducing both DNA damage and cell membrane asymmetry [7,8,10,14,17,18,20,21,42,45,50,57,67-72]. Necrotic cell death (for example, after glutamate exposure) is also be attenuated by EPO [9,69,73]. Why astroglial cultures are protected by EPO from nitric oxide- and staurosporine-induced death but not from arsenic trioxide- or glutamate-induced toxicity is not fully understood [73,74]. Figure 1 Multimodal neuroprotective profile of erythropoietin (EPO). BBB - blood brain barrier; EC - endothelial cells. Another tissue-protective mechanism of EPO is its ability to protect cells against oxidative damage [75,76]. EPO inhibits lipid peroxidation by increasing the activities of cytosolic antioxidant enzymes such as superoxide dismutase and glutathione peroxidase [77-79].

Figure 1 Multimodal neuroprotective profile of erythropoietin (EPO). BBB - blood brain barrier; EC - endothelial cells. Another tissue-protective mechanism of EPO is its ability to protect cells against oxidative damage [75,76]. EPO inhibits lipid peroxidation by increasing the activities of cytosolic antioxidant enzymes such as superoxide dismutase and glutathione peroxidase [77-79]. EPO attenuates inflammation by reducing reactive astrocytosis and microglia activation and by inhibiting immune cells recruitment into the injured area [47,58,59,70,80-85]. In cerebrovascular endothelial cell cultures EPO down-regulates TNF-α-induced gene expression of interleukin-6 (IL-6), IL-1beta, CXCR4, and IL-1alpha [86]. It also directly counteracts interferon-γ- and lipopolyssaccharide-induced cytotoxicity in oligodendrocytes, preserves white matter [87] and reduces TNF-α release and its effects in cultured Schwann cells [88].

PO down-regulates TNF-α-induced gene expression of interleukin-6 (IL-6), IL-1beta, CXCR4, and IL-1alpha [86]. It also directly counteracts interferon-γ- and lipopolyssaccharide-induced cytotoxicity in oligodendrocytes, preserves white matter [87] and reduces TNF-α release and its effects in cultured Schwann cells [88]. EPO protects vascular integrity and stimulates angiogenesis [89-92]. It preserves blood-brain barrier integrity during injury by restoring expression of tight junction proteins [90,93,94], by reducing vascular inflammation [95] and reactive free radical expression [90,93,94,96]. In vasculogenesis EPO stimulates proliferation of endothelial precursor cells, production of matrix metalloproteinase-2, migration of endothelial cells into vascular sites and formation of capillary tubes [90-92,97,98]. EPO displays direct antiapoptotic activity in cerebral endothelial cells during oxidative stress and ischemic injury as well [91]. Stimulation of endothelial nitric oxide synthase (eNOS) activity has been shown to contribute to the improvements by EPO after experimental cerebral hemorrhage [99-101]. Interestingly, plasma and tissue levels of NO are markedly increased in transgenic rats overexpressing EPO [102] whereas the vascular protection by EPO is abolished in eNOS-deficient mice [103].

hase (eNOS) activity has been shown to contribute to the improvements by EPO after experimental cerebral hemorrhage [99-101]. Interestingly, plasma and tissue levels of NO are markedly increased in transgenic rats overexpressing EPO [102] whereas the vascular protection by EPO is abolished in eNOS-deficient mice [103]. EPO promotes differentiation towards neurons in several neuroblastoma cell lines, in neural stem cell cultures derived from both embryonic and adult neuronal germinal zones, as well as in embryonic neural progenitor-cell cultures [21,45,46,54,89,104-108]. Neuronal differentiation of adult neural progenitor cell by EPO seem to require activation of the sonic hedgehog signaling and up-regulation of suppressor of cytokine signaling-2 (SOCS2) [54,105]. EPO increases proliferation of oligodendrocyte progenitors and promotes differentiation of oligodendrocytes in culture [33,34]. EPOR-/- fetuses exhibit increased apoptosis in the brain and a reduction in the number of neural progenitor cells, as well as increased sensitivity to hypoxia prior to significant anaemia or heart defects in the embryo proper [42,45,46]. Moreover, adult mice that lack EPOR in the brain have significantly reduced neurogenesis in the subventricular zone and demonstrate impaired migration of precursors into infracted cortex [46]. Nevertheless, expression of EPO or EPOR on neural cells is not indispensable for brain development [42,45,46].

oper [42,45,46]. Moreover, adult mice that lack EPOR in the brain have significantly reduced neurogenesis in the subventricular zone and demonstrate impaired migration of precursors into infracted cortex [46]. Nevertheless, expression of EPO or EPOR on neural cells is not indispensable for brain development [42,45,46]. The reported neurotrophic effects of EPO include the ability to stimulate axonal regrowth, neurite formation, dendritic sprouting, electrical activity and modulate intracellular calcium and neurotransmitter synthesis and release [9,13,22,23,25,26,46,109-113]. A recent study demonstrated a calcium sensitive activation of cAMP response element binding protein (CREB) and induction of brain-derived neurotrophic factor (BDNF) gene expression by EPO in primary hippocampal neurons [25]. In rat hippocampal slices, EPO improved synaptic transmission during and following oxygen and glucose deprivation [13]. However, it has not been directly shown that EPO induces formation of new synapses.

nd induction of brain-derived neurotrophic factor (BDNF) gene expression by EPO in primary hippocampal neurons [25]. In rat hippocampal slices, EPO improved synaptic transmission during and following oxygen and glucose deprivation [13]. However, it has not been directly shown that EPO induces formation of new synapses. Animal Studies The preclinical data in support of the use of EPO in human brain disease have explosively increased since the first discovery of its neuroprotective action. In particular, the preclinical evidence in support for testing EPO in human acute ischemic stroke fulfills most of the STAIR criteria [114] such as testing by several laboratories using both temporary and permanent stroke models, post-treatment at several doses and exploration of therapeutic window, characterization of pharmacokinetic profile in respect to blood-brain-barrier penetration after peripheral administration, measurement of histological and functional outcome with prolonged survival.

using both temporary and permanent stroke models, post-treatment at several doses and exploration of therapeutic window, characterization of pharmacokinetic profile in respect to blood-brain-barrier penetration after peripheral administration, measurement of histological and functional outcome with prolonged survival. Cerebral ischemia The in vivo potential of EPO to protect neurons against ischemic neuronal damage was first provided by the Sasaki group. Their landmark finding was that application of recombinant human (rh)EPO directly into the cerebral ventricles of gerbils prevented ischemia-induced learning disabilities and protected hippocampal pyramidal CA1 neurons from lethal ischemia while neutralization of the endogenous brain EPO by infusions of soluble EPOR before a nonlethal mild ischemia induced neuronal death [56]. Since the circulating EPO, as a large, highly glycosylated negatively charged molecule, was thought not to cross the blood-brain-barrier [91,115,116], the early studies used direct intracerebroventricular route of administration of EPO to demonstrate its potent tissue protective activity in focal and global models of cerebral ischemia [36,55,56]. The first evidence for a neuroprotective effect of EPO by peripheral route of administration was provided by Brines et al. (2000) who demonstrated in a focal stroke model reduction of infarct volumes by intraperitoneally applied high dose rhEPO (5000 U/kg) up to 6 h after reperfusion. Immunohistochemical detection of biotinylated rhEPO 5 hours after its intraperitoneal injection at the therapeutically effective dose (5000 U/kg) further provided evidence that EPO can cross the blood-brain barrier [117]. Studies in several species including man have confirmed the ability for high dose EPO to cross the blood-brain barrier in therapeutic effective concentrations [118-122]. To date EPO and its derivatives have shown to reduce histological damage and improve functional outcome when given as intraperitoneal or even intranasal post-treatment after experimental stroke [57-59,70,85,89,123], global cerebral ischemia [124], neonatal stroke and hypoxia-ischemia [107,125-128]. For example, a comprehensive dosing study using post-treatment with EPO and CEPO starting at 6 h after an embolic middle cerebral artery occlusion in rats demonstrated reduction of functional deficits and infarct volume up to 28 days models of middle cerebral artery occlusion [59].

e and hypoxia-ischemia [107,125-128]. For example, a comprehensive dosing study using post-treatment with EPO and CEPO starting at 6 h after an embolic middle cerebral artery occlusion in rats demonstrated reduction of functional deficits and infarct volume up to 28 days models of middle cerebral artery occlusion [59]. Induction of EPO and its intracellular signaling intermediates represents a significant component of tolerance induced by ischemic or hypoxic preconditioning [60-62,65]. Here activation of the antiapoptotic and anti-inflammatory signaling seems to play a major role in the EPO-induced neuroprotection [14,19,129].

e and hypoxia-ischemia [107,125-128]. For example, a comprehensive dosing study using post-treatment with EPO and CEPO starting at 6 h after an embolic middle cerebral artery occlusion in rats demonstrated reduction of functional deficits and infarct volume up to 28 days models of middle cerebral artery occlusion [59]. Induction of EPO and its intracellular signaling intermediates represents a significant component of tolerance induced by ischemic or hypoxic preconditioning [60-62,65]. Here activation of the antiapoptotic and anti-inflammatory signaling seems to play a major role in the EPO-induced neuroprotection [14,19,129]. Cerebral hemorrhage Post-treatment with EPO starting at 2 h after induction of intracerebral hemorrhage (ICH) by intraparenchymal injections of collagenase or autologous blood dose-dependently reduced volume of hemorrhage, brain edema, perihematomal apoptosis and inflammation in a rat model [99]. Functional recovery was faster and more efficient in the EPO-treated group and was associated with reduction in hemispheric brain atrophy 5 weeks after the induction of ICH [99]. Cerebral vasospasm and ischemic brain damage after subarachnoid hemorrhage (SAH) by autologous blood injections into the cisterna magna in rabbits are reduced by EPO administered either by intraperitoneal injections of rhEPO or by delivery of adenoviral vectors encoding the human EPO into cisterna magna immediately after induction of SAH [130-133]. Mortality and functional deficits 3 days after induction of SAH were reduced in EPO treated rabbits [130-132]. In a rat model of SAH, the impaired autoregulatory response of cerebral blood flow to intravenous noradrenaline was restored by a single subcutaneous bolus of EPO [134].

ately after induction of SAH [130-133]. Mortality and functional deficits 3 days after induction of SAH were reduced in EPO treated rabbits [130-132]. In a rat model of SAH, the impaired autoregulatory response of cerebral blood flow to intravenous noradrenaline was restored by a single subcutaneous bolus of EPO [134]. Traumatic brain and spinal cord injury Administration of EPO and EPO-analogs in experimental models of traumatic brain and spinal cord injury leads to morphological, functional and cognitive recovery that can be attributed to a multitude of cytoprotective mechanisms including inhibition of apoptosis, anti-inflammatory and anti-oxidant actions, restoration of blood-brain barrier integrity, stimulation of neurogenesis and angiogenesis [67,95,96,104,117,135-146]. Induction of EPO and its protective down stream signaling via Akt seems also to account for the protective effect of heat acclimation stress in a closed head injury model [147,148]. Brain edema after experimental brain injury can effectively be attenuated by post-treatment with EPO [95,138,140]. A reduction of cytotoxic and vasogenic edema may be anticipated based on the direct actions of EPO on glutamate release [112] and on the endothelial barrier function (see above). It is not clear to date which from the panoply of neurorestorative effects of EPO are responsible for the long-term prevention of trauma-induced brain atrophy, cognitive and neurobehavioral dysfunction [104,135-137,146]. In this context it is interesting to note that chronic peripheral administration of EPO has been reported to improve spatial memory function and cognitive functioning in the context of an aversion task also in healthy mice [119,149]. Improved hippocampal functioning after a single intravenous bolus of EPO was recently shown in a study using functional magnetic resonance imaging in healthy human volunteers [150].

n reported to improve spatial memory function and cognitive functioning in the context of an aversion task also in healthy mice [119,149]. Improved hippocampal functioning after a single intravenous bolus of EPO was recently shown in a study using functional magnetic resonance imaging in healthy human volunteers [150]. Degeneration & neuroinflammation EPO and its analogs offer protection also in models of neurodegenerative and neuroinflammatory disease. In experimental autoimmunencephalitis (EAE), an animal model for multiple sclerosis (MS), treatment with EPO and EPO analogs can improve functional recovery, reduce tissue damage, inflammatory responses and blood-brain barrier leakage [47,80-84,117]. Beneficial effects of EPO have also been reported in models of peripheral axonal nerve injury, injury-induced Wallerian degeneration and HIV-associated sensory neuropathy [88,151,152]. Here, the anti-cytokine, anti-apoptotic, anti-oxidative and trophic effects on both neurons and oligodendrocyte progenitor cells by EPO seem to play an important role in reducing inflammation and preserving myelination and neuronal function [35,47,80-84,86-88,151].

on and HIV-associated sensory neuropathy [88,151,152]. Here, the anti-cytokine, anti-apoptotic, anti-oxidative and trophic effects on both neurons and oligodendrocyte progenitor cells by EPO seem to play an important role in reducing inflammation and preserving myelination and neuronal function [35,47,80-84,86-88,151]. Chronic neurodegeneration might also be a target for EPO therapy as EPO and its analogs can counteract degenerative processes in experimental models of Parkinson disease and amyotrophic lateral sclerosis (ALS) by inducing anti-oxidant enzymes, inhibiting apoptosis and stimulating axonal regeneration [78,153-155]. EPO improved graft survival of embryonic ventral mesencephalic dopamine neurons when transplanted into the striatum of 6-hydroxy-dopamine lesioned rats [156]. However, not all degenerative diseases seem to respond to EPO therapy [157].

dant enzymes, inhibiting apoptosis and stimulating axonal regeneration [78,153-155]. EPO improved graft survival of embryonic ventral mesencephalic dopamine neurons when transplanted into the striatum of 6-hydroxy-dopamine lesioned rats [156]. However, not all degenerative diseases seem to respond to EPO therapy [157]. Translation to human brain disease EPO and its receptor are abundantly expressed in the developing human brain decreasing to a weak constitutive expression in the adult [30,39,41,66,119,158]. Hypoxia rapidly induces expression of brain EPO as evidenced by the increased expression of EPO in cerebrospinal fluid (CSF) or postmortem brain tissue in humans with traumatic brain injury, SAH, stroke and hypoxia [31,41,66,121,159,160]. Expression of EPOR has also been detected on myelin sheath of radicular nerves and in the epineurial blood vessels of sural nerves in the human peripheral nervous system [161,162]. Measurements of endogenous levels of EPO in CSF of patients with neurodegenerative diseases has revealed EPO in CSF of patients with ALS to be lower than in controls whereas patients with Alzheimer disease (AD) or vascular dementia had EPO levels comparable to control persons [163,164]. Astrocytic EPOR immunoreactivity in postmortem hippocampus and temporal cortex from subjects with AD or chronic schizophrenia has been reported to be increased as compared to age-matched control brains [119,158]. In AD, however, no association of EPOR-positive astrocytes was found with summary measures of global cognition or AD pathology [158].

tivity in postmortem hippocampus and temporal cortex from subjects with AD or chronic schizophrenia has been reported to be increased as compared to age-matched control brains [119,158]. In AD, however, no association of EPOR-positive astrocytes was found with summary measures of global cognition or AD pathology [158]. Millions of patients have already been receiving EPO as a highly efficacious and safe treatment for anemia [1]. Indeed, the first proof-of-concept study on use of EPO in human acute ischemic stroke has already demonstrated that treatment of stroke patients with intravenous high dose EPO is not only well tolerated but is associated with improvement in clinical outcome at 30 days [120]. Encouraging results with the use of EPO as a neuroprotective agent have been recently reported in clinical pilot studies after out-of-hospital cardiac arrest [165], ureamia-associated peripheral neuropathy [166], chronic schizophrenia [167] and MS [145]. A small pilot study to probe the safety and efficacy of EPO in SAH was recently preliminarily terminated with no conclusive results because of low recruitment rate [168]. To date a randomized multicenter double blinded placebo-controlled clinical trial in acute ischemic stroke is running (for details see http://www.epo-study.de) reflecting the beginning of exploration of EPO and its analogs for clinical neuroprotection in large phase II/III setting. Competing interests The authors declare that they have no competing interests.

Millions of patients have already been receiving EPO as a highly efficacious and safe treatment for anemia [1]. Indeed, the first proof-of-concept study on use of EPO in human acute ischemic stroke has already demonstrated that treatment of stroke patients with intravenous high dose EPO is not only well tolerated but is associated with improvement in clinical outcome at 30 days [120]. Encouraging results with the use of EPO as a neuroprotective agent have been recently reported in clinical pilot studies after out-of-hospital cardiac arrest [165], ureamia-associated peripheral neuropathy [166], chronic schizophrenia [167] and MS [145]. A small pilot study to probe the safety and efficacy of EPO in SAH was recently preliminarily terminated with no conclusive results because of low recruitment rate [168]. To date a randomized multicenter double blinded placebo-controlled clinical trial in acute ischemic stroke is running (for details see http://www.epo-study.de) reflecting the beginning of exploration of EPO and its analogs for clinical neuroprotection in large phase II/III setting. Competing interests The authors declare that they have no competing interests. Authors' contributions NB drafted the manuscript and designed the Figure. ALS corrected and wrote the final manuscript. Both authors read and approved the final manuscript.

s that died before the endpoint of 24 hours were not excluded from MRI, but SIS analysis. The nonparametric Kruskall-Wallis test evaluated the neuroscore with subsequent group comparisons by Mann-Whitney U test. The Chi square test with Yates correction for small numbers was used to test for differences mortality rate. Results Physiological variables There were no significant differences in the physiologic variables (data not shown). Survival Survival rate was 75% for the control group, the early G-CSF group, the early rt-PA group, and the deco group. 83% of the animals survived in com group. In contrast, most animals died in the delayed rt-PA group which was treated by rt-PA 180 min after TE. Only 5 animals survived 24 hours in this subgroup which corresponds to a survival rate of 41%. Therefore, the survival rate was significantly lower compared to all other experimental groups (p < 0.05). Animals that died prematurely were assessed for the cause of death and showed all large infarcts in the brain territory supplied by the MCA. Therefore, large infarct was probably responsible for the death.

Granulocyte Colony-Stimulating Factor (G-CSF) is neuroprotective in models of acute experimental cerebral ischemia [1-9]. During the acute phase of ischemic stroke, various neuroprotective effects of G-CSF have been described in different species [1,2,9]. G-CSF influences apoptotic pathways [3,4], suppresses edema formation and interleukin-1 beta expression [1,6] induces the cerebral G-CSF receptor [7], and diminishes glutamate induced neurotoxicity [1,10]. Moreover, reduction of infarct size is associated with an improved functional score [6,8,11]. Remarkably, G-CSF reduced the infarct size even when given 72 hours after induction of cerebral ischemia [8]. In addition, G-CSF stimulates endogenous neurogenesis and vascularisation [6,7,9,11,12]. As a result, clinical studies are currently conducted to test the safety and effectiveness of G-CSF after acute ischemic stroke [9,11,13].

reduced the infarct size even when given 72 hours after induction of cerebral ischemia [8]. In addition, G-CSF stimulates endogenous neurogenesis and vascularisation [6,7,9,11,12]. As a result, clinical studies are currently conducted to test the safety and effectiveness of G-CSF after acute ischemic stroke [9,11,13]. Because of its multiple ways of action and good clinical tolerability for other medical conditions [9], G-CSF might be an ideal drug for the treatment of acute ischemic stroke. So far, thrombolysis by recombinant tissue-plasminogen activator (rt-PA) within the first 4.5 hours after symptom onset is the only proven effective therapy for thromboembolic stroke [14,15]. As recanalisation and neuroprotection are probably the most promising therapeutical approaches in stroke, combination of rt-PA and G-CSF need to be tested experimentally before using it in patient trials. G-CSF might decrease infarct volume when combined with rt-PA, but may interfere with potentially beneficial effects of rt-PA such as improvement of cerebral blood flow (CBF). Additionally, pretreatment by G-CSF might influence infarct volume and overall outcome after delayed and therefore potentially harmful reperfusion. In this context, the interaction of rt-PA associated pathways and simultaneous treatment by hematopoetic growth factors is of high clinical relevance, since combination treatment by erythropoetin (EPO) and rt-PA lead to increased side effects in a recently published randomized study [16]. It is very remarkable that no animal experiments have been published for combination of rt-PA and EPO before starting the clinical study. Therefore, we investigated whether G-CSF was neuroprotective in a model of thromboembolic stroke. Moreover, the combination therapy with rt-PA was evaluated to eludicate whether G-CSF alters infarct growth in combination therapy. Pretreatment with G-CSF before delayed rt-PA administration was investigated to evaluate whether reperfusion associated injury could be influenced.

ive in a model of thromboembolic stroke. Moreover, the combination therapy with rt-PA was evaluated to eludicate whether G-CSF alters infarct growth in combination therapy. Pretreatment with G-CSF before delayed rt-PA administration was investigated to evaluate whether reperfusion associated injury could be influenced. Materials and methods Animals and experimental groups The animal experiments were performed after approval of the animal care committee (Regierungspräsidium Karlsruhe, Germany). Before surgery male Wistar rats (n = 72) weighing 280 to 320 g (Charles-River Deutschland, Sulzfeld, Germany) were assigned to one of the following groups. • Group control (n = 12): Thromboembolic cerebral ischemia (TE). No specific treatment. • Group early G-CSF(n = 12): TE followed by intravenous G-CSF treatment after 60 min. • Group early rt-PA(n = 12): TE followed by intravenous rt-PA treatment 60 min after TE. • Group com(n = 12): TE followed by intravenous rt-PA treatment 60 min after TE and intravenous G-CSF treatment 60 min after TE. • Group delayed rt-PA(n = 12): TE followed by intravenous rt-PA treatment 180 min after TE. • Group deco(n = 12): TE followed by intravenous G-CSF treatment after 60 min and intravenous rt-PA treatment 180 min after TE. All animals were subjected to MRI monitoring including perfusion weighted imaging (PWI), diffusion weighted imaging (DWI), T2, and T2* at 0.5, 2.5, 4, and 24 hours after TE followed by silver-infarct staining (SIS) as described below.

showed that ADR1 is present in the cell body, axon and dendrites, but is absent from the nucleus, of cultured cortical neurons (Figure 1D). Addition of globular adiponectin to cultured neurons increased the expression of ADR1 at 6 and 12 hours post-OGD exposure, but had no effect on ADR2 expression levels (Figure 1E). Figure 1 Neurons express ADR1 and ADR2 and respond to oxygen and glucose deprivation. (A) ADR1 and ADR2 mRNA are present in cultured cortical neurons determined by single-cell RT-PCR analysis. The numbers indicate the number of neurons from which RNA was amplified; 1 and 3 cortical neurons consistently yielded a positive PCR signal for the ADR1 and ADR2 with exactly predicted size. (B) Cortical neurons subjected to OGD for the indicated times show increased levels of ADR1 and ADR2 mRNA in a time-dependent manner. (C) Immunoblot analysis of proteins in cell lysates of neurons in control cultures and cultures subjected to OGD for 3-24 h. OGD resulted in increased levels of ADR1 and ADR2 (D) ADR1 immunoreactivity (red) in cultured neurons; cells were counterstained with DAPI (blue) to label all nuclei. Arrow points to the axon of a neuron and arrowheads point to dendrites of the same neuron. (E) Cortical neurons subjected to OGD following globular adiponectin treatment show increased levels of ADR1.

• Group deco(n = 12): TE followed by intravenous G-CSF treatment after 60 min and intravenous rt-PA treatment 180 min after TE. All animals were subjected to MRI monitoring including perfusion weighted imaging (PWI), diffusion weighted imaging (DWI), T2, and T2* at 0.5, 2.5, 4, and 24 hours after TE followed by silver-infarct staining (SIS) as described below. Animal preparation For induction of anesthesia, the animals inhaled a gas mixture of halothane (Forene; Abott, Wiesbaden, Germany), nitrous oxyide (70%) and oxygen (30%) via a precalibrated vaporizer (Fortec; Cyprane Keighley, United Kingdom). The right femoral artery and vein were cannulated using polyethylene catheters (PE-50; Labokron, Sinsheim, Germany). Body temperature was kept constant at 37°C with a temperature-controlled heating pad (Föhr Medical Instruments, Germany) during surgery. The correlation between body temperature, pericranial and intracranial temperature has been shown before [17]. TE was induced as previously described [18]. Briefly, the right common carotid (CCA), internal carotid (ICA), and external carotid artery (ECA) were exposed and further dissection identified the origin of the pterygopalatine artery (PPA). The ECA and the PPA were permanently ligated while the CCA was only temporarily clipped for embolization. A PE 50 catheter was inserted into the ECA proximal to its ligation and 12 red blood clots (each 0.35 mm in diameter and 3 mm in length) were injected at the origin of the right middle cerebral artery (MCA). The whole surgical procedure lasted 30 to 40 min.

d while the CCA was only temporarily clipped for embolization. A PE 50 catheter was inserted into the ECA proximal to its ligation and 12 red blood clots (each 0.35 mm in diameter and 3 mm in length) were injected at the origin of the right middle cerebral artery (MCA). The whole surgical procedure lasted 30 to 40 min. Rt-PA and G-CSF treatment Rt-PA (Alteplase, Boehringer Ingelheim, Ingelheim am Rhein, Germany) was infused intravenously at a dose of 10 mg/kg body weight (b.w.). Ten percent were given as a bolus at the beginning of thrombolysis followed by continuous infusion over a 30-minute period with a Harvard pump (Harvard Apparatus). The dose was in accordance with experimental studies [18,19]. Rt-PA dissolved with 1 ml of saline 0.9%, while pure saline was administered to the animals in the control group a. Recombinant G-CSF (Neupogen, Amgen, Thousand Oaks) in a dose of 60 μg/kg body weight was dissolved in 1 ml of 0.9% saline and administered over a period of 30 min 60 min after TE. The dose was in accordance with a previous study [1].

saline 0.9%, while pure saline was administered to the animals in the control group a. Recombinant G-CSF (Neupogen, Amgen, Thousand Oaks) in a dose of 60 μg/kg body weight was dissolved in 1 ml of 0.9% saline and administered over a period of 30 min 60 min after TE. The dose was in accordance with a previous study [1]. MRI protocol The animals were examined in a 2.35-T scanner (Biospec 24/40, BRUKER Medizintechnik, Ettlingen, Germany). An actively shielded gradient coil with an inner diameter of 120 cm was used. This coil was driven by the standard 150 V/100A gradient power supply. In this configuration, 180 mT/m could be reached in 180 ms. A home-built birdcage resonator with an inner diameter of 40 mm as RF coil was used. MRI examination started within 30 min after TE and was repeated at 2.5 h, 4 h, and 24 h. The protocol included T2-WI, DWI, T2*, and PWI as described previously [18,20,21]. For PWI, a bolus of 0.5 mmol/kg bw gadodiamide (Omniscan®, Amersham, Braunschweig, Germany) was injected at the scan at 30 min, 2.5 hours, and 4 hours.

examination started within 30 min after TE and was repeated at 2.5 h, 4 h, and 24 h. The protocol included T2-WI, DWI, T2*, and PWI as described previously [18,20,21]. For PWI, a bolus of 0.5 mmol/kg bw gadodiamide (Omniscan®, Amersham, Braunschweig, Germany) was injected at the scan at 30 min, 2.5 hours, and 4 hours. An investigator blinded for the treatment groups measured the lesion volumes in T2-weighted and DWI-weighted MR images by tracing the area of hyperintense regions. The infarct volume was calculated similar to the method described for SIS staining. The apparent rrCBV was calculated from the PWI data, as previously described [18,20,21]. PWI data was assessed in a region of interest (ROI) with an area of 3 × 3 pixels at the level of the lateral basal ganglia, which represents a typical ischemic region after occlusion of the MCA. Values were then calculated in percent of the healthy contralateral hemisphere. Measurement of infarct volume by the silver infarct staining (SIS) method The SIS method was used to measure infarct size [18,22]. Following the last MRI investigation after 24 hours, the animals were scarified and brain slices were stained according to the SIS method. A modified version of the semi automated method [23] was used to measure the cerebral infarct volume and calculated as follows:

SIS method was used to measure infarct size [18,22]. Following the last MRI investigation after 24 hours, the animals were scarified and brain slices were stained according to the SIS method. A modified version of the semi automated method [23] was used to measure the cerebral infarct volume and calculated as follows: Functional neurological outcome All surviving animal were tested for neurological outcome using the neuroscore according to Menzies [24]: 0 = no apparent deficit, 1 = contralateral forelimb flexion; 2 = decreased grip of contralateral forelimb grip while tail pulled; 3 = spontaneous movement in all directions, contralateral circling only if pulled by tail; 4 = spontaneous contralateral circling. The testing was performed by a co-worker who was blinded for the earlier treatment regimen. Statistical analysis Values of the result section and figures are presented as mean ± S.D. However, functional outcome is given as median and range. After acquiring all the data, the randomization code was broken. ANOVA and subsequent post hoc Fisher protected least significant difference tests were used. A value of p < 0.05 was considered statistically significant. Animals that died before the endpoint of 24 hours were not excluded from MRI, but SIS analysis. The nonparametric Kruskall-Wallis test evaluated the neuroscore with subsequent group comparisons by Mann-Whitney U test. The Chi square test with Yates correction for small numbers was used to test for differences mortality rate.

of 41%. Therefore, the survival rate was significantly lower compared to all other experimental groups (p < 0.05). Animals that died prematurely were assessed for the cause of death and showed all large infarcts in the brain territory supplied by the MCA. Therefore, large infarct was probably responsible for the death. Functional outcome Functional outcome was as follows: control group showed a median of 2 (range 2-3), the early G-CSF group had a median of 2 (range 1-2) and the early G-CSF group a median of 2 (range 1-2). Animals of the early combination group showed a non significant trend towards a better outcome compared to all other groups, since the median Menzies score was 1 (range 1-2). Animals of the group which received delayed rt-PA treatment had a median score of 3 (range 2-3). Pretreatment with G-CSF lead to a median Menzies score of 2 (1-3). Infarct size calculated from SIS after 24 hours The extent of cerebral infarction was 38 ± 4% of the right, ischemic hemisphere in the remaining animals of the delayed rt-PA group, 22 ± 3% in the control group, 17 ± 3% in the early G-CSF group, 14 ± 3% in com group, and 11 ± 3% in the early thrombolysis group. Therefore, the infarct size in the delayed rt-PA group was larger compared to all others (p < 0.05). Moreover, the infarct size in the control group exceeded the early G-CSF, early rt-PA group, and com group (p < 0.05). Animals of the deco group had larger infarcts compared to early rt-PA, (p < 0.05). There was no difference between the control group and the deco group (Figure 1 and 2).

pared to all others (p < 0.05). Moreover, the infarct size in the control group exceeded the early G-CSF, early rt-PA group, and com group (p < 0.05). Animals of the deco group had larger infarcts compared to early rt-PA, (p < 0.05). There was no difference between the control group and the deco group (Figure 1 and 2). Figure 1 Infarct volume in SIS. Cerebral infarct is shown as calculated from SIS after 24 hours. Infarct size is larger in the control group (control) compared to early G-CSF (G-CSF), early rt-PA (rt-PA), and early combination (com) as indicated by the asterisks (p < 0.05). Figure 2 Infarct volume in SIS after delayed rt-PA treatment. Cerebral infarct is shown as calculated from SIS after 24 hours. Infarct size is larger in the delayed rt-PA group (delayed rt-PA) compared to the control group (control), and delayed combination (deco) as indicated by the asterisks (p < 0.05).

Figure 1 Infarct volume in SIS. Cerebral infarct is shown as calculated from SIS after 24 hours. Infarct size is larger in the control group (control) compared to early G-CSF (G-CSF), early rt-PA (rt-PA), and early combination (com) as indicated by the asterisks (p < 0.05). Figure 2 Infarct volume in SIS after delayed rt-PA treatment. Cerebral infarct is shown as calculated from SIS after 24 hours. Infarct size is larger in the delayed rt-PA group (delayed rt-PA) compared to the control group (control), and delayed combination (deco) as indicated by the asterisks (p < 0.05). Perfusion weighted imaging Analysis of the relative regional cerebral blood volume (rrCBV) at the level of the basal ganglia showed a decrease to below 50% of the corresponding nonischemic ROIs 30 min after thromboembolic occlusion (data not shown). This effect was not different between the groups (p = ns). The rrCBV slightly was significantly larger in the rt-PA group and com group compared to all other groups, (p < 0.001). There were no significant differences in the groups not treated by rt-PA (p = ns). 4 hours after stroke onset, rrCBV was larger in the rt-PA group, delayed rt-PA group and the com group compared to the control group and the early G-CSF group, (p < 0.005). There was a significant difference between the rt-PA group (94 ± 11%), delayed rt-PA group (91 ± 7%) compared to the deco group (75 ± 4%; p < 0.05). Data is shown in Table 1. Table 1 Data of serial MRI. DWI in mm3 2.5 hrs 4 hrs 24 hrs control 43 ± 5 86 ± 14 102 ± 11 early G-CSF 31 ± 5 65 ± 17 71 ± 8 rt-PA 26 ± 7 57 ± 17 68 ± 9 com 21 ± 8 49 ± 14 65 ± 11

Perfusion weighted imaging Analysis of the relative regional cerebral blood volume (rrCBV) at the level of the basal ganglia showed a decrease to below 50% of the corresponding nonischemic ROIs 30 min after thromboembolic occlusion (data not shown). This effect was not different between the groups (p = ns). The rrCBV slightly was significantly larger in the rt-PA group and com group compared to all other groups, (p < 0.001). There were no significant differences in the groups not treated by rt-PA (p = ns). 4 hours after stroke onset, rrCBV was larger in the rt-PA group, delayed rt-PA group and the com group compared to the control group and the early G-CSF group, (p < 0.005). There was a significant difference between the rt-PA group (94 ± 11%), delayed rt-PA group (91 ± 7%) compared to the deco group (75 ± 4%; p < 0.05). Data is shown in Table 1. Table 1 Data of serial MRI. DWI in mm3 2.5 hrs 4 hrs 24 hrs control 43 ± 5 86 ± 14 102 ± 11 early G-CSF 31 ± 5 65 ± 17 71 ± 8 rt-PA 26 ± 7 57 ± 17 68 ± 9 com 21 ± 8 49 ± 14 65 ± 11 delayed rt-PA 43 ± 4 136 ± 22 180 ± 13 deco 33 ± 5 80 ± 14 87 ± 12 T2 in mm3 4 hrs 24 hrs control 79 ± 12 113 ± 7 early G-CSF 65 ± 18 85 ± 8 rt-PA 68 ± 13 69 ± 8 com 54 ± 16 70 ± 15 delayed rt-PA 129 ± 14 193 ± 15 deco 68 ± 13 93 ± 13 PWI in % 2.5 hrs 4 hrs control 28 ± 3 58 ± 7 early G-CSF 34 ± 5 53 ± 8 rt-PA 78 ± 4 94 ± 11 com 75 ± 6 91 ± 7 delayed rt-PA 34 ± 3 90 ± 5 deco 38 ± 7 75 ± 4 These are given as means ± SD for different time points after stroke onset. Diffusion weighted imaging There were no significant differences between the lesion size after 0.5 hours as calculated from DWI (p = ns; data not shown).

early G-CSF 34 ± 5 53 ± 8 rt-PA 78 ± 4 94 ± 11 com 75 ± 6 91 ± 7 delayed rt-PA 34 ± 3 90 ± 5 deco 38 ± 7 75 ± 4 These are given as means ± SD for different time points after stroke onset. Diffusion weighted imaging There were no significant differences between the lesion size after 0.5 hours as calculated from DWI (p = ns; data not shown). DWI lesions after 2.5 hours accounted to 43 ± 5 mm3 in the control group and were significantly larger than those observed for the rt-PA group by 26 ± 7 mm3, and early com group by 21 ± 8 mm3 (p < 0.05), (Data shown in Figure 3). There was no significant difference compared to the early G-CSF group by 31 ± 5 mm3, the deco group by 33 ± 5 mm3, and the delayed rt-PA group with 42.7 ± 4 mm3. Animals treated with early G-CSF tended towards smaller lesions, since the lesion volume was 26 ± 6 mm3 in the early G-CSF group and 21 ± 8 mm3 in the com group. Early thormbolysis lead to a lesion volume of 25.9 ± 9 mm3 (p < 0.05). 4 hours after stroke onset, lesion volume was larger in the delayed rt-PA group (135.8 ± 22 mm3) than in the deco group (80.2 ± 14 mm3; p < 0.05), the early G-CSF group (65.1 ± 17 mm3; p < 0.005), and early rt-PA (57 ± 17 mm3; p < 0.005). After 24 hours, the lesion volume of the delayed rt-PA group was 180 ± 13 mm3 and significantly larger than in the control group (102.4 ± 11 mm3; p > 0.001), the early G-CSF group (71.1 ± 8 mm3; p < 0.001), the deco group (87.3 ± 12 mm3; p > 0.001), and the early rt-PA group (68.3 ± 9 mm3; p > 0.001). Moreover, lesion volume in the early G-CSF group and the early rt-PA group was smaller than in the control group (p < 0.05).

in the control group (102.4 ± 11 mm3; p > 0.001), the early G-CSF group (71.1 ± 8 mm3; p < 0.001), the deco group (87.3 ± 12 mm3; p > 0.001), and the early rt-PA group (68.3 ± 9 mm3; p > 0.001). Moreover, lesion volume in the early G-CSF group and the early rt-PA group was smaller than in the control group (p < 0.05). Figure 3 Lesion volume in MRI. Lesion volume is shown as calculated from DWI and expressed as means ± SD. The lesion volume in the control group was larger compared to early rt-PA (rt-PA) and early combination group (com) after 2.5 hours, and larger than the early rt-PA, early G-CSF (G-CSF) and early combination group after 4 and 24 hours as shown by the single asterisks. Delayed rt-PA treatment (delayed rt-PA) resulted in larger lesion volume than the early rt-PA, early G-CSF and early combination group after 2.5 hours, and additionally larger than the control group after 4 hours. It was larger than all others after 24 hours as shown by the string sign. The late combination group (deco) showed larger infarcts than the early combination after 2.5 hours. After 4 hours, lesion volume was larger than all others with exception of delayed rt-PA. After 24 hours, there was a difference compared to the early rt-PA and early combination as indicated by the double asterisks. Levels of significance are given in table 2.

larger infarcts than the early combination after 2.5 hours. After 4 hours, lesion volume was larger than all others with exception of delayed rt-PA. After 24 hours, there was a difference compared to the early rt-PA and early combination as indicated by the double asterisks. Levels of significance are given in table 2. Values of DWI after 24 hours show significant differences between the treatment groups (Table 1 and 2). The infarct volume was larger in the control group a vs. early G-CSF group (p = 0.019), to early rt-PA group (p = 0.006), and the com group (p = 0.013). No treatment resulted in smaller infarct volume than G-CSF plus delayed rt-PA administration in group deco (p = 0.002). There was a non-significant trend towards smaller infarct volume in group deco compared to the control group (p = 0.25; n.s.). Comparing the early treatment group G-CSF and rt-PA, there was no difference for the infarct volume. Combining rt-Pa and G-CSF (group com) in the early phase lead to smaller infarct than for rt-PA treatment alone (p = 0.001). Early rt-PA treatment (group rt-PA) and combination (group com) lead to smaller infarcts than in group delayed rt-PA (p = 0.001). Pretreatment by G-CSF in combination with delayed rt-PA in group deco reduced the infarct volume compared to the delayed rt-PA group (p = 0.001). Table 2 Comparing diffusion-weighted imaging after 2.5 hours control early G-CSF rt-PA com delayed rt-PA deco control n.s. < 0.05 < 0.05 n.s. n.s. early G-CSF n.s. n.s. n.s. < 0.05 n.s. rt-PA < 0.05 n.s. n.s. < 0.05 n.s. com n.s. n.s. n.s. < 0.05 n.s.

Values of DWI after 24 hours show significant differences between the treatment groups (Table 1 and 2). The infarct volume was larger in the control group a vs. early G-CSF group (p = 0.019), to early rt-PA group (p = 0.006), and the com group (p = 0.013). No treatment resulted in smaller infarct volume than G-CSF plus delayed rt-PA administration in group deco (p = 0.002). There was a non-significant trend towards smaller infarct volume in group deco compared to the control group (p = 0.25; n.s.). Comparing the early treatment group G-CSF and rt-PA, there was no difference for the infarct volume. Combining rt-Pa and G-CSF (group com) in the early phase lead to smaller infarct than for rt-PA treatment alone (p = 0.001). Early rt-PA treatment (group rt-PA) and combination (group com) lead to smaller infarcts than in group delayed rt-PA (p = 0.001). Pretreatment by G-CSF in combination with delayed rt-PA in group deco reduced the infarct volume compared to the delayed rt-PA group (p = 0.001). Table 2 Comparing diffusion-weighted imaging after 2.5 hours control early G-CSF rt-PA com delayed rt-PA deco control n.s. < 0.05 < 0.05 n.s. n.s. early G-CSF n.s. n.s. n.s. < 0.05 n.s. rt-PA < 0.05 n.s. n.s. < 0.05 n.s. com n.s. n.s. n.s. < 0.05 n.s. delayed rt-PA n.s. < 0.05 < 0.05 < 0.05 n.s. deco n.s. n.s. n.s. < 0.05 n.s. after 4 hours control early G-CSF rt-PA com delayed rt-PA deco control n.s. p < 0.05 p < 0.01 < 0.05 n.s. early G-CSF n.s. n.s. n.s. < 0.005 n.s. rt-PA p < 0.05 n.s. n.s. < 0.005 n.s. com p < 0.01 n.s. n.s. < 0.005 n.s. delayed rt-PA < 0.05 < 0.005 < 0.005 < 0.005 < 0.05 deco n.s. n.s. n.s. n.s. < 0.05

delayed rt-PA n.s. < 0.05 < 0.05 < 0.05 n.s. deco n.s. n.s. n.s. < 0.05 n.s. after 4 hours control early G-CSF rt-PA com delayed rt-PA deco control n.s. p < 0.05 p < 0.01 < 0.05 n.s. early G-CSF n.s. n.s. n.s. < 0.005 n.s. rt-PA p < 0.05 n.s. n.s. < 0.005 n.s. com p < 0.01 n.s. n.s. < 0.005 n.s. delayed rt-PA < 0.05 < 0.005 < 0.005 < 0.005 < 0.05 deco n.s. n.s. n.s. n.s. < 0.05 after 24 hours control early G-CSF rt-PA com delayed rt-PA deco control < 0.05 < 0.05 < 0.05 < 0.001 n.s. early G-CSF < 0.05 n.s. n.s. < 0.001 n.s. rt-PA < 0.05 n.s. n.s. < 0.001 < 0.05 com < 0.05 n.s. n.s. < 0.001 < 0.05 delayed rt-PA < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 deco n.s. n.s. < 0.05 < 0.05 < 0.001 Levels of significant difference are given here as calculated from DWI-sequences.

control early G-CSF rt-PA com delayed rt-PA deco control < 0.05 < 0.05 < 0.05 < 0.001 n.s. early G-CSF < 0.05 n.s. n.s. < 0.001 n.s. rt-PA < 0.05 n.s. n.s. < 0.001 < 0.05 com < 0.05 n.s. n.s. < 0.001 < 0.05 delayed rt-PA < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 deco n.s. n.s. < 0.05 < 0.05 < 0.001 Levels of significant difference are given here as calculated from DWI-sequences. T2-WI and T2*-WI There were no significant differences among all groups at the first scan after 30 min (data not shown) and after 2.5 hours (p = ns). After 4 hours, lesion size in the delayed rt-PA group was 129.4 ± 14.4 mm3 and therefore larger than in the early G-CSF group (65.3 ± 18 mm3), the com group (68.1 ± 13 mm3), and the rt-PA group (58.1 ± 12 mm3); (p < 0.005). After 24 hours, infarct size was larger in the control (113 ± 7 mm3) compared to the early G-CSF group (85.5 ± 8 mm3; p < 0.05), the com group (70.3 ± 15 mm3; p < 0.01), and the early rt-PA group (59.1 ± 8 mm3); (p < 0.005). However, infarct size in the control group was smaller compared to the delayed rt-AP group (193.3 ± 5 mm3; p < 0.001). Moreover, the lesion size in the delayed rt-PA group was larger than in the com group, the early G-CSF group, and early rt-PA group (p < 0.001). T2*-WI showed 4 animals (one in each group excluding early thrombolysis) with an intra-cerebral hemorrhage.

ler compared to the delayed rt-AP group (193.3 ± 5 mm3; p < 0.001). Moreover, the lesion size in the delayed rt-PA group was larger than in the com group, the early G-CSF group, and early rt-PA group (p < 0.001). T2*-WI showed 4 animals (one in each group excluding early thrombolysis) with an intra-cerebral hemorrhage. 4. Discussion This study shows for the first time that G-CSF reduces infarct volume in a model of thromboembolic stroke. Moreover, it prevents some deleterious effects of delayed rt-PA treatment in terms of infarct growth and mortality during an observation time of 24 hours after stroke onset. These results are an important step towards further clinical investigations on G-CSF and acute stroke, since the EPO trial failed because of deleterious combination therapy of EPO and rt-PA [16].

delayed rt-PA treatment in terms of infarct growth and mortality during an observation time of 24 hours after stroke onset. These results are an important step towards further clinical investigations on G-CSF and acute stroke, since the EPO trial failed because of deleterious combination therapy of EPO and rt-PA [16]. Neuroprotective effects of G-CSF Recent experimental studies showed that G-CSF is beneficial after cerebral ischemia and brain injury [1-8]. While neuroprotective effects are described in the early stage of brain injury, G-CSF also stimulates neuronal progenitor cells providing a link to functional recovery [6,9]. G-CSF reduces infarct volume after transient suture occlusion of the MCA and protects neurons against glutamate-induced excitotoxicity in cell culture [1,10]. Further neuroprotective mechanisms include an increased STAT3 regulation in the penumbra of G-CSF-treated rats. Effects of G-CSF are probably mediated by a special neuronal G-CSF receptor [1,7], since G-CSF passes even the intact the blood-brain-barrier [7] and therefore reaches injured brain regions. Moreover, G-CSF seems to have additional regenerative effects as bone marrow cells are activated. Neuronal plasticity and vascularisation were proven in experimental studies of cerebral ischemia [6,7].

tor [1,7], since G-CSF passes even the intact the blood-brain-barrier [7] and therefore reaches injured brain regions. Moreover, G-CSF seems to have additional regenerative effects as bone marrow cells are activated. Neuronal plasticity and vascularisation were proven in experimental studies of cerebral ischemia [6,7]. G-CSF compared to early thrombolysis So far, G-CSF was not tested in models of thromboembolic stroke. However, this step is essential when transferring neuroprotective agents to stroke patients. If neuroprotectants are successful at all, the chance is the highest in early stages of stroke. In this study, MRI-data as well as SIS showed that G-CSF was as effective as early thrombolysis in terms of reduction of infarct volume. There was no difference in mortality as well. G-CSF did not influence rrCBV compared to the control group. As expected, early rt-PA treatment resulted in almost complete normalisation of rrCBV. It can be suggested that the mentioned multiple neuroprotective effects contribute to reduction of infarct for this subgroup. These results are in accordance to former studies in suture occlusion model [1,6].

ed to the control group. As expected, early rt-PA treatment resulted in almost complete normalisation of rrCBV. It can be suggested that the mentioned multiple neuroprotective effects contribute to reduction of infarct for this subgroup. These results are in accordance to former studies in suture occlusion model [1,6]. G-CSF prior to delayed thrombolysis G-CSF reduced infarct volume and mortality when given prior to delayed rt-PA treatment. Moreover, there was no significant difference to early rt-PA and G-CSF treatment alone. In accordance to data of the suture occlusion model, delayed restoration of CBF leads to larger infarct than early restoration. Aronowski et al. observed larger infarct volumes and edema as compared to permanent ischemia, when the MCA occluding suture was removed 120 to 300 min after induction of ischemia [25]. Clinical data shows that delayed thrombolysis beyond the three hours window in already demarked infarct increases the risk of side effects such as hemorrhage or enlargement of infarction [14]. Probably these side effects are caused by reperfusion-associated injury and neurotoxic properties of rt-PA [26,27]. In contrast to the patient, infarct mature much faster in rats and therefore a time window of three hours can be considered as late thrombolysis. In accordance, early rt-PA treatment lead to smaller infarcts compared to the control group, while delayed administration of rt-PA increased infarct volume and mortality in this study. It can be suggested that G-CSF helps to prolong the time window for thrombolysis. This is of importance, since the transport of patients to the clinic is often long and excludes them from thrombolytic treatment. Pre-treatment In the ambulance car or if the treatment with rt-PA is delayed in the hospital might be helpful for these subgroup of stroke patients.

prolong the time window for thrombolysis. This is of importance, since the transport of patients to the clinic is often long and excludes them from thrombolytic treatment. Pre-treatment In the ambulance car or if the treatment with rt-PA is delayed in the hospital might be helpful for these subgroup of stroke patients. However, there are several limitations of the study. Certainly, observation periods of more than 24 hours are necessary to test whether effects of neuroprotectants are transient or permanent. Moreover, further investigations should address whether G-CSF interacts with rt-PA and CBF. Autoradiographic techniques may answer these questions, but were not in the focus of the present study. Animals were exposed to anesthesia for several hours. While this could interfere with mortality overall, differences between the groups cannot explained with it. Although we did not investigate further pathways of G-CSF and combination with rt-PA, this study is essential when G-CSF will be investigated in patients treated by rt-PA.

exposed to anesthesia for several hours. While this could interfere with mortality overall, differences between the groups cannot explained with it. Although we did not investigate further pathways of G-CSF and combination with rt-PA, this study is essential when G-CSF will be investigated in patients treated by rt-PA. In conclusion, the results of the present study are encouraging on the path of new therapies for ischemic stroke. G-CSF represents an interesting and promising candidate for stroke therapy because of its neuroprotective properties, potential induction of stem cells and good clinical tolerance in hematological patients. Further experimental studies have to investigate combination therapy of G-CSF and rt-PA over longer time periods, since combining the so far best medical therapy rt-PA with new drugs represents a logical and potentially successful way for stroke treatment. Competing interests The authors RK and SS are involved into the AXIS-trial investigating safety and feasibility of G-CSF after acute stroke. Moreover, they own a patent on the use of growth-factors such as G-CSF for the treatment of stroke. Authors' contributions RK and SS designed the study, did the statistics and prepared the manuscript. NH and CU performed the experiments. NH investigated neurological examination of the animals. All authors read and approved the final manuscript.

Introduction Adiponectin is an abundantly expressed adipokine that is released into the circulation and self-associates to form homotrimers. Adiponectin trimers further associate to form hexamers, high molecular weight (HMW) oligomers and a globular fraction, generated by proteolytic cleavage of full-length adiponectin monomers [1,2]. Adiponectin receptor 1 (ADR1) and adiponectin receptor 2 (ADR2) are the major receptors for adiponectin. Both ADRs can be activated by all forms of adiponectin found in the circulation. However, ADR1 has a higher affinity for globular adiponectin (gAd) over the full-length forms, whereas ADR2 has a similar affinity for both isoforms [3]. In addition, HMW oligomers are reported to be a specific ligand for T-cadherin [4]. ADRs were shown to exert actions in the peripheral tissues by activating the AMP-activated protein kinase α (AMPKα) [5], p38 mitogen-activated protein kinase (p38-MAPK) [6] and nuclear factor-kappa B (NFκB) [reviewed in reference [7]]. In the brain, ADRs 1 and 2 are expressed in the arcuate and the paraventricular nuclei of the hypothalamus, where they regulate feeding behaviours [8,9]. However, the functions of adiponectin in other regions of the central nervous system (CNS) are still poorly understood.

or-kappa B (NFκB) [reviewed in reference [7]]. In the brain, ADRs 1 and 2 are expressed in the arcuate and the paraventricular nuclei of the hypothalamus, where they regulate feeding behaviours [8,9]. However, the functions of adiponectin in other regions of the central nervous system (CNS) are still poorly understood. The cerebral ischemia that occurs in brain cells affected by a stroke triggers a complex array of molecular and cellular alterations including activation of signaling pathways that may either contribute to neuronal damage or protect neurons. Among the pathways known to be activated in neurons in response to ischemia, are those involving AMPKα and P38-MAPK [10,11]. It was recently reported that levels of circulating adiponectin increase after an ischemic stroke [12]. However, it is not known whether ADRs are activated in neurons in response to ischemic stroke, nor have the consequences of ADR signaling on the clinical outcome of a cerebral ischemic event been established. In the present study we show that both ADR1 and ADR2 are expressed in cerebral cortical neurons, and that activation of ADR1 leads to neuronal cell death under ischemic conditions.

ponse to ischemic stroke, nor have the consequences of ADR signaling on the clinical outcome of a cerebral ischemic event been established. In the present study we show that both ADR1 and ADR2 are expressed in cerebral cortical neurons, and that activation of ADR1 leads to neuronal cell death under ischemic conditions. Materials and methods Animals and Stroke Model Three-month-old C57BL/6 male mice were used for all in vivo experiments. All animal experimental procedures performed were reviewed and approved by the University of Queensland Animal Care and Use Committee. Transient focal cerebral ischemia was induced by middle cerebral artery occlusion (MCAO) using the previously described intraluminal filament method [13]. Briefly, mice were anesthetized with isoflurane, a midline incision was made in the neck, and the left external carotid and pterygopalatine arteries were isolated and ligated with 5-0 silk thread. The internal carotid artery (ICA) was occluded at the peripheral site of the bifurcation of the ICA and the pterygopalatine artery with a small clip, and the common carotid artery (CCA) was ligated with 5-0 silk thread. The external carotid artery (ECA) was cut, and a 6-0 nylon monofilament with a tip that was blunted (0.2-0.22 mm) with a coagulator was inserted into the ECA. After the clip at the ICA was removed, the nylon thread was advanced into the middle cerebral artery (MCA) until light resistance was felt. The nylon thread and the CCA ligature were removed after 1 h of occlusion to initiate reperfusion. In the sham operated group, these arteries were surgically exposed but not disturbed. At different time points during the reperfusion period, mice were euthanized and brains were immediately removed and processed for immunoblot and immunohistochemical analysis.

re removed after 1 h of occlusion to initiate reperfusion. In the sham operated group, these arteries were surgically exposed but not disturbed. At different time points during the reperfusion period, mice were euthanized and brains were immediately removed and processed for immunoblot and immunohistochemical analysis. Patient tissue collection The case is a 39-year-old man who had an acute brainstem stroke. He died on the ninth day after the incident; due to massive infarcts and obstructive hydrocephalus. Acute basilar artery occlusion related to atherosclerosis and associated thrombi were suggested as the possible causes of death on autopsy [14]. Neuronal cultures Cortical tissues dissected from C57BL/6 mouse embryos at the E15 developmental stage were incubated for 15 min in a solution of 2 mg/ml trypsin in-Ca2+/Mg2+-free Hank's balanced salt solution (HBSS) (Invitrogen, USA) buffered with 10 mmol/L HEPES. Tissues were then dissociated and cells were plated in 60 or 100-mm diameter plastic dishes or 24-well plates and maintained at 37°C in Neurobasal medium containing B-27 supplements (Invitrogen, Carlsbad, CA, USA), 2 mmol/L L-glutamine, 0.001% gentamycin sulfate and 1 mmol/L HEPES (pH 7.2). Experiments were performed in 7 to 9-day-old cultures. Approximately 95% of the cells in such cultures were neurons and the remaining cells were astrocytes.

37°C in Neurobasal medium containing B-27 supplements (Invitrogen, Carlsbad, CA, USA), 2 mmol/L L-glutamine, 0.001% gentamycin sulfate and 1 mmol/L HEPES (pH 7.2). Experiments were performed in 7 to 9-day-old cultures. Approximately 95% of the cells in such cultures were neurons and the remaining cells were astrocytes. RT-PCR analysis PCR primers were designed using Primer3 software and synthesized by Integrated DNA Technologies, Inc. (Coralville, IA, USA). Total RNA from cultured neurons was extracted using Trizol reagent (Sigma, St-Louis, MO, USA). For single-cell RT-PCR analysis, individual neurons were visualized by using a phase-contrast microscope and collected into a micropipette. PCR products were electrophoresed on a 2% agarose gel, and were visualized by ethidium bromide staining. Primer sequences used in this study: ADR1 (forward primer) 5' TCC TGA CTG GCT GAA AGA CAA CGA 3', (reverse primer) 5' ACA GTG TGG AAG AGC CAG GAG AAA 3', ADR2 (forward primer) 5'-TGT GCT ACC GGA TTG GCT TAA GGA-3', (reverse primer) 5'-TAC ACC GTG TGG AAG AGC CAT GAA-3', Actin (forward primer) 5'GGC TGT GTC CCAT GTA T 3', (reverse primer) 5'CCG CTC ATT GCC GAT AGT G 3'.

primer) 5' TCC TGA CTG GCT GAA AGA CAA CGA 3', (reverse primer) 5' ACA GTG TGG AAG AGC CAG GAG AAA 3', ADR2 (forward primer) 5'-TGT GCT ACC GGA TTG GCT TAA GGA-3', (reverse primer) 5'-TAC ACC GTG TGG AAG AGC CAT GAA-3', Actin (forward primer) 5'GGC TGT GTC CCAT GTA T 3', (reverse primer) 5'CCG CTC ATT GCC GAT AGT G 3'. Immunoblot analysis Lysates of cultured cells were obtained by washing the cells in ice-cold PBS and resuspending them in cell lysis buffer. Proteins were extracted from ipsilateral mouse brain tissue specimens and 40 μg of protein was separated by SDS-PAGE (8-12%) and then transferred to a nitrocellulose membrane. The membrane was blocked in 5% non-fat milk for 1 h at room temperature, followed by an overnight incubation at 4°C with primary antibodies against: Actin (Sigma); ADR1 (Santacruz, USA; Alexis, USA), ADR2 (Alexis, USA), APPL-1, p-AMPK and Cleaved Caspase-3 (Cell Signaling). Membranes were then washed and incubated with secondary antibodies for 1 h at room temperature °C. Protein bands were visualized using a chemiluminescence detection kit (Amersham Biosciences, Piscataway, NJ, USA).

(Santacruz, USA; Alexis, USA), ADR2 (Alexis, USA), APPL-1, p-AMPK and Cleaved Caspase-3 (Cell Signaling). Membranes were then washed and incubated with secondary antibodies for 1 h at room temperature °C. Protein bands were visualized using a chemiluminescence detection kit (Amersham Biosciences, Piscataway, NJ, USA). Immunocytochemistry and Immunohistochemistry Neurons grown on 24-chamber microscope slides were fixed in 4% paraformaldehyde and incubated at 4°C with primary ADR1 (Santacruz, USA; Alexis, USA), ADR2 (Alexis, USA) antibodies overnight, followed by a 2 h incubation with FITC conjugated secondary antibody at room temperature. Frozen brain sections were incubated with primary antibodies against ADR1, ADR2 and NeuN (Millipore, Billerica, MA). Images of cells were acquired using a Zeiss Axiophot microscope (Oberkochen, Germany). Formalin-fixed, paraffin embedded human brain sections were incubated with primary antibodies against Adiponectin (Abcam, USA), and biotinylated secondary antibody and avidin-biotin peroxidase complex (Vector Elite Kit; Vector, Burlingame, CA, USA). The peroxidase reaction was developed using a peroxidase substrate kit (diaminobenzidine DAB, SK-4100; Vector).

rain sections were incubated with primary antibodies against Adiponectin (Abcam, USA), and biotinylated secondary antibody and avidin-biotin peroxidase complex (Vector Elite Kit; Vector, Burlingame, CA, USA). The peroxidase reaction was developed using a peroxidase substrate kit (diaminobenzidine DAB, SK-4100; Vector). Glucose and oxygen deprivation and Cell viability experiments For glucose deprivation (GD), cultured neurons were incubated in glucose-free Locke's medium containing (in mmol/L) 154 NaCl, 5.6 KCl, 2.3 CaCl2, 1 MgCl2, 3.6 NaHCO3, 5 HEPES, pH 7.2, supplemented with gentamycin (5 mg/L) for 6, 12 or 24 h. For combined oxygen and glucose deprivation (OGD), neurons were incubated in glucose-free Locke's medium in an oxygen-free chamber with 95% N2/5% CO2 atmosphere for either 3, 6, 9, 12 or 24 h. Effects of globular adiponectin (gAd) and trimer adiponectin (tAd) (CYT-432, CYT-247 Prospec Bio, Israel) against GD or OGD-induced neuronal cell death were determined by trypan blue dye exclusion assay. Statistical Analyses Statistical comparisons were made by using ANOVA, and Newman-Keuls post hoc tests for pairwise comparisons.

Glucose and oxygen deprivation and Cell viability experiments For glucose deprivation (GD), cultured neurons were incubated in glucose-free Locke's medium containing (in mmol/L) 154 NaCl, 5.6 KCl, 2.3 CaCl2, 1 MgCl2, 3.6 NaHCO3, 5 HEPES, pH 7.2, supplemented with gentamycin (5 mg/L) for 6, 12 or 24 h. For combined oxygen and glucose deprivation (OGD), neurons were incubated in glucose-free Locke's medium in an oxygen-free chamber with 95% N2/5% CO2 atmosphere for either 3, 6, 9, 12 or 24 h. Effects of globular adiponectin (gAd) and trimer adiponectin (tAd) (CYT-432, CYT-247 Prospec Bio, Israel) against GD or OGD-induced neuronal cell death were determined by trypan blue dye exclusion assay. Statistical Analyses Statistical comparisons were made by using ANOVA, and Newman-Keuls post hoc tests for pairwise comparisons. Results ADR1 and ADR2 are expressed in cortical neurons, and their levels increase in response to oxygen and glucose deprivation Using single-cell PCR analysis we found that cultured murine cortical neurons express mRNA for ADR1 and ADR2 (Figure 1A). In order to determine whether ADR signaling is involved in neuronal responses to ischemic conditions, we first evaluated ADR expression levels in cultured neurons subjected to OGD. OGD exposure resulted in an increase in the mRNA levels of ADR1, as well as a transient increase in the mRNA levels of ADR2 that peaked at 3 hours (Figure 1B). ADR1 protein levels were also increased during OGD while ADR2 showed a transient increase in protein expression that peaked at 12 hours after the onset of OGD (Figure 1C). Confocal microscopy of ADR1 immunoreactivity showed that ADR1 is present in the cell body, axon and dendrites, but is absent from the nucleus, of cultured cortical neurons (Figure 1D). Addition of globular adiponectin to cultured neurons increased the expression of ADR1 at 6 and 12 hours post-OGD exposure, but had no effect on ADR2 expression levels (Figure 1E).

R2 (D) ADR1 immunoreactivity (red) in cultured neurons; cells were counterstained with DAPI (blue) to label all nuclei. Arrow points to the axon of a neuron and arrowheads point to dendrites of the same neuron. (E) Cortical neurons subjected to OGD following globular adiponectin treatment show increased levels of ADR1. Cerebral ischemia induces a rapid increase in ADR immunoreactivity in the brain The extensive increase in expression of ADR1 and ADR2 following OGD in vitro encouraged us to examine whether similar effects occur in vivo following ischemic damage. Immunoblots of ischemic cortical tissues at different times following ischemic reperfusion demonstrated increased levels of ADR1 (Figure 2A). In contrast, the levels of ADR2 remained unchanged following MCAO and reperfusion (Figure 2A). APPL1 is an adapter protein involved in ADR1 and ADR2 signaling and enhances the binding affinities of the ADRs to adiponectin [15]. APPL1 levels were highly expressed in the cortex, and we observed an increase in its expression levels in response to MCAO and reperfusion (Figure 2A). In the ipsilateral cerebral cortex of sham-operated control mice, little or no immunoreactivity with ADR1 and ADR2 antibodies was observed. At 6 h after stroke, neurons in the ischemic cortex exhibited robust ADR1 immunoreactivity (Figure 2B). In order to see adiponectin accumulate in the human brain following ischemic stroke, we analysed brain tissue obtained from a stroke patient at National Taiwan University Hospital. We observed accumulation of adiponectin in vessel-like structres (Large arrows) as well as in parenchyma (small arrows) in human ischemic brainstem (2C).

e adiponectin accumulate in the human brain following ischemic stroke, we analysed brain tissue obtained from a stroke patient at National Taiwan University Hospital. We observed accumulation of adiponectin in vessel-like structres (Large arrows) as well as in parenchyma (small arrows) in human ischemic brainstem (2C). Figure 2 Cerebral ischemia increases ADRs immunoreactivity in the brain. (A) Immunoblot analysis of protein samples from the cerebral cortex of sham operated control mice and mice subjected to 1 h cerebral ischemia and 1-24 h reperfusion. Ischemia resulted in rapid increases in the levels of ADRs immunoreactivities in neurons in the penumbra area (P). (Scale bars: 50 μM). (B) Images of brain sections showing ADR1 immunoreactivities (green) and NeuN (neuronal marker) in mice subjected to cerebral ischemia (1 h) and reperfusion (24 h). (C) Adiponectin accumulates in vessels like structures (large yellow arrow) and in parenchyma (small yellow arrow) in the human ischemic brain. (D) Control brain tissue stained with secondary antibody shows no staining.

een) and NeuN (neuronal marker) in mice subjected to cerebral ischemia (1 h) and reperfusion (24 h). (C) Adiponectin accumulates in vessels like structures (large yellow arrow) and in parenchyma (small yellow arrow) in the human ischemic brain. (D) Control brain tissue stained with secondary antibody shows no staining. ADR activation enhances cortical neuronal death induced by oxygen and glucose deprivation Because the activation of AMPK, p38-MAPK and caspase-3 cleavage and their consequent mitochondrial alterations are implicated in ischemic neuronal death [10,11], and because ADR signaling activates AMPK and p38-MAPK, we next measured levels of cleaved caspase-3, phosphorylated-AMPK (p-AMPK) and p38-MAPK following GD and OGD in globular adiponectin-treated neurons compared with vehicle-treated neurons. Globular adiponectin treatment significantly increased p-AMPK, p38-MAPK and activated caspase-3 levels as compared to vehicle-treated neurons suggesting a pro-apoptotic role for ADRs in neurons under ischemia-like conditions (Figure 3A &3B respectively). Furthermore, we analysed GD- and OGD-induced cell death in adiponectin-treated neurons (10 μg/ml; Prospec, Israel) and compared it with vehicle-treated neurons. Our data showed that both globular and trimeric adiponectin significantly increased cell death in neurons subjected to GD and OGD (Figure 3C-F).

respectively). Furthermore, we analysed GD- and OGD-induced cell death in adiponectin-treated neurons (10 μg/ml; Prospec, Israel) and compared it with vehicle-treated neurons. Our data showed that both globular and trimeric adiponectin significantly increased cell death in neurons subjected to GD and OGD (Figure 3C-F). Figure 3 ADR activation mediates neuronal cell death following in vitro ischemia-like conditions. (A, B) Neuronal cultures were treated with 10 μg of the globular adiponectin (gAd) and then subjected to OGD for 12 h (A) or GD for 24 h (B). Proteins in cell lysates were then subjected to immunoblot analysis by using the indicated antibodies. The gAd treatment enhanced OGD or GD-induced increases in levels of p-AMPK, p-38 MAPK and activated caspase-3. (C-D) Globular adiponectin (gAd) and trimeric adiponectin (tAd) treatment exacerbates OGD induced death of cultured primary neurons. Neuronal cell death was quantified 12 h later. Values are mean ± s.e.m. (n = 6-10 cultures). ***P < 0.0001 compared to OGD or vehicle treated OGD value, *P < 0.05 compared to OGD or vehicle treated OGD value. (E) Globular adiponectin (gAd) treatment exacerbates GD induced death of cultured primary neurons. Neuronal cell death was quantified 24 h later. Values are mean ± s.e.m. (n = 6-10 cultures). ***P < 0.0001 compared to GD or vehicle treated GD value.

ted OGD value, *P < 0.05 compared to OGD or vehicle treated OGD value. (E) Globular adiponectin (gAd) treatment exacerbates GD induced death of cultured primary neurons. Neuronal cell death was quantified 24 h later. Values are mean ± s.e.m. (n = 6-10 cultures). ***P < 0.0001 compared to GD or vehicle treated GD value. Discussion In the brain, so far, adiponectin receptor expression has only been shown in the arcuate and the paraventricular nuclei of hypothalamus [8,9]. We have now identified the presence of both adiponectin receptors (ADRs 1 & 2) in mice cortical neurons. Confocal microscopy of ADR1 immunoreactivity shows that ADR1 is present in the cell body, axon and dendrites, but is absent from the nucleus, of cultured cortical neurons. As seen in the immunoblots of cortical neuron cultures, ADR1 expression was also more prominent in vivo following reperfusion, as compared to ADR2. Adiponectin is widely known to promote anti-inflammatory effects such as inhibition of NF-κB, TNF-α, IL-6 and IFN-γ, while increasing levels of IL-10 and IL1RA [16]. These effects are believed to confer protection against chronic disease conditions like atherosclerosis [17] and the metabolic syndrome [18]. To some extent, these protective effects are due to the ability of adiponectin to phosphorylate AMPK via ADRs [19,20]. Conversely, studies investigating the role of AMPK in neuronal survival/death have generated much controversy. The discrepancies, seen in studies performed in vitro, are most likely because of the differences in models, culture conditions [21], and the cell type used (e.g., transformed neural tumor cells versus primary cells). Recently, one group used a more direct approach by examining animals with selective gene deletion of AMPKα [10]. AMPKα knockout mice were protected from experimental ischemic stroke compared with wild-type controls. The beneficial effect of the AMPK inhibitor, Compound C, supported this detrimental effect of AMPK after stroke [10]. The combined results from these genetic and pharmacological approaches strongly suggest that activation of AMPK pathways in the brain is detrimental to neuronal survival following ischemia. Since ADR signaling is known to activate AMPK pathways, we hypothesized that the cerebral accumulation of adiponectin and its consequent ADRs activation following ischemic stroke could contribute to neuronal cell death.

hat activation of AMPK pathways in the brain is detrimental to neuronal survival following ischemia. Since ADR signaling is known to activate AMPK pathways, we hypothesized that the cerebral accumulation of adiponectin and its consequent ADRs activation following ischemic stroke could contribute to neuronal cell death. We therefore, examined the effect of adiponectin treatment in vitro on levels of cleaved caspase-3, a hallmark indicator of apoptosis, and phospho-AMPK, in cultured cortical neurons subjected to OGD. Consistent with our hypothesis, we found that ADR1 activation following adiponectin treatment during OGD, enhanced OGD-induced AMPK and caspase-3 activation, as well as neuronal death, thereby suggesting a pro-apoptotic role of ADR 1 activation by adiponectin contributes to neuronal death under ischemic conditions.

GD. Consistent with our hypothesis, we found that ADR1 activation following adiponectin treatment during OGD, enhanced OGD-induced AMPK and caspase-3 activation, as well as neuronal death, thereby suggesting a pro-apoptotic role of ADR 1 activation by adiponectin contributes to neuronal death under ischemic conditions. The stress-activated p38-MAPK plays important roles in transducing stress-related signals by phosphorylating intracellular enzymes, transcription factors and cytosolic proteins involved in apoptosis and inflammatory cytokine production. Sustained activation of p38-MAPK has been shown to be associated with neuronal cell death/apoptosis following ischemic stroke [22], and inhibition of this pathway is neuroprotective [23]. Our findings suggest that ADR-mediated p38-MAPK contributes to neuronal death after cerebral ischemia by promoting apoptotic cascades in neurons. Caspase-3 mediated apoptosis facilitates synaptic and neurite degeneration early in the ischemic neuronal death process [24], suggesting a role for this mechanism in the pathogenic actions of ADR signaling.

diated p38-MAPK contributes to neuronal death after cerebral ischemia by promoting apoptotic cascades in neurons. Caspase-3 mediated apoptosis facilitates synaptic and neurite degeneration early in the ischemic neuronal death process [24], suggesting a role for this mechanism in the pathogenic actions of ADR signaling. It has been proposed that the globular fragment of adiponectin is generated by proteolytic cleavage, and recently it has been shown that the cleavage of adiponectin by leukocyte elastase, secreted from activated monocytes and/or neutrophils, could be responsible for the generation of the globular fragment of adiponectin [25]. It has been shown previously that adiponectin exerts a cerebroprotective action through an endothelial nitric oxide synthase-dependent mechanism [26]. Nishimura and colleagues showed that Adiponectin-KO mice exhibited enlarged brain infarction and increased neurological deficits after ischemia-reperfusion compared with WT mice [26]. Conversely, systemic administration of adenoviral vectors expressing full-length murine adiponectin significantly reduced cerebral infarct size in WT and Adiponectin-KO mice. However, this murine adiponectin does not exclusively comprise the globular fraction and may have the ability to oligomerise into high, medium and low molecular weight oligomers, thereby inducing ADR2 activation. However, this study has not analysed the role of globular adiponectin which has a higher binding affinity towards ADR1.

s murine adiponectin does not exclusively comprise the globular fraction and may have the ability to oligomerise into high, medium and low molecular weight oligomers, thereby inducing ADR2 activation. However, this study has not analysed the role of globular adiponectin which has a higher binding affinity towards ADR1. The pathophysiological processes in stroke are complex and also involve disruption of the blood-brain barrier (BBB) [27]. It is noteworthy that in response to ischemic stroke, the microvasculature assumes an inflammatory phenotype characterized by leukocyte-endothelial cell adhesion, leukocyte capillary plugging, endothelial barrier dysfunction and activation of resident leukocytes including neutrophils [28]. Although the CNS is normally isolated from the immune system by the BBB, activated leukocytes can easily infiltrate the CNS once the BBB is disrupted during ischemic stroke. This disruption could also facilitate the penetration of full length adiponectin into injured brain tissues, that could be further cleaved by leukocytes elastases at the site of injury [25]. The pathophysiological importance of adiponectin cleavage by leukocyte elastase in vivo remains unclear. However, various studies using different cell types have reported a pro-inflammatory role for globular adiponectin [29-31]. These studies showed globular adiponectin to be a potent stimulator of NFκB and other pro-inflammatory genes, which could be detrimental during an inflammatory pathology like stroke. Another recently published study by Bråkenhielm and colleagues showed that adiponectin induced caspase mediated cell death in endothelial cells [32]. Notably, the physiological levels of adiponectin in both human and mouse serum have been reported to range from 2 to 17 μg/ml and also elevated following inflammatory disease conditions like preeclampsia and arthritis [29,33,34]. Thus, the concentration of adiponectin used in our study (10 μg/ml) is well within the observed physiological concentration.

of adiponectin in both human and mouse serum have been reported to range from 2 to 17 μg/ml and also elevated following inflammatory disease conditions like preeclampsia and arthritis [29,33,34]. Thus, the concentration of adiponectin used in our study (10 μg/ml) is well within the observed physiological concentration. Conclusions This study reveals a novel pathogenic role for adiponectin and adiponectin receptor activation in ischemic stroke. We show that cortical neurons express adiponectin receptors and levels of ADR1 increase substantially under ischemic conditions, and that addition of globular or trimeric adiponectin to neurons exacerbates cell death. Our results suggest that ischemia-induced neuronal ADR1 expression may increase the sensitivity of neurons to circulating levels of adiponectin following stroke, contributing to disease pathogenesis.

ally under ischemic conditions, and that addition of globular or trimeric adiponectin to neurons exacerbates cell death. Our results suggest that ischemia-induced neuronal ADR1 expression may increase the sensitivity of neurons to circulating levels of adiponectin following stroke, contributing to disease pathogenesis. List of abbreviations ADRS: Adiponectin receptors; AMP: Adenosine monophosphate; AMPK: AMP-activated protein kinase; APPL-1: Adaptor protein, phosphotyrosine interaction, PH domain and leucine zipper containing 1; BBB: Blood Brain Barrier; CNS: Central nervous system; CCA: Common carotid artery; ECA: External carotid artery; GAD: Globular adiponectin; GD: Glucose deprivation; ICA: Internal carotid artery; MAPK: Mitogen-activated protein kinase; MCA: Middle cerebral artery; MCAO: Middle cerebral artery occlusion; NFκB: Nuclear factor-kappa B; OGD: Oxygen and glucose deprivation; P38-MAPK: p38 mitogen-activated protein kinase; RT-PCR: Reverse transcription polymerase chain reaction; SDS-PAGE: Sodium dodecyl sulfate polyacrylamide gel electrophoresis; TAD: Trimeric adiponectin. Competing interests The authors declare that they have no competing interests.

List of abbreviations ADRS: Adiponectin receptors; AMP: Adenosine monophosphate; AMPK: AMP-activated protein kinase; APPL-1: Adaptor protein, phosphotyrosine interaction, PH domain and leucine zipper containing 1; BBB: Blood Brain Barrier; CNS: Central nervous system; CCA: Common carotid artery; ECA: External carotid artery; GAD: Globular adiponectin; GD: Glucose deprivation; ICA: Internal carotid artery; MAPK: Mitogen-activated protein kinase; MCA: Middle cerebral artery; MCAO: Middle cerebral artery occlusion; NFκB: Nuclear factor-kappa B; OGD: Oxygen and glucose deprivation; P38-MAPK: p38 mitogen-activated protein kinase; RT-PCR: Reverse transcription polymerase chain reaction; SDS-PAGE: Sodium dodecyl sulfate polyacrylamide gel electrophoresis; TAD: Trimeric adiponectin. Competing interests The authors declare that they have no competing interests. Authors' contributions JT designed the study, performed animal and cell culture experiments and analyzed data. ET and KS performed cell culture experiments and analyzed data. SCT and YIL collected human tissue and performed immunohistochemistry experiments. VTK, TMW, SMT, DGJ and MPM provided lab facilities, helped to interpret data and wrote the manuscript. TVM designed the study, provided lab facilities, helped to interpret data and wrote the manuscript. All authors read and approved the final manuscript. Acknowledgements This research was supported by the American Heart Association and Australian Heart Foundation (to T.V.A.) and by the Intramural Research Program of the National Institute on Aging.

Traumatic brain injury (TBI) is a result of an outside force causing immediate mechanical disruption of brain tissue and delayed pathogenic events which collectively mediate widespread neurodegeneration (reviewed by [1]). It is a heterogeneous disorder that can vary in the type of brain injury, distribution of brain damage and mechanisms of damage (Table 1). The primary damage of brain tissue can be diffuse or focal whereby the circumstances of injury determine the relative degree to which diffuse and focal trauma develops. Primary injury caused by direct impact to the head is considered to be largely focal, and results in cortical contusion, vascular injury and hemorrhages accompanied by ischemia. In contrast, diffuse brain injury characterized by diffuse axonal injury is caused by acceleration/deceleration forces. Depending upon the nature of primary injury, various cell responses are triggered that can exacerbate the injury. To date, these secondary injury processes are poorly understood. Table 1 Leading clinical causes and types of TBI in the United States 2002 - 2006 [2] Cause Percentage of Physical mechanism Primary brain injury TBI TBI-related deaths Falls 35.2% 18.9% impact resulting in the acceleration of the head and brain [125,126] closed head injury Road traffic accidents 17.3% 31.8% impact and acceleration of the head and brain [125,127] closed head injury

Table 1 Leading clinical causes and types of TBI in the United States 2002 - 2006 [2] Cause Percentage of Physical mechanism Primary brain injury TBI TBI-related deaths Falls 35.2% 18.9% impact resulting in the acceleration of the head and brain [125,126] closed head injury Road traffic accidents 17.3% 31.8% impact and acceleration of the head and brain [125,127] closed head injury Struck by/against events 16.5% 0.7% impact resulting in the acceleration of the head and brain [125,126] closed head injury TBI remains a leading cause of death and disability in the industralized countries [2,3] and represents a growing health problem also in the developing countries [4-7]; therefore even a modest outcome improvement could have major public health implications. As the immediate cell death resulting from the initial impact on the brain tissue is irreversible, treatments focus on interruption or inhibition of the secondary injury cascades expanding this primary injury. Nonetheless, no effective neuroprotective treatment is available so far [8-11]. The use of animal models is essential for better understanding of the secondary injury processes and for the development on novel therapies. Although large animal models may be necessary to investigate specific aspects of TBI, rodents (mice and rats) have emerged as the most commonly used species (for a review see [12]), since they are easily available to many researchers, normative data for a wide range of physiological and behavioral variables in rodents are well documented and transgenic technologies allow the generation of rodent lines with specific genetic alterations. A number of mouse and rat models have been developed to induce brain trauma. Of these the most commonly used are weight-drop injury, fluid percussion injury (FPI), and cortical contusion injury (CCI). However, the entire spectrum of events that might occur in TBI cannot be covered by one single rodent model. Therefore, this review evaluates the strengths and weaknesses of the currently available rodent models for TBI (Table 2).

e weight-drop injury, fluid percussion injury (FPI), and cortical contusion injury (CCI). However, the entire spectrum of events that might occur in TBI cannot be covered by one single rodent model. Therefore, this review evaluates the strengths and weaknesses of the currently available rodent models for TBI (Table 2). Table 2 Experimental rodent models of closed-head injury Model Species Injury Strengths Weaknesses Weight-drop models Feeney's weight-drop Shohami's weight-drop Marmarou's weight-drop rat [22] rat [14], mouse [15] rat [16,17], mouse [43] predominantly focal predominantly focal predominantly diffuse injury mechanism and inflicted injury is close to human TBI severity of injury can be adjusted well characterized neuroscoring immediately after injury allows randomization high mortality rate due to apnea and skull fractures not highly reproducible FP models MFP LFP rat [46,47] rat [48], mouse [49] mixed mixed severity of injury can be adjusted inflicted injury is highly reproducible within one laboratory requires craniotomy that may compensate for ICP increases no immediate post-injury neuroscoring possible inflicted injury is variable between laboratories high mortality rate due to apnea CCI rat [72], mouse [73] predominantly focal severity of injury can be adjusted inflicted injury is highly reproducible requires craniotomy no immediate post-injury neuroscoring Cryogenic brain lesion rat [92], mouse [93] focal severity of injury can be adjusted inflicted injury is highly reproducible and easily quantifiable mimics only conditionally human TBI

rat [72], mouse [73] predominantly focal severity of injury can be adjusted inflicted injury is highly reproducible requires craniotomy no immediate post-injury neuroscoring Cryogenic brain lesion rat [92], mouse [93] focal severity of injury can be adjusted inflicted injury is highly reproducible and easily quantifiable mimics only conditionally human TBI Weight-drop models The weight-drop models use the gravitational forces of a free falling weight to produce a largely focal [13-15] or diffuse [16-19] brain injury. The impact of the free falling weight is delivered to the exposed skull in rat [14] and mouse [20] or the intact dura in rat [21,22]. When the impact is delivered to the exposed skull, generally soft tips, e.g. silicon-covered [15] reduce the risk of skull fractures. For inducing focal brain injury, the animals are placed on non-flexible platforms in order to minimize dissipation of energy [13-15]. In contrast, crucial for inducing a diffuse brain injury is an impact widely distributed over the skull and the use of flexible platforms allowing the head to accelerate, e.g. foam-type platforms [16,17,19] or platforms with elastic springs [18]. The severity of head trauma can be varied by using different weights and/or heights of the weight-drop. The high mortality rate due to apnea can be reduced by early respiratory support and the usage of animals with a certain age and weight [16,17].

foam-type platforms [16,17,19] or platforms with elastic springs [18]. The severity of head trauma can be varied by using different weights and/or heights of the weight-drop. The high mortality rate due to apnea can be reduced by early respiratory support and the usage of animals with a certain age and weight [16,17]. Feeney's weight-drop model Typically this rat model in which an impact is delivered to the intact dura [21,22] results in a cortical contusion with hemorrhage [23] and damage of the blood-brain barrier [24,25]. Inflammatory processes lead to activation of microglia and astrocytes, activation of the complement system and invasion of neutrophils and macrophages [22,23,25-31]. Delayed microcirculatory disturbances and cortical spreading depression [32] have also been reported in this model. The pattern of posttraumatic cell death depends on the severity of impact [33]. Although the primary injury is largely focal, diffusely distributed axonal injury has been observed in the neuropil of the cortical lesion [23].

tory disturbances and cortical spreading depression [32] have also been reported in this model. The pattern of posttraumatic cell death depends on the severity of impact [33]. Although the primary injury is largely focal, diffusely distributed axonal injury has been observed in the neuropil of the cortical lesion [23]. Shohami's weight-drop model Shapira et al. and Chen et al., later introduced a model for closed-head injury using a weight-drop impact to one side of the unprotected skull in rat [14] and mouse [15,20], respectively. The injury severity in this model is dependent on the mass and falling height of the weight used. Thus, heavier weights and/or increased falling height produces an ipsilateral cortical brain contusion and blood-brain barrier disruption followed by brain edema, activation of the complement system, cell death evolving over time from the contusion site and invasion of inflammatory cells [13,15,23,34-37]. A modified model using lighter weights and/or shorter fall heights resulted in a concussive-like brain injury, bilateral cell loss, short duration of brain edema and long-lasting cognitive deficits [23]. Moreover, bilateral diffuse brain damage, cell death (bilateral and beneath the impact site), and inflammatory responses were reported [38-41]. In general mild weight-drop injuries are associated with a diffuse injury pattern whereas more severe weight-drop injuries produce a focal contusion. A disadvantage of the weight-drop model is the high variability of the injury severity. A major advantage of this model is that it can be quickly performed under gas-anesthesia and thus allows neurological scoring immediately after injury [15,20]. Thus clinically relevant randomization of animals into the various treatment groups is possible.

ight-drop model is the high variability of the injury severity. A major advantage of this model is that it can be quickly performed under gas-anesthesia and thus allows neurological scoring immediately after injury [15,20]. Thus clinically relevant randomization of animals into the various treatment groups is possible. Marmarou's weight-drop model (Impact acceleration model) To model "whole head" motion resulting in a diffuse brain injury, Marmarou et al. [16,17] allows the head to accelerate at impact. Depending on the severity of injury, the induced brain injury results in hemorrhages, neuronal cell death, astrogliosis, diffuse axonal injury, and cytotoxic brain edema [17,23,26,42,43]. This impact acceleration model using a weight-drop is a useful model for investigating diffuse brain injuries ranging from mild to severe. Taken together, weight-drop models provide a straightforward way to assess brain injuries close to the clinical conditions ranging from focal to diffuse brain injuries. Fluid percussion injury models Fluid percussion injury (FPI) models produce brain injury by rapidly injecting fluid volumes onto the intact dural surface through a craniotomy. The craniotomy is made either centrally (CFP, MFP), over the sagittal suture midway between bregma and lambda, or laterally (LFP), over the parietal cortex. Graded levels of injury severity can be achieved by adjusting the force of the fluid pressure pulse. Like in various other TBI models, a high mortality rate due to apnea is evident [44,45].

her centrally (CFP, MFP), over the sagittal suture midway between bregma and lambda, or laterally (LFP), over the parietal cortex. Graded levels of injury severity can be achieved by adjusting the force of the fluid pressure pulse. Like in various other TBI models, a high mortality rate due to apnea is evident [44,45]. The central (CFP) and lateral (LFP) fluid percussion injury models were adapted to rats in 1987 [46,47] and in 1989 [48,49], respectively. These models produce a mixed type of brain injury. Traumatic pathology includes cortical contusion, hemorrhage and a cytotoxic and/or vasogenic brain edema either typically bilateral for CFP injury or ipsilateral for LFP injury [23,26,50]. The delayed progression of brain damage is accompanied by astrogliosis, diffuse axonal injury, inflammatory events, cortical spreading depression and neurodegeneration [23,26,45,50-61]. Regardless of injury location, FPI leads to cognitive dysfunction [23,51,55,61,62] and thus it can be a useful model for posttraumatic dementia. Furthermore, FPI delivered laterally is an appropriate model for posttraumatic epilepsy [63].

ory events, cortical spreading depression and neurodegeneration [23,26,45,50-61]. Regardless of injury location, FPI leads to cognitive dysfunction [23,51,55,61,62] and thus it can be a useful model for posttraumatic dementia. Furthermore, FPI delivered laterally is an appropriate model for posttraumatic epilepsy [63]. The FPI model in the rat has been the most widely used model for TBI. Nevertheless, for both CFP and LFP variability's in injury parameters between laboratories are evident. For instance, initial studies using LFP detected an ipsilateral brain injury [64] whereas some later studies detected a widespread, bilateral brain injury [65-67]. One crucial factor determining the outcome severity in this model seems to be the positioning of the craniotomy as already a small shift in the craniotomy site is associated with marked differences in neurological outcome, lesion location and size [68,69]. Thus, establishing a FPI model necessitates extensive methodological fine-tuning to obtain a standardized outcome in respect to its severity and pathophysiology. Once the FPI model is established, the induced brain trauma seems to be highly reproducible. To enable the use of transgenic mice, Carbonell et al. [49] adapted the FPI model to the mouse. Similar to the rat, the inflicted injury in mice leads to cognitive dysfunction, microglial activation and neuronal and axonal damage [23,49,51,63,70,71].

The FPI model in the rat has been the most widely used model for TBI. Nevertheless, for both CFP and LFP variability's in injury parameters between laboratories are evident. For instance, initial studies using LFP detected an ipsilateral brain injury [64] whereas some later studies detected a widespread, bilateral brain injury [65-67]. One crucial factor determining the outcome severity in this model seems to be the positioning of the craniotomy as already a small shift in the craniotomy site is associated with marked differences in neurological outcome, lesion location and size [68,69]. Thus, establishing a FPI model necessitates extensive methodological fine-tuning to obtain a standardized outcome in respect to its severity and pathophysiology. Once the FPI model is established, the induced brain trauma seems to be highly reproducible. To enable the use of transgenic mice, Carbonell et al. [49] adapted the FPI model to the mouse. Similar to the rat, the inflicted injury in mice leads to cognitive dysfunction, microglial activation and neuronal and axonal damage [23,49,51,63,70,71]. Controlled cortical impact injury model Controlled cortical impact (CCI) models utilize a pneumatic pistol to deform laterally the exposed dura and provide controlled impact and quantifiable biomechanical parameters. This model was adapted to rat in 1991 [72] and to mouse in 1995 [73] and produces graded, reproducible brain injury.

rtical impact injury model Controlled cortical impact (CCI) models utilize a pneumatic pistol to deform laterally the exposed dura and provide controlled impact and quantifiable biomechanical parameters. This model was adapted to rat in 1991 [72] and to mouse in 1995 [73] and produces graded, reproducible brain injury. Dependent on the severity of injury, CCI results in an ipsilateral injury with cortical contusion, hemorrhage and blood-brain barrier disruption [74]. Neuronal cell death and degeneration, astrogliosis, microglial activation, inflammatory events, axonal damage, cognitive deficits, excitotoxicity and cortical spreading depressions are reported to ensue [23,26,30,73,75-82]. Particularly with regard to brain edema, CCI is an important model as it presumably causes a cytotoxic and a vasogenic brain edema [23,26,83-89] and thus it reflects the clinical situation of posttraumatic brain edema formation. The predominantly focal brain injury caused by CCI makes this model to a useful tool for studying the pathophysiology of the secondary processes induced by focal brain injury. Interestingly, CCI in rodents is associated with posttraumatic seizure activity similar to the injury-induced epilepsy in humans [90,91]. Thus this model is particularly suitable to study pathomechanisms of posttraumatic epilepsy.

or studying the pathophysiology of the secondary processes induced by focal brain injury. Interestingly, CCI in rodents is associated with posttraumatic seizure activity similar to the injury-induced epilepsy in humans [90,91]. Thus this model is particularly suitable to study pathomechanisms of posttraumatic epilepsy. Cryogenic injury model The method of cryogenic injury in rodents [92,93] leads to a focal brain lesion. The brain injury in this model is generally produced by applying a cold rod to the exposed dura in rats (e.g. on the parietal cortex using a copper cylinder filled with an mixture of acetone and dry ice (-78°C) [94]) or skull in mice (e.g. on the parietal cortex using a copper cylinder filled with liquid nitrogen (-183°C) [95]). In some studies, a dry ice pellet was directly applied to the skull of the rat or mouse [96,97]. Different injury severities can be achieved by varying the contact time to the exposed cortex [98].

°C) [94]) or skull in mice (e.g. on the parietal cortex using a copper cylinder filled with liquid nitrogen (-183°C) [95]). In some studies, a dry ice pellet was directly applied to the skull of the rat or mouse [96,97]. Different injury severities can be achieved by varying the contact time to the exposed cortex [98]. In rodents, cortical cryogenic injury results in a focal brain lesion and breakdown of the blood-brain barrier [94,95]. The primary lesion is surrounded by a penumbral zone where secondary processes lead to an extension of lesion size accompanied by neuronal cell death and cytotoxic and vasogenic edema [98,99]. These secondary processes also include activation of astrocytes and inflammation [95,96,100-103]. Moreover, it was reported recently that a discrete cryogenic lesion to the parietal cortex of juvenile mice causes delayed global neurodegeneration [104]. Due to epileptic activities surrounding the focal lesion, this method is also used for mimicking certain aspects of epilepsy [105-107].

inflammation [95,96,100-103]. Moreover, it was reported recently that a discrete cryogenic lesion to the parietal cortex of juvenile mice causes delayed global neurodegeneration [104]. Due to epileptic activities surrounding the focal lesion, this method is also used for mimicking certain aspects of epilepsy [105-107]. The cryogenic brain lesion model is particularly suited for investigating TBI-associated blood-brain barrier leakage and vasogenic brain edema. However, this focal trauma model lacks the countracoup and diffuse axonal injuries that typically complicate human head injuries [1]. Thus the cryogenic brain lesion model only conditionally mimics the clinical situation. Although various other models reflect more realistic the pathophysiological characteristic of TBI, the cryogenic brain lesion model has one major advantage: The lesions caused by the cryogenic injury model are clearly circumscribed and highly reproducible in size, location and pathophysiological processes of the secondary lesion expansion at the cortical impact site. The high reproducibility of the cortical lesion is particularly useful to screen the impact of pharmacological treatments or gene knockout on secondary lesion development after focal brain injury.

roducible in size, location and pathophysiological processes of the secondary lesion expansion at the cortical impact site. The high reproducibility of the cortical lesion is particularly useful to screen the impact of pharmacological treatments or gene knockout on secondary lesion development after focal brain injury. Other models Models to induce diffuse brain injury In addition to the original Marmarou's weight-drop model, various other impact acceleration models that induce diffuse brain injuries have been described in the literature. As an example, in one model the rat is placed on its back while the head is accelerated upward by a piston [108]. The severity of injury depends on the impact velocity of the piston. In another study, rats were subjected to impact acceleration head injury, using a pneumatic impact targeted to a steel disc centered onto their skull. The animal's head was supported by a soft pad to decelerate the head after the impact [109]. To induce moderate head concussion without focal injury, a pendulum can be used that stroke on the skull midline of rats [110].

eleration head injury, using a pneumatic impact targeted to a steel disc centered onto their skull. The animal's head was supported by a soft pad to decelerate the head after the impact [109]. To induce moderate head concussion without focal injury, a pendulum can be used that stroke on the skull midline of rats [110]. Models to induce focal brain injury In an attempt to create a model of focal cerebral contusion without diffuse brain injury, Shreiber et al. (1999) generated cerebral contusions and associated evolving damage by a transient non-ablative vacuum pulse applied to the exposed cerebral cortex [111]. Other models designed to generate focal cortical injury inject fluids leading to an inflammatory response and a progressive cavitation [112], apply a mechanical suction force through the intact dura [113] or apply a stab wound [114]. Each of these models result in clearly circumcised focal lesions and thus, similar to the cryogenic injury model, they might be helpful in studies evaluating putative treatments by monitoring the focal lesion size.

tation [112], apply a mechanical suction force through the intact dura [113] or apply a stab wound [114]. Each of these models result in clearly circumcised focal lesions and thus, similar to the cryogenic injury model, they might be helpful in studies evaluating putative treatments by monitoring the focal lesion size. Models to mimic blast-induced neurotrauma In recent years, exposure to blast is becoming more frequent foremost in military populations. Brain injuries due to blast are caused by particles propelled by blast-force, acceleration and declaration forces and/or the blast wave itself [115]. The non-impact blast injury exhibits an interesting pathophysiology characterized by diffuse cerebral brain edema, extreme hyperemia and a delayed vasospasm [115]. To investigate blast-induced neurotrauma different models have been established. As an example, to mimic a non-impact blast-induced neurotrauma, rodents were fixed and exposed to blast waves caused by detonation of explosive [116] or compressed air [117]. Recently the pathobiology of TBI caused by blast and the animal models for non-impact blast injury have been recently reviewed by Cernak and Noble-Haeusslein [115].

imic a non-impact blast-induced neurotrauma, rodents were fixed and exposed to blast waves caused by detonation of explosive [116] or compressed air [117]. Recently the pathobiology of TBI caused by blast and the animal models for non-impact blast injury have been recently reviewed by Cernak and Noble-Haeusslein [115]. Combined and modified injury models If no convenient model is available to address specific research topics, the modification of already existing animal models might be useful. As an example human TBI is often induced by angular (a combination of linear and rotational) accelerations, e.g. TBI caused by car accidents. This clinical scenario was mimicked in rats by instantly rotating the animal to reproduce rotational acceleration after it had sustained the impact that produced linear acceleration using the Marmarou's weight drop model [118]. Another example is the Maryland model, in which Kilbourne et al. mimicked a frontal impact by modifying the impact-acceleration model of Marmarou [119]. To simulate concussions in National Football League players, a rat model was developed in which a pneumatic pressure in the style of CCI models is used to impact laterally the helmet-protected head [120]. Clinical TBI is frequently accompanied by complications such as hypoxic episodes and sepsis. In order to mimic those clinical situations, they can be integrated in the study design (hypoxia [121,122] and sepsis [123]).

atic pressure in the style of CCI models is used to impact laterally the helmet-protected head [120]. Clinical TBI is frequently accompanied by complications such as hypoxic episodes and sepsis. In order to mimic those clinical situations, they can be integrated in the study design (hypoxia [121,122] and sepsis [123]). Cell culture models Cell culture is currently the most promising alternative to animal research. The use of cell culture models simulating TBI might be useful for certain research goals, such as high throughput drug screenings or the assessment of the effect of trauma on individual cell types. The current available cell culture models include models using disruption of various cell cultures by laceration, compression, acceleration or stretch injury (reviewed by [124]).

seful for certain research goals, such as high throughput drug screenings or the assessment of the effect of trauma on individual cell types. The current available cell culture models include models using disruption of various cell cultures by laceration, compression, acceleration or stretch injury (reviewed by [124]). Outlook Initially, the rodent models for TBI were designed to mimic closely the clinical sequelae of human TBI. In this respect, the most straightforward rodent models are the weight-drop models by Marmarou and Shohami as they closely mimic the real life TBI. The inflicted injuries are predominantly diffuse or focal in nature, respectively. Similarly the FPI model and CCI model mimic various injury processes associated with human TBI. Probably due to the excellent reproducibility of induced brain trauma, FPI and CCI are the most widely used rodent models for TBI. However, even small modifications in the experimental design often lead to differences in primary injury and hence to differences in pathobiological processes leading to secondary injury. Considering the heterogeneity of human TBI, scientific hypothesis should be tested in multiple rodent models resulting in distinct types of injury. Thus, models solely mimicking focal or diffuse injury are needed. In conclusion, there are numerous rodent models of TBI available, widely varying in their ability to model pathomechanisms associated with human TBI. They provide the experimental backbone for investigating TBI pathomechanisms and for the initial testing of neuroprotective compounds.

Outlook Initially, the rodent models for TBI were designed to mimic closely the clinical sequelae of human TBI. In this respect, the most straightforward rodent models are the weight-drop models by Marmarou and Shohami as they closely mimic the real life TBI. The inflicted injuries are predominantly diffuse or focal in nature, respectively. Similarly the FPI model and CCI model mimic various injury processes associated with human TBI. Probably due to the excellent reproducibility of induced brain trauma, FPI and CCI are the most widely used rodent models for TBI. However, even small modifications in the experimental design often lead to differences in primary injury and hence to differences in pathobiological processes leading to secondary injury. Considering the heterogeneity of human TBI, scientific hypothesis should be tested in multiple rodent models resulting in distinct types of injury. Thus, models solely mimicking focal or diffuse injury are needed. In conclusion, there are numerous rodent models of TBI available, widely varying in their ability to model pathomechanisms associated with human TBI. They provide the experimental backbone for investigating TBI pathomechanisms and for the initial testing of neuroprotective compounds. Competing interests The authors declare that they have no competing interests. Authors' contributions CAW drafted the manuscript. ALS corrected and wrote the final manuscript. Both authors read and approved the final manuscript. Acknowledgements This work received financial support from the Wilhelm Sander Foundation (Munich, Germany).

Introduction Thromboembolic occlusion of intracerebral vessels is responsible for the majority of ischemic strokes [1]. Studies on the early use of anticoagulant drugs (e.g. heparin) in cerebral ischemia failed to demonstrate overall benefit in that reduced lesion progression was counterbalanced by an increase in hemorrhages [2]. In addition, long-term anticoagulation for secondary prevention of cardioembolic stroke, mainly accomplished by warfarin prescription, is inevitably associated with increased bleeding-related morbidity and mortality [3]. Hence, identification of novel targets for more effective and safer anticoagulation in patients with imminent stroke is badly needed.

tion for secondary prevention of cardioembolic stroke, mainly accomplished by warfarin prescription, is inevitably associated with increased bleeding-related morbidity and mortality [3]. Hence, identification of novel targets for more effective and safer anticoagulation in patients with imminent stroke is badly needed. In the classical "waterfall model" of blood coagulation the formation of a fibrin thrombus can be initiated by two distinct pathways, the extrinsic and the intrinsic pathway [4]. Both cascades consist of a series of proteolytic reactions involving trypsin-like serine proteases [5]. Fibrin formation via the extrinsic pathway occurs when tissue factor (TF), located on cell membranes in the subendothelium of a vessel, forms a complex with activated coagulation factor VII (FVIIa) after endothelial injury [6]. According to the original assumption, the intrinsic pathway is initiated when coagulation factor XII (FXII) becomes activated on a negatively charged surface followed by successive activation of factor XI (FXI) and factor IX (FIX) [7]. FXII has long been considered to be dispensable for clot formation because humans with hereditary FXII deficiency suffer from neither spontaneous nor injury-related abnormal bleedings [8,9]. This concept was recently called into question by studies demonstrating that FXII-deficient mice are profoundly protected from pathological thrombus formation in different models of arterial thrombosis but, like FXII-deficient humans, do not show impaired hemostasis [10,11]. Consequently, it was anticipated that the use of FXII inhibitors would be associated with relatively low rates of therapy-related hemorrhages, the major clinical complication associated with current anticoagulant therapies [1]. Indeed, wild-type mice treated with D-Pro-Phe-Arg chloromethyl ketone (PCK), which blocks the amidolytic activity of FXIIa, and subjected to ischemic stroke afterwards, developed less vessel occlusive thrombi in the cerebral microvasculature but did not show increased bleeding tendencies [11]. However, PCK is not selective over FXII and also interacts with other components of the plasmatic coagulation cascade [12].

olytic activity of FXIIa, and subjected to ischemic stroke afterwards, developed less vessel occlusive thrombi in the cerebral microvasculature but did not show increased bleeding tendencies [11]. However, PCK is not selective over FXII and also interacts with other components of the plasmatic coagulation cascade [12]. Pochet and co-workers recently described the synthesis of new 3-carboxamide-coumarins which are the first selective nonpeptidic inhibitors of FXIIa [12]. COU254 is a member of this novel class of FXII inhibitors. In the present study we assessed the effect of COU254 on stroke development, intracerebral fibrinogen clotting and post stroke functional outcome in mice.

nthesis of new 3-carboxamide-coumarins which are the first selective nonpeptidic inhibitors of FXIIa [12]. COU254 is a member of this novel class of FXII inhibitors. In the present study we assessed the effect of COU254 on stroke development, intracerebral fibrinogen clotting and post stroke functional outcome in mice. Methods Animal experiments A total of 26 mice were used in this study. Animal experiments were approved by legal state authorities (Bezirksregierung of Unterfranken) and conducted according to the recommendations for research in basic stroke studies [13]. Focal cerebral ischemia was induced in 6-8-weeks old male C57Bl/6 mice (Harlan Winkelmann, Borchen, Germany) by 60 min transient middle cerebral artery occlusion (tMCAO) as described [11,14]. Mice were anesthetized with 2.5% isoflurane (Abbott, Wiesbaden, Germany). Following a midline skin incision in the neck, the proximal common carotid artery and the external carotid artery were ligated and a standardized silicon rubber-coated 6.0 nylon monofilament (6021; Doccol Corp., Redlands, CA, USA) was inserted and advanced via the right internal carotid artery to occlude the origin of the right MCA. The operator was blinded to the treatment groups and operation time per animal did not exceed 15 minutes. The intraluminal suture was left in situ for 60 minutes. Then animals were re-anesthetized and the occluding monofilament was withdrawn to allow reperfusion.

l carotid artery to occlude the origin of the right MCA. The operator was blinded to the treatment groups and operation time per animal did not exceed 15 minutes. The intraluminal suture was left in situ for 60 minutes. Then animals were re-anesthetized and the occluding monofilament was withdrawn to allow reperfusion. COU254 dissolved in 25% DMSO was administered intraperitoneally (i.p.) 2 h before the induction of tMCAO at a dosage of 40 mg/kg bodyweight. Vehicle-treated control mice receiving 25% DMSO without COU254 served as controls. The i.p. route of administration was chosen because i.v. application was afflicted with significant acute toxicity and mortality in preliminary experiments. Determination of infarct size and histology Edema-corrected infarct volumes were quantified by planimetry from 2,3,5-Triphenyltetrazoliumchloride (TTC)-stained brain sections 24 h after ischemic stroke as described [11,14,15]. For morphological assessment, paraffin embedded brains were stained with hematoxylin and eosin (H&E). Protein extraction and Western blot analysis Following TTC staining cortices were dissected from formalin-fixed brain slices and homogenized in RIPA buffer (25 mM Tris pH 7.4, 150 mM NaCl, 1% NP40) containing 2% SDS. The samples were incubated for 20 min at 100°C followed by incubation at 60°C for 2 h [16]. After that, tissue lysates were centrifuged at 15.000 × g for 20 min at 4°C and supernatants were used for BCA protein assay and subsequent Western blot analysis.

ffer (25 mM Tris pH 7.4, 150 mM NaCl, 1% NP40) containing 2% SDS. The samples were incubated for 20 min at 100°C followed by incubation at 60°C for 2 h [16]. After that, tissue lysates were centrifuged at 15.000 × g for 20 min at 4°C and supernatants were used for BCA protein assay and subsequent Western blot analysis. The total lysates were treated with SDS-PAGE loading buffer (final conc. 65 mM Tris, 5% 2-mercaptoethanol, 3% SDS, 10% glycerol) at 95°C for 5 min. 30 μg of total protein were electrophoresed and transferred to a PVDF membrane. After blocking for 1 h with blocking buffer (5% nonfat dry milk, 50 mM Tris-HCl pH 7.5, 150 mM NaCl, 0.05% Tween-20) membranes were incubated with the primary antibody at 4°C over night at the following dilutions: anti-Fibrinogen (cross-reactive against fibrin) pAb 1:500 (Acris Antibodies) and anti-Actin mAb 1:10,000 (Dianova). After a washing step with TBS-T (50 mM Tris-HCl pH 7.5, 150 mM NaCl, 0.05% Tween-20), membranes were incubated for 1 h with HRP-conjugated donkey anti-rabbit IgG (for Fibrinogen) or donkey anti-mouse IgG (for Actin) at a dilution of 1:5000 and were finally developed using ECLplus (GE Healthcare).

Ab 1:10,000 (Dianova). After a washing step with TBS-T (50 mM Tris-HCl pH 7.5, 150 mM NaCl, 0.05% Tween-20), membranes were incubated for 1 h with HRP-conjugated donkey anti-rabbit IgG (for Fibrinogen) or donkey anti-mouse IgG (for Actin) at a dilution of 1:5000 and were finally developed using ECLplus (GE Healthcare). Assessment of functional outcome 24 h after tMCAO the modified Bederson score [17] was used to determine global neurological function according to the following scoring system: 0, no deficit; 1, forelimb flexion; 2, decreased resistance to lateral push; 3, unidirectional circling; 4, longitudinal spinning; 5, no movement. Motor function and coordination were evaluated by the grip test [18]. For this test, the mouse was placed midway on a string between two supports and rated as follows: 0, falls off; 1, hangs onto string by one or both forepaws; 2, as for 1, and attempts to climb onto string; 3, hangs onto string by one or both forepaws plus one or both hindpaws; 4, hangs onto string by fore- and hindpaws plus tail wrapped around string; 5, escape (to the supports). Neurological scores were always assessed by an independent and blinded investigator. Laser-Doppler flowmetry Laser-Doppler flowmetry (Moor Instruments, Axminster, U.K.) was used in some animals (n = 3/group) to monitor regional cerebral blood flow (rCBF) in the MCA territory (6 mm lateral and 2 mm posterior from bregma) [19].

Assessment of functional outcome 24 h after tMCAO the modified Bederson score [17] was used to determine global neurological function according to the following scoring system: 0, no deficit; 1, forelimb flexion; 2, decreased resistance to lateral push; 3, unidirectional circling; 4, longitudinal spinning; 5, no movement. Motor function and coordination were evaluated by the grip test [18]. For this test, the mouse was placed midway on a string between two supports and rated as follows: 0, falls off; 1, hangs onto string by one or both forepaws; 2, as for 1, and attempts to climb onto string; 3, hangs onto string by one or both forepaws plus one or both hindpaws; 4, hangs onto string by fore- and hindpaws plus tail wrapped around string; 5, escape (to the supports). Neurological scores were always assessed by an independent and blinded investigator. Laser-Doppler flowmetry Laser-Doppler flowmetry (Moor Instruments, Axminster, U.K.) was used in some animals (n = 3/group) to monitor regional cerebral blood flow (rCBF) in the MCA territory (6 mm lateral and 2 mm posterior from bregma) [19]. Statistics Data are expressed as mean ± standard deviation (SD). For statistical analysis, PrismGraph 4.0 software package (La Jolla, CA, USA) was used. Infarct volumes and neurological scores were analyzed using the non-parametric Mann Whitney test. Laser Doppler flowmetry data were compared by 1-way ANOVA followed by Bonferroni post hoc test. P-values < 0.05 were considered to be statistically significant.

PrismGraph 4.0 software package (La Jolla, CA, USA) was used. Infarct volumes and neurological scores were analyzed using the non-parametric Mann Whitney test. Laser Doppler flowmetry data were compared by 1-way ANOVA followed by Bonferroni post hoc test. P-values < 0.05 were considered to be statistically significant. Results The transient middle cerebral artery filament occlusion model (tMCAO) was used to induce focal cerebral ischemia in mice [11,14,15]. After advancing the filament to the origin of the MCA the decrease in rCBF was similar between control mice and COU254-treated mice (5.4 ± 7.5% of baseline levels vs. 11.5 ± 5.2% of baseline levels; p > 0.05) (Figure 1). Ten minutes after reperfusion rCBF in the MCA territory was reconstituted to >70% of baseline levels and again did not significantly differ between treated and untreated mice (70.0 ± 10.3% of baseline levels vs. 74.3 ± 4.2% of baseline levels; p > 0.05) (Figure 1). Taken together, these findings exclude any significant rCBF alterations related to COU254 or vehicle treatment and prove that MCA occlusion and reperfusion were sufficient in our model.

y differ between treated and untreated mice (70.0 ± 10.3% of baseline levels vs. 74.3 ± 4.2% of baseline levels; p > 0.05) (Figure 1). Taken together, these findings exclude any significant rCBF alterations related to COU254 or vehicle treatment and prove that MCA occlusion and reperfusion were sufficient in our model. Figure 1 rCBF does not differ between COU254-treated mice and controls after tMCAO. Determination of regional cerebral blood flow (rCBF) using Laser Doppler flowmetry before the occlusion of the middle cerebral artery (baseline), 10 min after the occlusion (ischemia) and again 10 min after the removal of the filament (reperfusion) in COU254-treated mice and vehicle-treated controls (n = 3/group). No significant differences in rCBF were observed between the two groups. 1-way ANOVA, Bonferroni post hoc test, ns = not significant.

tery (baseline), 10 min after the occlusion (ischemia) and again 10 min after the removal of the filament (reperfusion) in COU254-treated mice and vehicle-treated controls (n = 3/group). No significant differences in rCBF were observed between the two groups. 1-way ANOVA, Bonferroni post hoc test, ns = not significant. As a next step, we determined infarct sizes and the extent of neuronal damage in control mice and mice treated with COU254. 24 h after tMCAO no significant differences in infarct volumes were observed between the two groups as revealed by 2,3,5-Triphenyltetrazoliumchloride (TTC) staining and successive infarct planimetry (101.5 ± 31.4 mm3 vs. 110.0 ± 27.2 mm3; p > 0.05) (Figure 2a). In line with these findings, H&E staining confirmed widespread ischemic neurodegeneration in both groups which regularly expanded to the basal ganglia and the neocortex (Figure 2a). Detailed analysis of the neurological status using two different functional scores also could not reveal any beneficial effects of COU254 in acute ischemic stroke (Bederson score: 2.8 ± 1.4 vs. 2.4 ± 1.8; p > 0.05; grip test: 2.0 ± 1.4 vs. 2.5 ± 1.9; p > 0.05) (Figure 2b). Finally, no differences in thrombus formation within the infarcted brain areas of COU254-treated mice or controls were detectable by immunoblot (Figure 2c).

eneficial effects of COU254 in acute ischemic stroke (Bederson score: 2.8 ± 1.4 vs. 2.4 ± 1.8; p > 0.05; grip test: 2.0 ± 1.4 vs. 2.5 ± 1.9; p > 0.05) (Figure 2b). Finally, no differences in thrombus formation within the infarcted brain areas of COU254-treated mice or controls were detectable by immunoblot (Figure 2c). Figure 2 COU254 does not improve outcome after experimental stroke in mice. Infarct size and functional outcome in COU254-treated mice and controls (vehicle) 24 h after 60 min transient middle cerebral artery occlusion (tMCAO). (a) (top) Representative 2,3,5-Triphenyltetrazoliumchloride (TTC)-stained coronal brain sections from the two animal groups. Ischemic infarctions appear white and regularly include the neocortex and basal ganglia as confirmed by hematoxylin and eosin (H&E) staining (bar represents 250 μm). (bottom) Infarct volumes on day 1 after tMCAO in COU254-treated mice and vehicle-treated controls as determined by planimetry (n = 10/group). Non-parametric Mann Whitney test, ns = not significant. (b) Neurological Bederson score and grip test score on day 1 after tMCAO in COU254-treated mice and vehicle-treated controls. In line with the results on infarct volumes, no significant functional differences became apparent between the treatment groups. Non-parametric Mann Whitney test, ns = not significant. (c) Accumulation of fibrin(ogen) in the infarcted (+) and contralateral (-) cortices of COU254-treated mice or vehicle-treated controls. Fibrinogen clotting 24 h after ischemia was analyzed by immunoblotting. Two representative immunoblots of each group are shown.

Non-parametric Mann Whitney test, ns = not significant. (c) Accumulation of fibrin(ogen) in the infarcted (+) and contralateral (-) cortices of COU254-treated mice or vehicle-treated controls. Fibrinogen clotting 24 h after ischemia was analyzed by immunoblotting. Two representative immunoblots of each group are shown. Discussion Unexpectedly, the selective nonpeptidic FXIIa inhibitor COU254 could not ameliorate ischemic brain damage after tMCAO in mice.

Non-parametric Mann Whitney test, ns = not significant. (c) Accumulation of fibrin(ogen) in the infarcted (+) and contralateral (-) cortices of COU254-treated mice or vehicle-treated controls. Fibrinogen clotting 24 h after ischemia was analyzed by immunoblotting. Two representative immunoblots of each group are shown. Discussion Unexpectedly, the selective nonpeptidic FXIIa inhibitor COU254 could not ameliorate ischemic brain damage after tMCAO in mice. COU254 belongs to a new group of recently described 3-carboxamide-coumarins which represent the first selective inhibitors of FXIIa [12]. In contrast to conventional FXII inhibitors it does not interfere with other components of the contact activation system or plasmatic coagulation cascade potentially involved in stroke development such as kallikrein, FXa or the TF/FVIIa complex [15]. Mechanistically, COU254 mediates anti-FXII activity through the formation of an acyl enzyme instead of an alkyl enzyme as observed with thrombin and 6-chloromethyl ester coumarins [12]. Several reports have meanwhile highlighted the role of FXII in pathological thrombogenesis [20]. FXII-deficient mice were protected against arterial thrombosis, collagen-induced venous thromboembolism and ischemic stroke [10,11]. Importantly, FXII-/- mice display normal hemostasis and consequently FXIIa inhibition was not associated with increased bleeding complications supporting the intriguing hypothesis that hemostasis and thrombosis are two mechanistically different processes [20]. Hence, FXIIa is considered an attractive target for pharmacological inhibitors designed to treat or prevent thromboembolic disorders. Such a safe therapy might be particularly advantageous for the treatment of acute ischemic stroke, where the conventional anticoagulants used against stroke progression or recurrence are inheritably associated with increased bleeding-related morbidity and mortality [1].

ned to treat or prevent thromboembolic disorders. Such a safe therapy might be particularly advantageous for the treatment of acute ischemic stroke, where the conventional anticoagulants used against stroke progression or recurrence are inheritably associated with increased bleeding-related morbidity and mortality [1]. We recently established D-Pro-Phe-Arg chloromethyl ketone (PCK) as FXIIa inhibitor in experimental stroke [11]. Mice treated with PCK immediately before the occlusion of the middle cerebral artery (MCA) developed smaller infarcts and less severe neurological deficits compared to controls. Moreover, the formation of fibrin within the infarcted brains was significantly reduced. Although PCK irreversibly inhibits the amidolytic activity of FXIIa it is not specific over FXII. Rather, PCK has been shown to block other components of the contact activation system or plasmatic coagulation cascade bearing the potential risk of undesired adverse effects [12]. Moreover, the peptidic structure and the alkylating behavior of the chloromethyl function prevent the application of PCK as oral anticoagulant. Natural anticoagulant proteins displaying anti-FXIIa activity were also reported, e.g. from leguminous plants [21], hematophagous insects [22-24], helminth parasites [25] and bacteria [26]. Again, despite their proven efficacy, all these proteins were generally not selective over blood coagulation proteases.

coagulant. Natural anticoagulant proteins displaying anti-FXIIa activity were also reported, e.g. from leguminous plants [21], hematophagous insects [22-24], helminth parasites [25] and bacteria [26]. Again, despite their proven efficacy, all these proteins were generally not selective over blood coagulation proteases. Several reasons might account for the negative results in present study. Besides "true" inefficacy of COU254 in acute experimental stroke related for example to the relatively low FXIIa inhibitory potency of COU254 compared to PCK [12] technical limitations could have been responsible. The anti-FXIIa activity of COU254 has only been established from in vitro dose-response curves so far and pharmacodynamic or pharmacokinetic data on COU254 in animals, especially rodents, are lacking. Moreover, the optimum dosage or route of application of COU254 in mice is yet unknown as is the ideal time point of administration during the course of ischemic stroke. Because human and mouse FXII share a high degree of sequence homology and the established human FXII inhibitors usually also block murine FXII [10,11], species specific differences of COU254 mode of action between humans and rodents seem unlikely but cannot be completely ruled out. In summary, 3-carboxamide-coumarins represent a promising new class of selective FXII inhibitors but further preclinical evaluation of these compounds in animal models is clearly needed before any firm conclusions on their antithrombotic potential can be drawn. Competing interests The authors declare that they have no competing interests.

In summary, 3-carboxamide-coumarins represent a promising new class of selective FXII inhibitors but further preclinical evaluation of these compounds in animal models is clearly needed before any firm conclusions on their antithrombotic potential can be drawn. Competing interests The authors declare that they have no competing interests. Authors' contributions All authors have read and approved the final manuscript. PK operated the animals, assessed the functional scores and interpreted the data. TS performed the immunoblots. LP provided COU254 and finalized the manuscript. GS and CK conceived the experiments, funded the project and wrote the manuscript. Acknowledgements The expert technical assistance of Melanie Glaser and Daniela Urlaub is highly appreciated. This work was supported by the Deutsche Forschungsgemeinschaft (SFB 688 TP B1 and A13).

Introduction Investigating cerebral ischemia requires animal models relevant to human stroke. A precise knowledge of the strengths and shortcomings of available models is mandatory for effective research in neuroprotection [1,2]. Initially described in the rat [3] and subsequently adapted in mice [4], permanent middle cerebral artery (MCA) occlusion (pMCAO) by electrocoagulation is a widely used model of focal ischemia. Invasive surgical procedures are required: temporal muscle dissection, in some cases by electrical cauterization [5,6], subtemporal craniotomy and MCA electrocoagulation. This model, by interrupting blood flow at the level of the parietal cerebral artery branch of the MCA (distal occlusion), has the advantage of producing smaller, cortical-restricted, more reproducible and better-tolerated infarcts compared to suture MCAO, which endoluminal occluder is situated in the internal artery, at the birth of the MCA (proximal occlusion) and gives rise to extensive cortico-striatal infarcts [7,8]. It may however induce traumatic brain damage. To our knowledge, sham-operated animals have only been studied using immunohistology at the subacute stage of cerebral ischemia. Early monitoring with magnetic resonance imaging (MRI) may facilitate in vivo identification of traumatic brain injury during pMCAO.

Introduction Investigating cerebral ischemia requires animal models relevant to human stroke. A precise knowledge of the strengths and shortcomings of available models is mandatory for effective research in neuroprotection [1,2]. Initially described in the rat [3] and subsequently adapted in mice [4], permanent middle cerebral artery (MCA) occlusion (pMCAO) by electrocoagulation is a widely used model of focal ischemia. Invasive surgical procedures are required: temporal muscle dissection, in some cases by electrical cauterization [5,6], subtemporal craniotomy and MCA electrocoagulation. This model, by interrupting blood flow at the level of the parietal cerebral artery branch of the MCA (distal occlusion), has the advantage of producing smaller, cortical-restricted, more reproducible and better-tolerated infarcts compared to suture MCAO, which endoluminal occluder is situated in the internal artery, at the birth of the MCA (proximal occlusion) and gives rise to extensive cortico-striatal infarcts [7,8]. It may however induce traumatic brain damage. To our knowledge, sham-operated animals have only been studied using immunohistology at the subacute stage of cerebral ischemia. Early monitoring with magnetic resonance imaging (MRI) may facilitate in vivo identification of traumatic brain injury during pMCAO. Materials and methods Animals and surgical procedure We designed an analytic appraisal of pMCAO procedure which included two methods of temporal muscle dissection: cauterization and blade incision. After intraperitoneal anesthesia with 12 mg/Kg xylazine and 90 mg/Kg ketamine, 22 male Swiss mice (28-30 g, Charles River, France) were allotted as follows:

surgical procedure We designed an analytic appraisal of pMCAO procedure which included two methods of temporal muscle dissection: cauterization and blade incision. After intraperitoneal anesthesia with 12 mg/Kg xylazine and 90 mg/Kg ketamine, 22 male Swiss mice (28-30 g, Charles River, France) were allotted as follows: • group A (n = 6): temporal muscle cauterization without craniotomy nor MCA occlusion; • group B (n = 4): temporal muscle incision without craniotomy nor MCA occlusion; • group C (n = 4): temporal muscle incision followed by craniotomy without MCA occlusion; • group D (n = 6): temporal muscle incision followed by craniotomy and MCA electrocoagulation; • group E (n = 2): temporal muscle cauterization followed by craniotomy and MCA electrocoagulation. Craniotomy and MCA electrocoagulation were performed as previously described [9]. Whatever the group be, surgery was performed in less than 15 min.

• group D (n = 6): temporal muscle incision followed by craniotomy and MCA electrocoagulation; • group E (n = 2): temporal muscle cauterization followed by craniotomy and MCA electrocoagulation. Craniotomy and MCA electrocoagulation were performed as previously described [9]. Whatever the group be, surgery was performed in less than 15 min. MRI MRI experiments were performed on a 7T/12 cm magnet (Bruker BioSpin GmbH, Ettlingen, Germany) using a 72 mm inner diameter birdcage for RF transmission and a 15 mm diameter surface coil for reception. T2-weighted images (T2WI) were acquired using a RARE sequence with TE/TR = 75/3000 ms. Diffusion weighted spin-echo images (DWI) were acquired with a TE/TR = 14/2000 ms using 2 b-values (139 and 1061 s/mm2). Apparent diffusion coefficients (ADC, in mm2/s) were calculated by fitting a monoexponential model function on a pixel-by-pixel basis. The field of view was 20 × 20 mm2 and slice thickness 1.0 mm. For T2WI, 15 slices were acquired with 256 × 256 matrix and for DWI, 7 slices with 128 × 128 matrix. Mice anesthesia was maintained during the MRI with isofluorane (1.5% in air). Body temperature was kept at 37 ± 1°C with an integrated heating system, and a pressure probe monitored mice respiration. MRI was started immediately following the end of surgical procedure and all MR acquisitions were performed between 30 min and 60 min after the start of surgery. MRI was repeated on day 1.

% in air). Body temperature was kept at 37 ± 1°C with an integrated heating system, and a pressure probe monitored mice respiration. MRI was started immediately following the end of surgical procedure and all MR acquisitions were performed between 30 min and 60 min after the start of surgery. MRI was repeated on day 1. Results Temporal muscle cauterization in the absence of MCA electrocoagulation (group A) consistently produced an extensive lesion across the frontoparietal cortex (n = 6/6). These lesions appeared hyperintense on T2WI within the first hour after surgery, with an early mass effect and reduced ADC (Figure 1A). Temporal muscle incision and craniotomy in the absence of MCA electrocoagulation (group B and C, respectively) caused no visible brain injury on day 0 or day 1 MRI (n = 4/4 in groups B and C, Figure 1B and 1C). MCA electrocoagulation (group D) induced an ischemic lesion in the frontoparietal cortex with reduced ADC on day 0 (n = 6/6). These lesions were not visible on T2WI within 1 h post-surgery (Figure 1D). No bleeding was noted during the surgery, but MCA electrocoagulation occasionally produced a small superficial traumatic lesion (n = 2/6 in group D, figure 1D). When pMCAO was performed using temporal muscle cauterization (group E), the lesion appeared as a superimposition of the almond-shaped cauterization lesion and of the MCA-territory bounded ischemic lesion (n = 2/2, Figure 2). Infarcts from group A, D and E were clearly delineated on day 1 T2WI.

6 in group D, figure 1D). When pMCAO was performed using temporal muscle cauterization (group E), the lesion appeared as a superimposition of the almond-shaped cauterization lesion and of the MCA-territory bounded ischemic lesion (n = 2/2, Figure 2). Infarcts from group A, D and E were clearly delineated on day 1 T2WI. Figure 1 MRI within one hour of surgery: T2WI (upper row) and ADC map (lower row). A: temporal muscle cauterization alone (group A). Note the lesion with early T2WI hyperintensity and reduced ADC (arrows). B: temporal muscle incision alone (group B). C: temporal muscle incision and craniotomy (group C). D: temporal muscle incision, craniotomy and MCA electrocoagulation. Note the ischemic lesion with low ADC (arrow) and normal T2WI with a limited superficial traumatic lesion (arrow). Figure 2 Impact of temporal muscle cauterization on pMCAO with T2WI. A: temporal muscle cauterization alone (group A) showing day 0 traumatic lesion (interrupted lines). B: temporal muscle cauterization, craniotomy and MCA electrocoagulation (group E). Color-coded magnification shows the boundaries of the ischemic lesion (dotted line) and traumatic lesion (interrupted lines) 6 h after pMCAO. C: same mice at day 1 showing the limits of the final lesion (dotted line).

interrupted lines). B: temporal muscle cauterization, craniotomy and MCA electrocoagulation (group E). Color-coded magnification shows the boundaries of the ischemic lesion (dotted line) and traumatic lesion (interrupted lines) 6 h after pMCAO. C: same mice at day 1 showing the limits of the final lesion (dotted line). Discussion Our results showed that temporal muscle cauterization in the absence of ischemic stimulus produced a hyperintense area on T2WI highly suggestive of traumatic injury. Indeed, T2 increase is not expected in ischemic lesions within the first hours of occlusion in adults, although early T2 increase has been reported in neonates after MCA electrocoagulation [10] or hypoxia-ischaemia [11]. Early mass effect is uncommon after ischaemia in the adult rat [12] or mouse [13]. Acute ADC decrease is not specific of ischemic insult and has been described in both experimental and human traumatic brain injury [14,15]. Signal suggestive of intra- or extra-cranial surgery-related bleeding (T2 hypointensity) would have been ascertained using T2*WI, but were not noted. No traumatic injury was observed after temporal muscle incision. Accordingly, incision should be preferred over cauterization for temporal muscle dissection.

ury [14,15]. Signal suggestive of intra- or extra-cranial surgery-related bleeding (T2 hypointensity) would have been ascertained using T2*WI, but were not noted. No traumatic injury was observed after temporal muscle incision. Accordingly, incision should be preferred over cauterization for temporal muscle dissection. The MRI appearance of lesions resulting from both temporal muscle cauterization (traumatic damage) and occlusion (ischemic damage) may mimic an infarction, especially if imaging is done at later time points. Histopathological analyses are usually performed after a delay of 24-72 h, when traumatic and ischemic damage may not be discernable, while early histological examination of intracerebral coagulation necrosis would be required to ascertain the thermal origin of traumatic damage induced by muscle cauterization. In the last decade, high resolution MRI has become a valuable tool for monitoring tissue damage in rodent models of cerebral ischemia. Early MRI monitoring may help to identify non-specific brain injury that could hamper neuroprotective drugs assessment. Competing interests The authors declare that they have no competing interests.

In the last decade, high resolution MRI has become a valuable tool for monitoring tissue damage in rodent models of cerebral ischemia. Early MRI monitoring may help to identify non-specific brain injury that could hamper neuroprotective drugs assessment. Competing interests The authors declare that they have no competing interests. Authors' contributions FC carried out the MRI experiments, participated in the design of the study and helped to draft the manuscript. SM carried out surgery. MW designed and optimized the MRI acquisition protocol, and helped to draft the manuscript. JCB participated in the MRI experiments. YB and NN conceived the study, and participated in its design and coordination. THC conceived the study, participated in its design, performed image analysis, and drafted the manuscript. All authors read and approved the final manuscript. Acknowledgements We are grateful to Jean-Baptiste Langlois, from the Small Animal Multimodal Imaging Facility, Animage (CERMEP, Lyon, France) for his help in MRI experiments.

Review Surviving a sublethal noxious insult may result in a more powerful state against a following lethal insult, referring to Nietszche; "What doesn't kill you, makes you stronger." This phenomenon named as preconditioning (PC) and tolerance has been shown to exist in many organs, most extensively in the heart. The first in vivo evidence of preconditioning and tolerance in brain was provided in 1960's [1,2], but almost three decades passed without any interest from researchers on this unique phenomenon, until Kitagawa et al. opened the research era of cerebral ischemic tolerance (IT) [3].

rgans, most extensively in the heart. The first in vivo evidence of preconditioning and tolerance in brain was provided in 1960's [1,2], but almost three decades passed without any interest from researchers on this unique phenomenon, until Kitagawa et al. opened the research era of cerebral ischemic tolerance (IT) [3]. The ability to withstand, respond to, and to cope with ongoing stress is a fundamental property of all living organisms [4]. The fate of the brain tissue after focal cerebral ischemia is determined by the degree and duration of ischemia, and even without preconditioning, resident brain cells naturally respond to brain ischemia by mobilizing a host of defences and counter responses to mitigate cell injury and death [5]. If the subthreshold noxious stimulus is too mild or negligibly mild, it may not induce any response, whereas if it is sufficient enough, it may serve as a PC trigger, or if it is too severe, over the threshold, may permanently injure tissues. The hallmark of PC stimulus is not being injurious. In the scenario of IT, PC stimulus primes the brain for subsequent injurious ischemic injury. Danger signal evoked in the brain by the stressing preconditioning stimulus induces complex endogenous protective mechanisms resulting to a latent protective phenotype. When the lethal ischemic insult is applied onto this latent protective phenotype, a separate set of responses are triggered that constitute ischemia-tolerant phenotype, which strikingly differs from the unprimed or unpreconditioned brain's phenotype (Figure 1). Therefore, the outcome of the brain cells is shifted by PC from death to survival.

insult is applied onto this latent protective phenotype, a separate set of responses are triggered that constitute ischemia-tolerant phenotype, which strikingly differs from the unprimed or unpreconditioned brain's phenotype (Figure 1). Therefore, the outcome of the brain cells is shifted by PC from death to survival. Figure 1 Outcomes in the different settings with middle cerebral artery occlusion (MCAO). In the absence of ischemic preconditioning (IPC), MCAO induces a large infarction (ischemia-sensitive phenotype). IPC results early ischemic tolerance (IT) in minutes; if MCAO is applied during this phase, mainly cortical areas are spared (early ischemia-tolerant phenotype). In hours, the brain regresses to its naïve state. Delayed phase of IT occurs in days, when the latent cerebroprotective phenotype is complete, the brain is again ischemia-tolerant. During the last years, the mechanisms underlying cerebral IT were intensely studied, and although incomplete, a vast amount of knowledge has been accumulated. The salient features of cerebral IT are presented in Table 1. There are two temporally distinct windows of protection from ischemia afforded by PC. Early protection, i.e., early IT, has been observed in relatively fewer studies than those exposing late IT in the brain. Exploring the functional relevance of these findings has proved difficult, however. Table 1 Main futures of cerebral ischemic tolerance

During the last years, the mechanisms underlying cerebral IT were intensely studied, and although incomplete, a vast amount of knowledge has been accumulated. The salient features of cerebral IT are presented in Table 1. There are two temporally distinct windows of protection from ischemia afforded by PC. Early protection, i.e., early IT, has been observed in relatively fewer studies than those exposing late IT in the brain. Exploring the functional relevance of these findings has proved difficult, however. Table 1 Main futures of cerebral ischemic tolerance General Preconditioning specific • Robust cerebroprotection  • Two phased: early and delayed • The interval between preconditioning and ischemia determines the fate  • Early tolerance starts in minutes, delayed tolerance not usually before 24 h • Ischemic tolerance is transient  • After early phase, but before delayed phase no tolerance is achieved • Ischemic tolerance can be induced by a variety of stimuli  • Early phase is short-lasting, delayed phase longer, up to 1 week • Transient ischemic attacks confer ischemic tolerance in humans  • Preconditioning preserves cortical/penumbral tissue in focal ischemia models In this review article, we first attempted to clarify the IT nomenclature. Various triggers induce cerebral IT; these are mentioned in a separate section discussing the models for IT. This variety among preconditioning triggers indicates that the downstream signaling pathways converge on some common fundamental mechanisms [5], major mechanisms are discussed briefly. A number of tools serve for examination of the efficacy of PC, chief methods are exampled. Lastly, we addressed the challenging issues of IT to encourage further research.

riggers indicates that the downstream signaling pathways converge on some common fundamental mechanisms [5], major mechanisms are discussed briefly. A number of tools serve for examination of the efficacy of PC, chief methods are exampled. Lastly, we addressed the challenging issues of IT to encourage further research. I. Nomenclature The nomenclature used in the studies addressing the IT phenomenon is not entirely consistent. In this article, in order to keep with consistency while defining the methodology of the IT experiments and to provide ease for reading, the following terms will be used according to the definitions given: PC The stimulus or the method applied in an experiment that triggers IT in the brain. Ischemic preconditioning (IPC) Method of PC by inducing either global or focal transient cerebral ischemia. When the PC stimulus is different than ischemia, PC will be named according to the nature of the trigger (e.g., hypoxic-PC, anesthetic-PC, chemical-PC). In the literature, sublethal or priming ischemia have been used as alternative terms to IPC. For subsequent ischemic insult given after PC, the term test ischemia is often used [6,7], among others (final, lethal, reference, or full ischemia). We will prefer to name the subsequent ischemic event as final ischemia.

-PC). In the literature, sublethal or priming ischemia have been used as alternative terms to IPC. For subsequent ischemic insult given after PC, the term test ischemia is often used [6,7], among others (final, lethal, reference, or full ischemia). We will prefer to name the subsequent ischemic event as final ischemia. IT Briefly, the protection from final ischemia provided by prior PC refers to as IT. Depending on the media used to study the phenomenon of IT, IT may refer to the cell, tissue, or organ's post-ischemic state wherein, due to previous PC exposure, the response to ischemia is different from one observed without previous PC. In this article therefore, PC and IT define two different (but related) entities and are not used interchangeably. Ischemia-tolerant phenotype (Figure 1) It is the consequence of both pre- and post-ischemic protective responses induced by PC [5]; in other words, it is the resulting phenotype from both PC and final ischemia. Latent cerebroprotective phenotype It determines the status of the cell, tissue, or organ exposed to PC that experiences changes triggered by PC, and it occurs before the application of final ischemia [5]. Hence, the latent cerebroprotective phenotype differs from the ischemia-tolerant phenotype by the lack of exposure to final ischemia (Figure 1).

type It determines the status of the cell, tissue, or organ exposed to PC that experiences changes triggered by PC, and it occurs before the application of final ischemia [5]. Hence, the latent cerebroprotective phenotype differs from the ischemia-tolerant phenotype by the lack of exposure to final ischemia (Figure 1). Cerebroprotection or protection With the better understanding of the concept of neurovascular unit (i.e., the contribution of glial and vascular endothelial cells and their interactions with neurons in physiological and pathological conditions), researchers' attention shifted from neurons towards cerebrum. Hence, instead of "neuroprotection", we prefer to use "cerebroprotection", which covers not only neurons but all the cerebral cell populations experiencing IT. To interpret IT afforded on single cell type (hippocampal CA1 neurons in global ischemic models or type of cell slice used in vitro study), the protective effect provided by PC will be discussed as "protection".

se "cerebroprotection", which covers not only neurons but all the cerebral cell populations experiencing IT. To interpret IT afforded on single cell type (hippocampal CA1 neurons in global ischemic models or type of cell slice used in vitro study), the protective effect provided by PC will be discussed as "protection". II. Two phases of IT Preconditioning induces two phases of IT with different temporal profiles and, to some extent, with different mechanisms of protection: early and delayed IT (Figure 1); the latter plays the major role in the brain. Early IT is a short-lasting protection induced within minutes of exposure to PC and wanes within hours. In this process, rapid changes in activity and posttranslational modifications of existing proteins are involved, whereas delayed IT requires gene induction and de novo protein synthesis, that represent a long-term response through genetic reprogramming [4]. If the final ischemia is induced during the unprotected window, which exists between early and late IT (usually 30 min-1 hour after PC, lasting up to 24 hours), no tolerance occurs (Figure 1). In the literature, early IT has been termed as the first window of protection [8], rapid IT [5], immediate IT [9], short-term protection [8], classical IT [10], or acute IT [4]. We will prefer to use early IT [10]. Alternative terms for delayed IT are: second window of protection (a term widely used in heart IT studies), classical IT [5], and late IT [11].

ermed as the first window of protection [8], rapid IT [5], immediate IT [9], short-term protection [8], classical IT [10], or acute IT [4]. We will prefer to use early IT [10]. Alternative terms for delayed IT are: second window of protection (a term widely used in heart IT studies), classical IT [5], and late IT [11]. III. Models for IT Study setups for investigating potential phenotypes induced by PC are exampled in Figure 2. A summary of the available rodent models of IT is included in the following link as Additional File 1. Figure 2 Exemplary protocol of a series of ischemic tolerance (IT) experiments. Preconditioning ischemia (IPC) lasts 5 min, final ischemia 10 min. Cross indicates the end of the study where the brains are collected for ex vivo evaluations. Set I experiments evaluate early IT: final ischemia is applied 30 min after IPC (upper row), early ischemia-tolerant phenotype is tested by collecting the brains 25 min after PC, and in the control experiment, final ischemia is induced without prior IPC (lower row). In Set II experiments, delayed IT (upper row, final ischemia induction 24 h after the PC) and delayed ischemia-tolerant phenotype (middle row, brains are collected 24 h after PC) are investigated, along with a control experiment (lower row). Set III experiments address long-lasting effects of delayed-IT, with a follow-up lasting several weeks after final ischemia.

final ischemia induction 24 h after the PC) and delayed ischemia-tolerant phenotype (middle row, brains are collected 24 h after PC) are investigated, along with a control experiment (lower row). Set III experiments address long-lasting effects of delayed-IT, with a follow-up lasting several weeks after final ischemia. Global-Global Animal models of global cerebral ischemia are designed to mimic cardiac arrest in humans. Global-global IT models include different durations of transient global ischemia as PC and final insult. Four-vessel occlusion Rat forebrain ischemia experiments necessitate the occlusion of both posterior and anterior blood circulation of the brain. Four-vessel occlusion was originally described by Pulsinelli et al. [12] as a two-stage procedure wherein first both vertebral arteries are permanently closed and the following day, both carotid arteries are occluded. The model has been modified later by the same authors [13] and others [14]. In rats, delayed IT was provided by 3 [15] and 5 min [16] of four-vessel occlusion to 6 to 20 min of final global ischemia [15-17]. Appropriate durations of PC and the interval between PC and final ischemia in this scenario have been studied by monitoring postischemic Hsp72 protein expression as a marker of IT [18]. Neither 3, nor 8 min of IPC induced sufficient synthesis of Hsp72. Once PC was fixed to 4 to 5 min, the minimum 2 days of interval was required between PC and final ischemia.

rval between PC and final ischemia in this scenario have been studied by monitoring postischemic Hsp72 protein expression as a marker of IT [18]. Neither 3, nor 8 min of IPC induced sufficient synthesis of Hsp72. Once PC was fixed to 4 to 5 min, the minimum 2 days of interval was required between PC and final ischemia. Two-vessel occlusion and hypotension This model, originally described by Smith et al. [19] in the rat, includes bilateral carotid artery occlusion and systemic hypotension induced by withdrawal of arterial blood. Compared to four-vessel occlusion, it is less invasive and more reproducible, as the depth of ischemia depends on hypotension, rather than on surgical attenuation of collateral perfusion [20]. PC by 2 to 3 min of two-vessel occlusion induces delayed IT in rats to 5 to 10 min of final global ischemia [21-23]. Early IT was also achieved in this model [24].

t is less invasive and more reproducible, as the depth of ischemia depends on hypotension, rather than on surgical attenuation of collateral perfusion [20]. PC by 2 to 3 min of two-vessel occlusion induces delayed IT in rats to 5 to 10 min of final global ischemia [21-23]. Early IT was also achieved in this model [24]. Two-vessel occlusion in gerbils The gerbil mostly lacks a functioning circle of Willis [25], a short period of bilateral occlusion of carotid arteries (3-5 min) therefore leads to severe damage in CA1 pyramidal neurons [26,27]. First introduced by Kitagawa et al. [3], PC by single 2-min bilateral carotid occlusion (or two times), 1 to 7 days before final ischemia in the gerbil brain has been a well-standardized method to study IT [28-30]. A disadvantage of the gerbil two-vessel occlusion model is that the severity of forebrain ischemia is highly influenced by the anatomical variations, which are not seldom [31]. Here, we should note that thresholds for severity (i.e., duration of the bilateral carotid occlusion), differentiating the outcome as either PC or final insult, are in a narrow scale. To overcome this issue and ensure a better control over the ischemia severity, Abe and colleagues [7] provided a useful modification of the model. By monitoring depolarizations, they largely eliminated the variability of the ischemia and IT. This approach later was introduced in a rat global-global IT model using four-vessel occlusion [32].

and ensure a better control over the ischemia severity, Abe and colleagues [7] provided a useful modification of the model. By monitoring depolarizations, they largely eliminated the variability of the ischemia and IT. This approach later was introduced in a rat global-global IT model using four-vessel occlusion [32]. Two-vessel occlusion in mice This method, borrowed from its equivalent in gerbils, may induce reproducible striatal injury in mice [33]. For a delayed IT paradigm, Wu et al. applied 6 min of two-vessel occlusion as PC and 18 min of bilateral carotid artery occlusion as final ischemia in C57BL/6 mice [34]. As this strain is a common subject of transgenic technology, the model proved useful for investigating the molecular mechanisms of IT in gene-modified mice. In a such scenario, a much longer two-vessel occlusion period (20 min) has induced delayed IT [35].

arotid artery occlusion as final ischemia in C57BL/6 mice [34]. As this strain is a common subject of transgenic technology, the model proved useful for investigating the molecular mechanisms of IT in gene-modified mice. In a such scenario, a much longer two-vessel occlusion period (20 min) has induced delayed IT [35]. Focal-Focal Transient focal-permanent focal Transient occlusion of the middle cerebral artery (MCA) by intraluminal insertion of a nylon monofilament, which was originally described by Koizumi et al. [36] and modified by others [37], is the most common model to induce focal cerebral ischemia in rats [38-41] and also available in mice [42-45]. This method was introduced first time in a rat IT experiment, applying 10 min of transient MCA occlusion (tMCAO) as the PC stimulus and permanent MCAO as the final ischemia [46]. Authors evaluated IT phenomenon with several reperfusion periods between IPC and final ischemia and showed that ischemic lesions involving both cortex and basal ganglia could be reduced when final ischemia was applied 1, 2, and 7 days after PC, but not 2, 6, and 12 hours or 14 and 21 days after PC. This model was applied successfully by others to obtain delayed IT [47,48]. Repeated brief transient ischemia regimen was also proved as a preconditioning paradigm inducing early IT in mice subjected to permanent focal ischemia [49,50].

d 1, 2, and 7 days after PC, but not 2, 6, and 12 hours or 14 and 21 days after PC. This model was applied successfully by others to obtain delayed IT [47,48]. Repeated brief transient ischemia regimen was also proved as a preconditioning paradigm inducing early IT in mice subjected to permanent focal ischemia [49,50]. Transient focal-transient focal One [51,52] or 3 times of 10 min transient focal cerebral ischemia protects from subsequent 120 min of tMCAO in rats [53-55]. Shorter durations (2 and 3 min) of tMCAO were severe enough to induce delayed IT, but did not provide early IT to transient ischemia [56,57]. Transient focal-focal IT paradigm induced IT also in mice and spontaneously hypertensive rats [58,59]. A recent mouse model of delayed-IT involves 2 periods of 5-min tMCAO as the PC method, against 90-min tMCAO applied in 3 days, but not in 2 or 4 days [6]. Global-Focal Brief global ischemia can protect from both subsequent transient and permanent focal ischemia [60,61]. Focal-Global Brief unilateral occlusion of the MCA induced significant protection from global ischemia in both gerbils [62] and rats [63]. Interestingly, transient (20 min) occlusion of the distal MCA protected only ipsilateral parietal cortex of the rat from global ischemia (10 min) [64].

Global-Focal Brief global ischemia can protect from both subsequent transient and permanent focal ischemia [60,61]. Focal-Global Brief unilateral occlusion of the MCA induced significant protection from global ischemia in both gerbils [62] and rats [63]. Interestingly, transient (20 min) occlusion of the distal MCA protected only ipsilateral parietal cortex of the rat from global ischemia (10 min) [64]. Cross-Tolerance Cross-tolerance is tolerance to ischemia provided by miscellaneous noxious stimuli, rather than ischemia. These differ greatly in nature, nevertheless, because of a common reason (most likely by inducing genetic reprogramming), all furnish IT. Examples of cross-tolerance in the scenario of transient focal cerebral ischemia are provided in Table 2. Table 2 Amount of histological protection afforded by preconditioning in selected studies of focal cerebral ischemia

Cross-Tolerance Cross-tolerance is tolerance to ischemia provided by miscellaneous noxious stimuli, rather than ischemia. These differ greatly in nature, nevertheless, because of a common reason (most likely by inducing genetic reprogramming), all furnish IT. Examples of cross-tolerance in the scenario of transient focal cerebral ischemia are provided in Table 2. Table 2 Amount of histological protection afforded by preconditioning in selected studies of focal cerebral ischemia Protection* (%) Follow-up** Ref. Focal-focal, in rats 15 min MCAO - 72 h later pMCAO 41 24 h [47] 10 min MCAO - 72 h later 60 min tMCAO 44 7 d [59] 3 min tMCAO - 72 h later 60 min tMCAO 35 24 h [57] Focal-focal, in mice 3 × 5 min tMCAO - 30 min later pMCAO 23 24 h [50] 3 × 5 min tMCAO - 30 min later 60 min tMCAO 32 24 h [49] 15 min MCAO - 72 h later 45 min tMCAO 70 24 h [58] LPS LPS 0.5 mg/kg - 72 h later 60 min tMCAO in rats 35 24 h [98] Hypoxia 11% oxygen for 2 h - 48 h later 90 min tMCAO in mice 46-64 24 h [66] Anestesia Isoflurane 1.4% for 3 h - 0, 12, and 24 h later pMCAO in rats 31-35 4 d [113] Halothane 1.2% for 3 h - 24 h later pMCAO in rats 35 4 d [113] Spreading depression KCl application - 4 days later 120 min tMCAO in rats 43 4 d [191] Hyperbaric oxygen 100% oxygen for 1 h, 5 days-24 h later pMCAO in mice 27 24 h [69] Ref, references; *Reduction in the ischemic damage due to preconditioning; **Time-point of the histopathological analysis after final ischemia

[113] Spreading depression KCl application - 4 days later 120 min tMCAO in rats 43 4 d [191] Hyperbaric oxygen 100% oxygen for 1 h, 5 days-24 h later pMCAO in mice 27 24 h [69] Ref, references; *Reduction in the ischemic damage due to preconditioning; **Time-point of the histopathological analysis after final ischemia Hypoxia Exposure of neonatal rats to 8% oxygen for 3 hours provides cerebroprotection from a combined hypoxia/ischemia model [65] and also from both transient and permanent focal cerebral ischemia [66,67]. Varying hypoxia durations (1, 3, or 6 hours) result in similar extent of protection, but when the interval between hypoxia and final ischemia exceeds 72 hours, IT abolishes [67]. Hyperbaric oxygen Hyperbaric oxygen was found protective from subsequent global ischemia in gerbils [68] and from permanent focal ischemia in SV129 mice [69], whereas it did not induce IT to transient focal ischemia in these mice [69]. Rats were protected from transient ischemia by hyperbaric oxygen PC, but they were not protected from permanent ischemia [70]. Repeated hyperbaric oxygen application seems to induce IT to global ischemia in the rat brain for a shorter period than 72 h [71]. Hyperthermia In rodent experiments, indirect brain temperature can be measured with a probe placed under the temporal muscle and can be maintained at a desired level by heaters allowing feedback adjustments. Chopp et al. first time observed the PC effect of hyperthermia in rats subjected to global ischemia [72]. Hyperthermia was protected as well neonatal rats from hypoxia/ischemia [73].

measured with a probe placed under the temporal muscle and can be maintained at a desired level by heaters allowing feedback adjustments. Chopp et al. first time observed the PC effect of hyperthermia in rats subjected to global ischemia [72]. Hyperthermia was protected as well neonatal rats from hypoxia/ischemia [73]. Hypothermia The hypothermic-PC has been described in a rat model of focal transient ischemia [74] and later was studied systematically in order to define the optimal depth, duration, and the method of application (global versus focal hypothermia) [75]. The extent of protection was dependent on the depth and duration of the hypothermia, focal cooling being as effective as systemic cooling. Although the deeper the hypothermia, the bigger the IT response was, mild to moderate levels of hypothermia, which are safe in humans [76], were efficient as well. This may encourage clinicians to test hypothermia as a preconditioning strategy, for instance before vascular surgical interventions with high risk of ischemic events.

e deeper the hypothermia, the bigger the IT response was, mild to moderate levels of hypothermia, which are safe in humans [76], were efficient as well. This may encourage clinicians to test hypothermia as a preconditioning strategy, for instance before vascular surgical interventions with high risk of ischemic events. Spreading depression Leão's spreading depression is a generalized and stereotyped response of the cerebral cortex to a variety of noxious stimuli and is characterized by a slowly moving, transient, and reversible depression of cortical electrical activity that spreads like ripples in a pond; these waves, from the site of onset, spread usually to the whole cortex of the ipsilateral brain hemisphere with a speed of 2 to 5 mm per minute [77]. Topical application of high concentration of potassium chloride onto the cortex induces spreading depression that repetitively extends from the sites of increased extracellular potassium concentration with a frequency of approximately 7/100 min [78]. This method has been an effective PC trigger in both global [79] and focal ischemia models in rats [80-82]. IT induced by spreading depression seems to develop in a delayed manner (in 3-6 days) [83,84] and was shown to persist up to 15 days [85].

potassium concentration with a frequency of approximately 7/100 min [78]. This method has been an effective PC trigger in both global [79] and focal ischemia models in rats [80-82]. IT induced by spreading depression seems to develop in a delayed manner (in 3-6 days) [83,84] and was shown to persist up to 15 days [85]. Remote IPC Limb ischemia by bilateral femoral artery occlusion protects rat from either global ischemia [86] or transient focal ischemia [87]. It can be applied as well repeatedly (5-10 min for 3 times, with 10 minutes intervals in between). This PC approach was successfully tested in humans to induce IT in the heart [88]. Mesenteric artery occlusion for 15 min was protective against bilateral carotid occlusion in mice [89]. 3-nitropropionic acid (3-NPA) This is the most extensively studied chemical PC agent, which inhibits oxidative phosphorylation. Intraperitoneal administration of 3-NPA, 72 hours before transient focal ischemia, is a well-established PC trigger for rats [90-93]. Regarding the efficacy of 3-NPA as a PC trigger, some contradictory results came from gerbil models of global ischemia [94-96], but these may be related to the doses used [97]. Lipopolysaccharide (LPS) LPS is a cell-wall component of gram-negative bacteria. A small dose provides IT in the brain. This has been proven in a number of experiments including both transient [98,99] and permanent [100,101] focal ischemia models in rats, as well as in a mouse model of transient focal ischemia [102]. With higher doses no PC effect occurs [98].

l component of gram-negative bacteria. A small dose provides IT in the brain. This has been proven in a number of experiments including both transient [98,99] and permanent [100,101] focal ischemia models in rats, as well as in a mouse model of transient focal ischemia [102]. With higher doses no PC effect occurs [98]. Anesthetic-PC Potential protective effects of anesthetics from an ischemic insult have been known for long time [2] and were well-studied in experimental stroke models as a cerebroprotective strategy (for reviews see [103-105]). Among anesthetics, isoflurane is the most commonly used volatile anesthetic in IT experiments. Different concentrations (0.5-4%) and varying durations (15 min-3 hours) of isoflurane inhalation have been efficient to induce both early [106,107] and delayed IT in vivo [108,109] and in vitro models [110,111]. Among other anesthetics, xenon [112], halothane [113], and sevoflurane [114] may also induce IT in animal models. Anestetic-PC has been proven a promising PC method for heart in humans [115]. Pharmacological PC Several clinically available drugs, including estrogen [116], erythromycin [117,118], and erythropoietin [119,120], are capable of inducing IT in animal models. Acetylsalicylic acid [121] and kanamycin [122] were effective as PC agents in vitro.

Anesthetic-PC Potential protective effects of anesthetics from an ischemic insult have been known for long time [2] and were well-studied in experimental stroke models as a cerebroprotective strategy (for reviews see [103-105]). Among anesthetics, isoflurane is the most commonly used volatile anesthetic in IT experiments. Different concentrations (0.5-4%) and varying durations (15 min-3 hours) of isoflurane inhalation have been efficient to induce both early [106,107] and delayed IT in vivo [108,109] and in vitro models [110,111]. Among other anesthetics, xenon [112], halothane [113], and sevoflurane [114] may also induce IT in animal models. Anestetic-PC has been proven a promising PC method for heart in humans [115]. Pharmacological PC Several clinically available drugs, including estrogen [116], erythromycin [117,118], and erythropoietin [119,120], are capable of inducing IT in animal models. Acetylsalicylic acid [121] and kanamycin [122] were effective as PC agents in vitro. Other models for IT Repeated electroacupuncture [123,124], electrical stimulation of cerebellar fastigial nucleus [125], and dietary restriction [126] protected rats from subsequent transient focal ischemia. In global ischemia rat models, repetitive transcranial magnetic stimulation [127], electroconvulsive shock [128], kainite-induced epileptic seizures [129], and sleep deprivation [130] all have served as PC stimuli.

ial nucleus [125], and dietary restriction [126] protected rats from subsequent transient focal ischemia. In global ischemia rat models, repetitive transcranial magnetic stimulation [127], electroconvulsive shock [128], kainite-induced epileptic seizures [129], and sleep deprivation [130] all have served as PC stimuli. In vitro models Neuronal cell culture systems provide an ideal microenvironment to study PC, because they lack a vascular compartment and the environment is easily controlled for confounding factors (e.g., see the fascinating work by Gonzalez-Zulueta et al. [131]). In vitro modeling for ischemia consists of oxygen and glucose deprivation (OGD) in the culture medium, and perhaps the most widely used method is the one described by Goldberg and Choi [132,133]. This model includes the transfer of neocortical cell cultures for several hours to an anaerobic chamber containing a gas mixture of 5% CO2, 10%H2, and 85%N2 (oxygen deprivation), followed by application of a deoxygenated glucose-free medium (glucose deprivation). Organotypic hippocampal slice cultures offer an attractive alternative method, because many aspects of in vivo ischemia, such as delayed death of CA1 neurons and selective vulnerability in response to OGD, can be addressed [134]. Hassen et al. has introduced a new model of IT by isolating hippocampal slices from young rats, to abolish age-dependent resistance to ischemic injury [135]. Also mixed neocortical cultures are available to study IT in vitro [136].

h of CA1 neurons and selective vulnerability in response to OGD, can be addressed [134]. Hassen et al. has introduced a new model of IT by isolating hippocampal slices from young rats, to abolish age-dependent resistance to ischemic injury [135]. Also mixed neocortical cultures are available to study IT in vitro [136]. IV. Methods for detecting IT In most of the IT studies, the ischemia-tolerant phenotype is addressed with assessments performed after the final ischemia; however, to expose the molecular substrates of latent cerebroprotective phenotype, the tissues should be collected after PC (Figure 2). Studies, which used the latter approach, have been recently reviewed [137]. To increase the relevance to the human condition, IT models should include both histological and functional evaluations. However, these imply more challenges for IT researchers [138], because not always a correlation between these two outcome parameters is present [30].

tter approach, have been recently reviewed [137]. To increase the relevance to the human condition, IT models should include both histological and functional evaluations. However, these imply more challenges for IT researchers [138], because not always a correlation between these two outcome parameters is present [30]. Histological techniques Determining the extent of injury after focal ischemia is relatively simpler than after global ischemia. For this purpose, traditional histological staining techniques, such as hematoxylin-eosine and 2% solution of 2,3,5-triphenyltetrazolium chloride, are often used. Digital camera-based image analysis systems enable lesion area and volume calculations. Ischemic lesion volume is calculated preferably with the correction of edema effect [38,39,139,140]. In IT experiments, reduction in lesion volume due to PC (lesion size in the näive brain -- lesion size in the preconditioned brain) can be calculated as a percent ratio to the lesion size in the näive brain (Table 2). In global ischemia models, ischemic damage is assessed in hippocampal sections stained with toluidine blue [141], cresyl violet [34,94], or thionin [142] by counting CA1 neurons, which are highly susceptible to global ischemia and easy to quantify due to their laminar distribution and large size [138]. Protection due to PC can be reported as the percentage of preserved healthy hippocampal CA1 neurons or number of viable CA1 neurons [7,30,129]. In vitro models of IT use cellular injury assessments, such as lactate dehydrogenase assay [111,143].

d easy to quantify due to their laminar distribution and large size [138]. Protection due to PC can be reported as the percentage of preserved healthy hippocampal CA1 neurons or number of viable CA1 neurons [7,30,129]. In vitro models of IT use cellular injury assessments, such as lactate dehydrogenase assay [111,143]. Functional assessment Gross measures of sensorimotor abilities are available for rodents [37,144], and were introduced in IT experiments [46,47,97,107]. However, in these species, gross sensorimotor deficits tend to recover rapidly. That is, more complex tests are needed, especially if outcome is assessed in long-term. A number of somatosensory tests (e.g. limb placing, beam walking, grid walking, rotarod) are available to apply in focal ischemia rat models [145]. In global ischemia models, tests of learning ability, and working and reference memory are particularly useful [138,146].

d, especially if outcome is assessed in long-term. A number of somatosensory tests (e.g. limb placing, beam walking, grid walking, rotarod) are available to apply in focal ischemia rat models [145]. In global ischemia models, tests of learning ability, and working and reference memory are particularly useful [138,146]. Lesion evaluation by magnetic resonance imaging (MRI) MRI technology allows for temporal and spatial monitoring of ischemic lesion and enables to conduct longitudinal studies [147-151]. Besides requiring anesthesia, MRI is risk-free for experimental animals. First MRI-based lesion evaluation in an experimental IT study was reported by Mullins et al. [152]. In a delayed IT model (focal--focal ischemia), rats were imaged 24 and 72 hours after final ischemia. Interestingly, lesion reduction due to PC was greater at 72 hours (70%) compared to that at 24 hours (53%). Authors concluded that 24 hours post-ischemia, which is a common time-point for lesion evaluation in experimental stroke studies [153-155], may not be the best time-point for experimental IT studies. Furuya et al., imaged rats serially (at 6 and 24 hours and 4, 7, and 14 days), following a delayed IT paradigm (LPC-PC--focal ischemia) [156]. They evaluated whether decreased lesion size due to PC would increase in long-term. No delayed lesion progression was found.

be the best time-point for experimental IT studies. Furuya et al., imaged rats serially (at 6 and 24 hours and 4, 7, and 14 days), following a delayed IT paradigm (LPC-PC--focal ischemia) [156]. They evaluated whether decreased lesion size due to PC would increase in long-term. No delayed lesion progression was found. Means for depicting the mechanisms of IT Immmunohistochemistry is widely applied in IT research and serves to evaluate tissue alterations by means of antigen-antibody interactions. In situ hybridization and Western blotting techniques are applied to examine the effect of PC on the investigated protein's mRNA expression and abundance [141]. DNA microarray technology, which allows quantification and differential expression of thousands of genes simultaneously, has been used to investigate global changes occurring between ischemia-tolerant and näive brains (see below "genomic reprogramming"). Real-time PCR can be used for confirmation of the selected genes, which found upregulated by microarray analysis [157]; proteomics may provide supplemental insights [158]. Alterations in neurotransmitter receptor density can be evaluated by quantitative in vitro receptor autoradiography [159,160]. Autoradiographical methods may show changes in the global protein status [29,161].

nes, which found upregulated by microarray analysis [157]; proteomics may provide supplemental insights [158]. Alterations in neurotransmitter receptor density can be evaluated by quantitative in vitro receptor autoradiography [159,160]. Autoradiographical methods may show changes in the global protein status [29,161]. Showing attenuation or abolishment of IT by pharmacological inhibition of a molecule before or after PC stimulus proves a robust approach. With this approach, necessary or mandatory components of IT can be explored [162,163]. Complementary information may come from genetically modified animals by proving abolishment of IT in mutants lacking a functional molecule or protein and showing reestablishment of IT in rescue experiments [164]. Maintenance of IT, despite pharmacological inhibition of a molecule of interest or despite the lack of this molecule in the mutant mouse, may rule out the hypothetic causative role for the investigated molecule in the acquisition of IT [35,84,163]. However, it should be noted that, the main effectors of IT can be model- or trigger-specific that, for instance, a specific molecule proven mandatory for hypoxia-induced IT in the rat brain [165] may not necessarily be required in OGD-induced IT in vitro [166].

nvestigated molecule in the acquisition of IT [35,84,163]. However, it should be noted that, the main effectors of IT can be model- or trigger-specific that, for instance, a specific molecule proven mandatory for hypoxia-induced IT in the rat brain [165] may not necessarily be required in OGD-induced IT in vitro [166]. V. Mechanisms of IT IT is achieved by the attenuation of broad categories of injury-inducing mechanisms, including excitotoxicity, ion and pH imbalance, oxidative and nitrosative stress, metabolic dysfunction, inflammation, and apoptotic cell death. Additionally, innate survival mechanisms and enhanced endogenous repair mechanisms are involved [5]. Preservation of energy metabolism and mitochondrial functions during the ischemic event is improved [167,168]. Our knowledge on the underlying mechanisms of cerebral IT is yet patchy. Additionally, different mechanisms may dominate different models. Here, we will review only the major molecular aspects contributing to delayed cerebral IT. The mechanisms of early IT will be discussed separately at the end of this section. Readers seeking for more comprehensive information should consult the recent excellent review of Obrenovitch [169] as well as its antecedents [9,133,167].

will review only the major molecular aspects contributing to delayed cerebral IT. The mechanisms of early IT will be discussed separately at the end of this section. Readers seeking for more comprehensive information should consult the recent excellent review of Obrenovitch [169] as well as its antecedents [9,133,167]. Hypoxia-inducible factor-1 (HIF-1) Among several transcription factors sensitive to regulation by hypoxia/ischemia, HIF isoforms have gained the most experimental support [5]. HIF-1 proteins are increased in the brain in the setting of hypoxia resistance [170] and hypoxic PC [67]. Pharmacological activators of HIF-1 (deferoxamine or cobalt chloride) promote PC in hypoxia/ischemia model in neonatal rats [170]. Over the past decade, the signalling pathways involved in HIF-1 activation have been deciphered in detail [171]. Briefly, hypoxia stabilizes alfa subunit of HIF-1, which enters the nucleus in a dimerized form and results in the induction of HIF target genes. Several HIF target genes contribute protection from ischemia [67,172,173], and their products involve in wide range of adaptive and pro-survival events, including cellular metabolism, proliferation, vascularization, iron homeostasis, and glucose metabolism [5,133].

form and results in the induction of HIF target genes. Several HIF target genes contribute protection from ischemia [67,172,173], and their products involve in wide range of adaptive and pro-survival events, including cellular metabolism, proliferation, vascularization, iron homeostasis, and glucose metabolism [5,133]. Protein kinase C (PKC) The role of protein kinase C in mediating stroke injury has been reviewed recently [174]. There are 10 isozymes in the PKC family. Previously, PKC was thought not to have a role in IT phenomenon, because blockage of PKC did not prevent IT [175], and PKC activation did not induce IT [176]. However, accumulating data suggest opposite roles for different PKC isozymes in the brain: γPKC contributes in ischemic cell death in organotypic hippocampal cell cultures, and NMDA triggered IT models require εPKC translocation [177]. Even though non-selective activation of PKC does not induce IT, specific εPKC activation leads to IT [143,177]. It seems that adenosine-mediated activation of εPKC and subsequent signal transduction pathways through MAPK-K, ERK [178], and cyclooxygenase-2 induction are involved in IT [143].

re εPKC translocation [177]. Even though non-selective activation of PKC does not induce IT, specific εPKC activation leads to IT [143,177]. It seems that adenosine-mediated activation of εPKC and subsequent signal transduction pathways through MAPK-K, ERK [178], and cyclooxygenase-2 induction are involved in IT [143]. Anti-excitotoxic mechanisms, NMDA, and calcium Exogenous application of NMDA or glutamate alone suffice to induce ischemia resistance in cell cultures, and NMDA receptor blockade during preconditioning eliminates IT both in vitro [179] and in vivo [180,181]. Specific AMPA or kainate receptor blockade do not eliminate or only partially attenuates IT [131,180]. Contradictory findings exist, however [136,182]. In gerbils, IPC increased inhibitory γ-amino butyric acid A (GABAA) receptor binding in hippocampus, whereas final ischemia did not [160]. Moreover, microdialysis experiments revealed a temporary increase in GABA release in preconditioned rat hippocampus early after final ischemia, with a decrease in glutamate concentration [21]. Thus, anti-excitotoxic mechanism induced in ischemia-tolerant state in global ischemia models involve a shift between inhibitory and excitatory hippocampal neurotransmission. In vitro, GABAB activation operates as a PC trigger [21], but not GABAA activation [21,183]. Regarding the role of Ca2+ in IT, it seems not always mandatory [184], but chelation of Ca2+ before and during both OGD- and NMDA-PC prevents IT in vitro [177].

t between inhibitory and excitatory hippocampal neurotransmission. In vitro, GABAB activation operates as a PC trigger [21], but not GABAA activation [21,183]. Regarding the role of Ca2+ in IT, it seems not always mandatory [184], but chelation of Ca2+ before and during both OGD- and NMDA-PC prevents IT in vitro [177]. Adenosine and ATP-sensitive K+ (KATP) channels Adenosine, an ischemia-induced degradation product of ATP, activates A1 receptors, which leads to a cascade of signaling events including KATPchannels. This cascade results in increased resistance to subsequent ischemic damage [185]. The general role of KATP channels, which are named for the inhibitory effect of ATP reducing channel opening probability, is to set membrane potential according to its metabolic state by sensing intracellular nucleotide concentrations [186]. Plasma membrane KATP channels are found widely throughout the brain [186]. The mandatory role of KATP channels for acquisition of IT was demonstrated in a rat delayed IT model (global-global) [187] and in vitro [188]. Interestingly, early IT is also blocked by pharmacological inhibition of KATP channels in vitro [176]. Opening of KATP channels is thought to relate to adenosine A1 receptor activation. Both specific and nonspecific adenosine A1 receptor antagonists attenuate or cancel the IT phenomenon [187,189,190]. However, SUR1-containing KATP channels seem not to be involved in IPC [35] and in spreading depression-PC model in rats and inhibition of KATP channels did not block IT [191]. Some authors emphasize a more pronounced role for mitochondrial KATP channel in IT [169,190].

tenuate or cancel the IT phenomenon [187,189,190]. However, SUR1-containing KATP channels seem not to be involved in IPC [35] and in spreading depression-PC model in rats and inhibition of KATP channels did not block IT [191]. Some authors emphasize a more pronounced role for mitochondrial KATP channel in IT [169,190]. Nitric oxide (NO) NO is one of the most extensively studied molecules in IT experiments (for reviews see [192,193]). Data suggest that generation of NO is crucial for the induction of IT, as a dependence on endothelial NO synthase (eNOS), but not on the neuronal NOS (nNOS) in newborn rats subjected to hypoxic-PC [165]. Whereas, nNOS was required to induce tolerance in vitro [166]. OGD tolerance in cortical cell cultures occurred via the activation of the Ras/extracellular signal-regulated kinase cascade by NO [131]. Atochin's early IT model proved an indispensable role for both eNOS and nNOS [50]. Puisieux et al. used a delayed IT (focal-focal) model in adult rats and showed no effect of NOS blockade on IT, but when the PC stimulus was LPS, IT was abolished by NOS inhibition [194]. Inducible NOS (iNOS) lacking mice experience no IT [164] and iNOS inhibition may nullify delayed IT to permanent focal ischemia, that otherwise follows isoflurane- or halotane-PC [113].

in adult rats and showed no effect of NOS blockade on IT, but when the PC stimulus was LPS, IT was abolished by NOS inhibition [194]. Inducible NOS (iNOS) lacking mice experience no IT [164] and iNOS inhibition may nullify delayed IT to permanent focal ischemia, that otherwise follows isoflurane- or halotane-PC [113]. Anti-inflammatory mechanisms Interleukin-1 (IL-1) and tumor necrosis factor-α (TNF-α) are implicated in IT induction: both cytokines are found increased in ischemic-tolerant state, both act as PC trigger when administered systemically, and their inhibition or lack significantly attenuate or block IT [195-198]. Pradillo et. al. explored the involvement of the TNF-α/nuclear factor-κB (NF-κB) signal transduction pathway in IT [48]. This pathway includes at least 131 interactors [199]. Activation of NF-κB is involved in IT in several models [200,201], likely via the induction of neuroprotective gene products, such as manganese superoxide dismutase and Bcl-2 [9]. Preconditioning with ligands of toll-like receptors 4 and 9 may alter innate inflammatory responses to ischemia by causing an initial activation of inflammatory mediators followed by a burst of inflammation inhibitors [202].

duction of neuroprotective gene products, such as manganese superoxide dismutase and Bcl-2 [9]. Preconditioning with ligands of toll-like receptors 4 and 9 may alter innate inflammatory responses to ischemia by causing an initial activation of inflammatory mediators followed by a burst of inflammation inhibitors [202]. Anti-apoptotic mechanisms PC blocks enhanced phosphorylation occurring after ischemia [9]. On the other hand, phosphorylation of transcriptional factors can induce long-term changes by regulating the expression of genes. IT is also characterized by reduced apoptosis [5,142]. Phosphaphatidylinositol 3-kinase/Akt pathway seems to act in two ways: 1) in relation to anti-apoptotic mechanisms and 2) by activating NFkB. In vitro, p21 Ras is required and sufficient to induce IT and Ras/Erk pathway is activated through NMDA receptor and NO production [131]. However, increasing evidences support the existence of a link between Akt activation and anti-apoptosis in IT [157,203-205], perhaps more persistently in penumbral regions in focal IT models [206]. Anti-apoptotic mechanisms induced by PC are several: induction of Bcl-2, reductions in caspase-3 synthesis and p-53 activation, and reductions in mitochondrial cytochrome c [9,185].

nk between Akt activation and anti-apoptosis in IT [157,203-205], perhaps more persistently in penumbral regions in focal IT models [206]. Anti-apoptotic mechanisms induced by PC are several: induction of Bcl-2, reductions in caspase-3 synthesis and p-53 activation, and reductions in mitochondrial cytochrome c [9,185]. Genomic reprogramming With the contribution of DNA microarray analysis method to IT research, we gained a better understanding of the preconditioned brain on the genetic level. In 2003, Stenzel-Poore and colleagues published a study, a cornerstone in the field, which introduced the concept of "genomic reprogramming" defining the altered transcriptional response of the ischemia-tolerant brain [207]. Followed by others [208,209], profiled the genetics of IT induced by IPC in rats were profiled. In the setting of IT, overall transcriptional response to injury was found altered as downregulation, which was strikingly different from that in the naïve brain's postischemic transcriptome. Suppression of gene expression in the ischemia-tolerant state was not simply the lack of response to injurious insult, but rather a reprogramming of the genetic response to ischemia [210]. Most of the genes suppressed are involved in the pathways that regulate metabolism, molecular transport, or cell-cycle control. Genomic transcriptional profile shows a substantial difference also between latent cerebroprotective and ischemia-tolerant states. None but one of the differentially regulated genes compared to healthy hemisphere are in common [208]; however, in both states, overall response is downregulation of genes involved in metabolism and transport/synaptic transmission. Using GeneChip analysis, Dhodda et al. evaluated temporal changes in gene expression after IPC in spontaneously hypertensive rats [158]. At the time-points studied (3, 6, 12, 24, and 72 h after PC), overall 40 transcripts were found up-regulated, among which 30 transcripts were overexpressed at all time-points, and the six HSP70 transcripts showed the highest increase. Other major families of transcripts, which were upregulated during PC, were those that control signal transduction, transcription, ionic homeostasis, and plasticity. Moreover, transcripts that showed upregulation after ischemia in näive brains were not found upregulated in ischemia-tolerant brains [158].

hest increase. Other major families of transcripts, which were upregulated during PC, were those that control signal transduction, transcription, ionic homeostasis, and plasticity. Moreover, transcripts that showed upregulation after ischemia in näive brains were not found upregulated in ischemia-tolerant brains [158]. Gene expression response to hypoxic-PC was also studied [173]. As early as 1 hour after hypoxia, but at a greater extent at 6 hours, expression of many genes, which are regulated by HIF-1, were increased. Compared to näive ischemic brains, in the ischemia-tolerant brains preconditioned with hypoxia, several genes were differentially upregulated. Genes with decreased expression in näive ischemic brains were no longer or only to a small degree underexpressed in ischemia-tolerant brains. Genetic response to hyperbaric oxygen-PC was studied in the rat, in the latent cerebroprotective state (at 6, 12, 24 after PC) [71]. Most of the differential regulations, including overexpression of genes and proteins related to neurotrophin and inflammatory-immune system, occurred around 12 and 24 hours. Genetic reprogramming was described as well for IT induced by erythromycin [118]. Mechanisms of early IT In vivo models demonstrating early IT in the brain are limited: global-global model in the rat [24], and focal-focal model both in the rat [211], and mouse [49,50], and anesthetic-PC against focal permanent ischemia in the rat [113].

Genetic response to hyperbaric oxygen-PC was studied in the rat, in the latent cerebroprotective state (at 6, 12, 24 after PC) [71]. Most of the differential regulations, including overexpression of genes and proteins related to neurotrophin and inflammatory-immune system, occurred around 12 and 24 hours. Genetic reprogramming was described as well for IT induced by erythromycin [118]. Mechanisms of early IT In vivo models demonstrating early IT in the brain are limited: global-global model in the rat [24], and focal-focal model both in the rat [211], and mouse [49,50], and anesthetic-PC against focal permanent ischemia in the rat [113]. The molecular mediators of early IT are little known. Changes in membrane channel activity and posttranslational modifications of existing proteins are among the few, which are well-described. Roles for adenosine receptor in vivo [211] and for KATP channel in vitro were also explored [176]. Several immediate-early genes (c-fos, c-jun), growth factors (brain-derived neurotrophic factor, nerve growth factor), and heat shock protein 70 were overexpressed during early latent ischemia-tolerant state [212]. According to Kariko et al. [213], during early tolerance, production of proinflammatory cytokines are suppressed, whereas in delayed tolerance, production of the very same cytokines are induced.

Introduction It is with great pleasure and enthusiasm that we introduce the new non-profit "Association for Supporting Young Scientists in the Field of Neurology in Germany" ("Verein zur Förderung des Wissenschaftlichen Nachwuchses in der Neurologie", NEUROWIND e.V.) http://www.neurowind.de. As its name suggests, the association is intended to promote the work of young neurologists and neuroscientists in German-speaking European countries. Founded by the neurologists Ralf Linker, Bochum, Thomas Korn, Munich, Tim Magnus, Hamburg, Sven G. Meuth, and Christoph Kleinschnitz, Würzburg, Germany, NEUROWIND e. V. aims to provide an interdisciplinary and interactive platform for young researchers in order to gather and disseminate new knowledge in the fields of clinical and basic neurosciences. NEUROWIND e. V. focuses on three main topics: [1] cerebrovascular diseases, [2] neuroinflammation, and [3] neurodegeneration.

factor, nerve growth factor), and heat shock protein 70 were overexpressed during early latent ischemia-tolerant state [212]. According to Kariko et al. [213], during early tolerance, production of proinflammatory cytokines are suppressed, whereas in delayed tolerance, production of the very same cytokines are induced. VI. Open Issues and Challenging Features of IT Several specific questions arise by an overview of past IT experiments. The nature of the PC stimulus and the duration of the interval between PC and final ischemia are among the main parameters that may affect the results. The strain and gender of the experimental animal are additional sources of variability, as we are familiar from stroke experiments. Therefore, findings of an experimental IT study should be interpreted considering the following issues. Trigger-dependent differences Experimental data amounted for the last 20 years clearly demonstrated that IT can be afforded in animals by miscellaneous PC triggers. Thus, one can both easily and reasonably make the following assumption: diverse PC are sharing a common or overlapping pathway. As discussed above, a number of effector mechanisms confer ischemia-tolerant phenotype, and recently, genetic reprogramming was proposed as the underlying common process set into motion by these mechanisms [207,208]. Below, we will have a closer look to studies comparing the mechanistic or molecular features of IT triggered by different PC stimuli.

f effector mechanisms confer ischemia-tolerant phenotype, and recently, genetic reprogramming was proposed as the underlying common process set into motion by these mechanisms [207,208]. Below, we will have a closer look to studies comparing the mechanistic or molecular features of IT triggered by different PC stimuli. IPC versus LPS-PC was compared in a transient ischemia rat model and found inducing similar degree of protection (35% reduction in infarct volume) [194]. An interesting finding was that NO synthase inhibition abolished the protective effect of LPS, but not of IPC. IPC induced the expression of heat shock protein 70 in the cerebral cortex, but LPS did not. Recently, ischemia-tolerant phenotypes induced by two well-known preconditioning stimuli -LPS and transient focal ischemia- have been evaluated from the genetic aspect [209]. Authors disclosed that a substantial subset of regulated genes were unique to each PC stimulus. In case of IPC, mainly metabolism and channel/transport-related genes were suppressed; whereas, LPS-PC induced expression of pro-inflammatory molecules and suppressed those genes related to deleterious inflammatory reactions. However, suppression of Toll-like receptor-mediated inflammation is a common mechanism triggered by both PC triggers [213]. Another comparative study of different PC stimuli (IPC and chemical PC with 3-NPA) addressed cytokine mRNA expression after final ischemia [214]. Both PC strategies exerted very similar effects on proinflammatory and cytotoxic cytokine expressions. Later, same authors studied the expression of nerve growth factor separately with IPC and 3-NPA-PC paradigms [215]. Neither trigger showed any effect on nerve growth factor expression, which in another study was found increased by PC with brief global ischemia in both early and delayed IT [212].

c cytokine expressions. Later, same authors studied the expression of nerve growth factor separately with IPC and 3-NPA-PC paradigms [215]. Neither trigger showed any effect on nerve growth factor expression, which in another study was found increased by PC with brief global ischemia in both early and delayed IT [212]. Intermodel differences In focal-global IT paradigm, PC may confer IT in neurons outside the primary area subjected to IPC that is in proximity, but not in the further regions such as contralateral hippocampus [62]. Similar IT paradigm in rats resulted in bilateral protection of hippocampi, however [63]. A functional direct pathway from the entorhinal cortex to both hippocampi was suggested to reflect the changes afforded by IPC to both hemispheres [63]. In the global-global IT paradigm, c-fos expression during the tolerant state was found specific to the cell type [216], which may explain selectivity of IT induction to certain brain areas. Prass et al. studied the confounding effects of strain and reperfusion on the IT phenomenon [69]. Hyperbaric oxygen was applied as PC stimulus to two common background strains for knockout mice, SV129 and C57BL/6. Final ischemia was either permanent or transient focal ischemia. In SV129 mice, PC induced tolerance to permanent ischemia but not to transient ischemia. In C57BL/6 mice, IT did not occur at all. Consequently, questions to answer with further study are: 1. For what reasons the very same trigger induced IT in a strain but not in another, and 2) Can reperfusion nullify the protection afforded by PC?

PC induced tolerance to permanent ischemia but not to transient ischemia. In C57BL/6 mice, IT did not occur at all. Consequently, questions to answer with further study are: 1. For what reasons the very same trigger induced IT in a strain but not in another, and 2) Can reperfusion nullify the protection afforded by PC? Gender Female rats sustain smaller infarcts after MCAO than males [217] and estrogen is neuroprotective in ovariectomized females and in males subjected to ischemic stroke [218]. Data from heart IT experiments show a clear gender-dependency of the IT phenomenon [219,220], this issue seems valid also in cerebral IT. Estrogen provided IT in a model of hippocampal organotypic slice culture, which was generated from neonatal female rats [116], and isoflurane induced IT only in male mice and increased the infarction in young female mice [221]. Age IT phenomenon is preserved in aged animals [222], but may not be as effective as it is in young animals. This aspect was tested with a global-global IT paradigm applied in 4- and 24-month-old rats [223]. The degree of protection due to PC was significantly diminished in aged rats compared to young rats. A retrospective clinical study indicated that IT may not be occurring in the elderly, aged around 75 [224].

ung animals. This aspect was tested with a global-global IT paradigm applied in 4- and 24-month-old rats [223]. The degree of protection due to PC was significantly diminished in aged rats compared to young rats. A retrospective clinical study indicated that IT may not be occurring in the elderly, aged around 75 [224]. Repeated PC Cumulative injurious effect of repeated cerebral ischemia is a well-known phenomenon. For example, three periods of 5-min forebrain ischemia, induced at 1-hour intervals, result in more extensive brain injury than one single episode of 15-min ischemia in gerbils [225]. However, if PC insults are applied repeatedly, a larger IT response may be gained. This was tested in a mice model of early IT, in which animals underwent either single or 3 episodes of 5-min focal cerebral ischemia, 30 min before permanent ischemia [49]. Only repeated insults conferred IT, the single brief ischemia was insufficient to induce IT. Similarly, a single episode of 2 min OGD is under the threshold to act as a PC stimulus, but four times repeated 2 min of OGD show a cumulative effect and protects from subsequent injurious insult [226]. Hyperbaric oxygen-PC, when applied singly or repeatedly, provide similar degree of protection from transient focal ischemia (63% vs 73% lesion reduction) [227], perhaps this is the maximum affordable protection by hyperbaric oxygen. In a clinical study however, anesthetic-PC with a single application induced no IT in the heart, whereas repeated application did [115]. In the pig heart, PC by repetitive ischemic insults was shown to induce a different set of genetic regulations from those induced with PC with single ischemic episode [228]. A corresponding study in cerebral IT is needed.

c-PC with a single application induced no IT in the heart, whereas repeated application did [115]. In the pig heart, PC by repetitive ischemic insults was shown to induce a different set of genetic regulations from those induced with PC with single ischemic episode [228]. A corresponding study in cerebral IT is needed. "Sublethality" of PC Although PC is defined as a sublethal stimulus, which per se causes no injury, several studies used relatively severe focal ischemia as the PC trigger and were able to induce IT, despite the injurious nature of the PC itself [63,64,190]. As pointed out by Sommer [137], with extended follow-up after the PC insult, some injury or structural changes can be detected. Therefore, it is suggested that PC is postponing these changes [137]. If that holds true, in the long-term, näive ischemic brains and IT experienced brains may have similar outcomes. This issue is discussed next.

mmer [137], with extended follow-up after the PC insult, some injury or structural changes can be detected. Therefore, it is suggested that PC is postponing these changes [137]. If that holds true, in the long-term, näive ischemic brains and IT experienced brains may have similar outcomes. This issue is discussed next. IT and long-term effects Early IT is a short-lasting phenomenon, its protection vanishes around 7 days [24]. In delayed IT models, protection lasts longer and tends to decline after 30-60 days. Ohno et al. applied a global-global IT model to rats [229] and showed that improvement in learning and memory due to IPC was preserved up to 3 weeks. Protective effects of spreading depression-PC and LPS-PC sustained up to 14 days [84,156]. Ma et al. found a sustained improvement in neurological scores up to 30 days in xenon-preconditioned neonatal mice subjected to global ischemia [112], a similar finding was reported with a focal-focal IT model in rats [51]. In global-global IT models, histological protection is longer preserved in rats (up to 90 days) than in gerbils (up to 60 days) [30,32,222,230]. Optimizing time interval between PC and final ischemia, together with the optimization of the PC stimulus (single or repetitive application) and the severity of final ischemic insult, may result in long-term preservation of protection achieved by PC [32], on which increased neurogenesis after PC [231] may have a potential role.

g time interval between PC and final ischemia, together with the optimization of the PC stimulus (single or repetitive application) and the severity of final ischemic insult, may result in long-term preservation of protection achieved by PC [32], on which increased neurogenesis after PC [231] may have a potential role. VII. Clinical Aspects To date, a body of evidence, which supports the hypothesis that TIAs may confer IT in humans, exists. In a retrospective study, preceding TIA was found to be associated with less-severe stroke on admission and improved outcome on follow-up, compared to stroke patients without preceding TIA [232]. Another retrospective case-control study, found no evidence of PC by TIA in baseline neurological scores, but favorable outcome was associated with the presence of TIA [233]. This study presented "potentially preconditioning" TIA characteristics as: 0-7 day interval between TIA and stroke, 2 or 3 times repeated TIA, and TIA with <20 min duration. Moncayo et al. reported a cohort of 65 patients with acute anterior circulation stroke, among whom those with previous TIAs (lasting less than 20 minutes), had a more favorable outcome than those without [234]. Apparently, duration of TIA should be taken into account while evaluating whether IT exists in humans or not [235]. An MRI study provided the tissue evidence for TIA-induced tolerance to ischemic stroke [236]. Ischemic lesions tended to be smaller on the baseline images and final infarct volumes were smaller in stroke patients with prior TIA than in those without.

to account while evaluating whether IT exists in humans or not [235]. An MRI study provided the tissue evidence for TIA-induced tolerance to ischemic stroke [236]. Ischemic lesions tended to be smaller on the baseline images and final infarct volumes were smaller in stroke patients with prior TIA than in those without. Although these findings strongly suggest TIA as the clinical correlate of IPC, other explanations for milder strokes after preceding TIA must be considered. In these patients, a carotid disease with slowly progressing stenosis, which improves collateral circulation may predominate [237]. Another point is that, patients with cardioembolic stoke have lower incidence of TIA than those with atherosclerotic vascular disease, and probably because of larger-sized emboli they sustain larger infarcts and poorer outcome [238]. Several clinical conditions may benefit from strategies using principles of ischemic tolerance, as discussed elegantly by Dirnagl et al. in a recent review article on cerebral IT [239]. Mediators of IT could be used as biochemical markers of IT in stroke patients. Castillo et al. tested this hypothesis by evaluating blood levels of TNF-α and IL-6 in acute stroke patients with or without prior ipsilateral TIA [240]. Better outcome was found in patients with TIA, who showed high plasma concentrations of TNF-α and low concentrations of IL-6. Hence, authors proposed the index of TNF-α/IL-6 as a marker of IT phenomenon in humans.

ating blood levels of TNF-α and IL-6 in acute stroke patients with or without prior ipsilateral TIA [240]. Better outcome was found in patients with TIA, who showed high plasma concentrations of TNF-α and low concentrations of IL-6. Hence, authors proposed the index of TNF-α/IL-6 as a marker of IT phenomenon in humans. Conclusions Experimental IT paradigms investigate the endogenous pathways by which the brain might protect itself from ischemia when geared with an appropriate stimulus. Attempts to elucidate the mechanisms underlying cerebral IT are increasing exponentially, but diversity of models, including PC stimuli, hardens interpretation of the data. In addition, narrow safety margin of PC may prove a limiting factor of the therapeutic utility of PC in clinics. On the other hand, accumulating clinical data suggest that IT might be a clinically relevant phenomenon. Several approaches, including ICP [241], remote-PC by limb ischemia [88], pharmacological-PC with nitroglycerine [242], and anesthetic-PC [115], are tested in clinical trials to protect the heart from cardiovascular interventions with high risk of cardiac ischemic event. Results are promising and give hope that clinical trials of PC to protect brain in situations with a high risk of ischemia can be designed, once PC is proven safe.

42], and anesthetic-PC [115], are tested in clinical trials to protect the heart from cardiovascular interventions with high risk of cardiac ischemic event. Results are promising and give hope that clinical trials of PC to protect brain in situations with a high risk of ischemia can be designed, once PC is proven safe. Abbreviations 3-NPA: 3-nitropropionic acid; GABA: γ-amino butyric acid; HIF-1: hypoxia-inducible factor-1; IL-1: interleukin-1; IT: ischemic tolerance; IPC: ischemic preconditioning; KATP, ATP-sensitive K+; LPS: lipopolysaccharide; MCA: middle cerebral artery; MCAO: middle cerebral artery occlusion; MRI: magnetic resonance imaging; NO: nitric oxide; nNOS: neuronal NO synthase; iNOS: inducible NOS; eNOS: endothelial NOS; NF-KB: nuclear factor KB; OGD: oxygen-glucose deprivation; PC: preconditioning; PKC: protein kinase C; TIA: transient ischemic attack; TNF-α: tumor necrosis factor-α; tMCA: transient MCAO. Competing interests The authors declare that they have no competing interests. Authors' contributions Both AD and TT made the conception and design of the manuscript and analysis and interpretation of the data, drafted and revised the manuscript, and have given approval of its final version. Supplementary Material Additional file 1 Models for ischemic tolerance in rodents. Click here for file Acknowledgements This work was supported in part by the Helsinki University Central Hospital and the Finnish Academy of Sciences.

Background There is increasing evidence that inflammatory processes play a detrimental role in ischemic stroke [1,2]. On the other hand, the postischemic immune response may also be beneficial with respect to neuroprotection and tissue remodelling. The proinflammatory cytokine IL-18 is an interleukin-1 family member identified as and named interferon-γ-inducing factor [3,4]. The effects of IL-18 are complex and pleiotropic involving activation of T cells and NK cells in autoimmune disorders (for review, see Reddy [5]). In the central nervous system, IL-18 can locally be produced by activated microglia [6,7]. Increased IL-18 serum levels have been detected within 24 hours in patients with acute ischemic stroke [8], and elevated IL-18 plasma levels at 48 hours were associated with unfavorable clinical outcome at 3 months [9]. Moreover, in hypoxic-ischemic brain injury in neonatal rats, an early IL-18 activation (already within hours) and a progressive increase for at least 14 days have been described [10]. At 3 days after hypoxia-ischemia, IL-18 deficiency has been shown to ameliorate infarct volume and grey matter injury [10] as well as white matter injury [11] in neonatal mice. These studies suggest that IL-18 may play a pathophysiological role in stroke development. To elucidate a functional role of IL-18 in cerebral ischemia we investigated infarct size and functional outcome 24 hours and 48 hours after tMCAO in adult mice with IL-18 deficiency.

as white matter injury [11] in neonatal mice. These studies suggest that IL-18 may play a pathophysiological role in stroke development. To elucidate a functional role of IL-18 in cerebral ischemia we investigated infarct size and functional outcome 24 hours and 48 hours after tMCAO in adult mice with IL-18 deficiency. Methods Animal studies were conducted in accordance with institutional guidelines and approved by the appropriate authorities (Regierung von Unterfranken). Wild-type and IL-18 knock-out mice that had been generated [12] and backcrossed onto BALB/C mice [13] as described were kindly provided by Drs. X-Q Wei and FY Liew, Glasgow, Scotland, and bred and raised in our laboratory animal facility. A total of 35 wild-type and 29 IL-18 knock-out animals (plus additional 11 wild-type and 10 knock-out mice for laser-Doppler flowmetry and ink perfusion) were used.

o BALB/C mice [13] as described were kindly provided by Drs. X-Q Wei and FY Liew, Glasgow, Scotland, and bred and raised in our laboratory animal facility. A total of 35 wild-type and 29 IL-18 knock-out animals (plus additional 11 wild-type and 10 knock-out mice for laser-Doppler flowmetry and ink perfusion) were used. Mice weighting 18-24 g were subjected to transient focal cerebral ischemia in the right middle cerebral artery (MCA) territory for 1 hour using the intraluminal suture MCA occlusion method [14]. In brief, mice were anesthetized with 2% to 2.5% enflurane in a 70% N2O/30% O2 mixture. A servo-controlled heating blanket was used to maintain core body temperature close to 37°C throughout surgery. A silicon rubber-coated 6.0 nylon monofilament (Doccol, Albuquerque, NM) was inserted into the right common carotid artery and advanced via the internal carotid artery to occlude the origin of the MCA, causing focal ischemic brain injury in the right MCA territory. The occluding filament was removed after 1 hour to allow reperfusion. Animals were sacrificed 24 hours or 48 hours after tMCAO. Brains were harvested and 2 mm-thick coronal slices were sectioned in a mouse brain matrix. After staining with 2% 2,3,5-triphenyltetrazolium chloride (TTC; Sigma-Aldrich, St. Louis, MO) in PBS, the pale infarctions were readily discernable from the brick-red non-ischemic areas and planimetric measurements were obtained using the ImageJ software package (available at http://rsb.info.nih.gov/ij/; National Institutes of Health, Bethesda, MD). The calculated lesion volume was corrected for brain swelling as described by Ginsberg et al. [15].

dily discernable from the brick-red non-ischemic areas and planimetric measurements were obtained using the ImageJ software package (available at http://rsb.info.nih.gov/ij/; National Institutes of Health, Bethesda, MD). The calculated lesion volume was corrected for brain swelling as described by Ginsberg et al. [15]. Additionally, we assessed the functional outcome in representative subsets of the animals. Immediately after recovery from anesthesia, and 24 hours and 48 hours later, a modified Bederson score [16] was determined according to the following scoring system: 0, no deficit; 1, forelimb flexion; 2, as for 1, plus decreased resistance to lateral push; 3, unidirectional circling; 4, longitudinal spinning or seizure activity; 5, no movement. 24 hours and 48 hours after surgery, the foot fault test and grip test were performed. The foot fault test was done as described by Gibson et al. [17], with the following modifications: Mice were placed on an elevated grid with 1.44 cm2 openings and allowed to take 25 paired steps. Animals not moving spontaneously for at least 25 steps were excluded. The number of foot faults of the ipsilateral and contralateral limbs was counted. Foot faults were given as the percentage of contralateral (left) limb foot fault errors of all errors made. The grip test, also known as string test, was adopted from Moran et al. [18], with modified scoring system. For this test, the mouse was placed midway on a string between two supports and rated as follows: 0, falls off; 1, hangs onto string by one or both forepaws; 2, as for 1, and attempts to climb onto string; 3, hangs onto string by one or both forepaws plus one or both hindpaws; 4, hangs onto string by fore- and hindpaws plus tail wrapped around string; 5, escape (to the supports).

two supports and rated as follows: 0, falls off; 1, hangs onto string by one or both forepaws; 2, as for 1, and attempts to climb onto string; 3, hangs onto string by one or both forepaws plus one or both hindpaws; 4, hangs onto string by fore- and hindpaws plus tail wrapped around string; 5, escape (to the supports). Laser-Doppler flowmetry (Moor Instruments, Axminster, United Kingdom) was used to monitor cerebral blood flow in 6 IL-18 +/+ and 5 IL-18 -/- animals before surgery (baseline), immediately after tMCAO, and 5 minutes after removal of the occluding monofilament (reperfusion). For this, a flexible laser-Doppler probe was positioned perpendicular to the exposed skull 2 mm posterior and 6 mm dexterolateral to the bregma, corresponding to the laser-Doppler probe position in murine MCA occlusion reported by Connolly et al. [19]. These animals were not included in the infarct size and functional outcome evaluations, because laser-Doppler flowmetric measurements in addition to tMCAO inevitably leads to prolonged operation times.

he bregma, corresponding to the laser-Doppler probe position in murine MCA occlusion reported by Connolly et al. [19]. These animals were not included in the infarct size and functional outcome evaluations, because laser-Doppler flowmetric measurements in addition to tMCAO inevitably leads to prolonged operation times. To exclude anatomic differences that could cause different susceptibility to tMCAO in wild-type and knock-out mice, we studied the cerebrovasculature in 5 IL-18 +/+ and 5 IL-18 -/- mice. Animals were deeply anesthetized with CO2 and transcardially perfused with 4% paraformaldehyde and then with black ink (T25; Edding, Ahrensburg, Germany). After brain removal and overnight fixation in 4% paraformaldehyde, the circle of Willis was visualized under a dissecting microscope. Special attention was paid to the posterior communicating arteries, whose level of plasticity was rated as described by Murakami et al. [20]: 0, absent; 1, capillary anastomosis; 2, small truncal vessel; 3, patent. A posterior communicating artery with a score of 0 or 1 is regarded as hypoplastic, and with a score of 2 or 3 as normal. Corrected infarct volumes, data from the functional outcome tests and from the posterior communicating artery score, and laser-Doppler flow measurements were statistically analyzed in two-tailed Mann-Whitney U tests using Prism 4 (GraphPad Software, San Diego, CA).

To exclude anatomic differences that could cause different susceptibility to tMCAO in wild-type and knock-out mice, we studied the cerebrovasculature in 5 IL-18 +/+ and 5 IL-18 -/- mice. Animals were deeply anesthetized with CO2 and transcardially perfused with 4% paraformaldehyde and then with black ink (T25; Edding, Ahrensburg, Germany). After brain removal and overnight fixation in 4% paraformaldehyde, the circle of Willis was visualized under a dissecting microscope. Special attention was paid to the posterior communicating arteries, whose level of plasticity was rated as described by Murakami et al. [20]: 0, absent; 1, capillary anastomosis; 2, small truncal vessel; 3, patent. A posterior communicating artery with a score of 0 or 1 is regarded as hypoplastic, and with a score of 2 or 3 as normal. Corrected infarct volumes, data from the functional outcome tests and from the posterior communicating artery score, and laser-Doppler flow measurements were statistically analyzed in two-tailed Mann-Whitney U tests using Prism 4 (GraphPad Software, San Diego, CA). Results Ink perfusion was performed in IL-18 wild-type and knock-out mice to visualize the complete circle of Willis. No gross anatomic differences were noted that could influence stroke outcome (Figure 1). The score assessing formation of the posterior communicating arteries of both hemispheres, which are pivotal in collateral blood flow between the anterior and posterior circulation, did not differ significantly in wild-type and knock-out mice (median of 2 in both groups; p = 0.91). Moreover, Laser-Doppler flowmetry ensured technical accuracy and similar basic characteristics in IL-18 +/+ and IL-18 -/- mice, since it did not show any significant differences between wild-type and knock-out animals. After right MCA occlusion, there was a similar substantial reduction of right hemispheric cerebral blood flow (median, 16.15% and 16.4%, respectively; p = 0.93). The blood flow recovered to a median of approximately 50% of baseline blood flow already within minutes after removal of the occluding intraluminal monofilament (median, 48.9% and 50.9%, respectively; p = 1.00).

ntial reduction of right hemispheric cerebral blood flow (median, 16.15% and 16.4%, respectively; p = 0.93). The blood flow recovered to a median of approximately 50% of baseline blood flow already within minutes after removal of the occluding intraluminal monofilament (median, 48.9% and 50.9%, respectively; p = 1.00). Figure 1 Ink perfusion. Images of a BALB/C wild-type mouse brain (A) and of a brain from an interleukin-18 knock-out mouse on a BALB/C background (B) with ink-perfused cerebrovasculature. We next assessed the influence of IL-18 on infarct size and on functional outcome. There were no significant differences in edema-corrected infarct size on standardized TTC-stained brain slices between wild-type and IL-18 knock-out mice after 24 hours (median, 79.25 mm3 and 86.1 mm3, respectively; interquartile range, 41.9 - 90.3 mm3 and 38.1 - 95.6 mm3, respectively; p = 0.51) or 48 hours (median, 93.05 mm3 and 85.2 mm3, respectively; interquartile range, 75.3 - 104.6 mm3 and 73.9 - 98.7 mm3, respectively; p = 0.36). These results are presented in Figure 2A. The functional outcome scores were not significantly different in IL-18 +/+ as compared to IL-18 -/- mice, too (Fig. 2B, C, D).

= 0.51) or 48 hours (median, 93.05 mm3 and 85.2 mm3, respectively; interquartile range, 75.3 - 104.6 mm3 and 73.9 - 98.7 mm3, respectively; p = 0.36). These results are presented in Figure 2A. The functional outcome scores were not significantly different in IL-18 +/+ as compared to IL-18 -/- mice, too (Fig. 2B, C, D). Figure 2 Infarct volumes and functional outcome. Infarct volumes (A) and functional outcome scores (B: modified Bederson test score; C: foot fault test; D: grip test score) of interleukin-18 wild-type and knock-out mice. The results are diagrammed as whisker boxes with medians. Boxes represent interquartile ranges and whiskers indicate extreme values. The p values resulting from Mann-Whitney U tests are given; all p values were greater than 0.05 and were thus considered insignificant. Abbreviations: wt, wild-type animals; ko, interleukin-18 knock-out animals; n, number of animals; 0 h, 0 hours (postoperatively after recovery from anesthesia); 24 h and 48 h, 24 hours and 48 hours after 1-hour middle cerebral artery occlusion.

all p values were greater than 0.05 and were thus considered insignificant. Abbreviations: wt, wild-type animals; ko, interleukin-18 knock-out animals; n, number of animals; 0 h, 0 hours (postoperatively after recovery from anesthesia); 24 h and 48 h, 24 hours and 48 hours after 1-hour middle cerebral artery occlusion. Discussion As principal finding, we show that deficiency of IL-18 does not protect mice from ischemic brain damage after tMCAO. These findings are surprising given the reported upregulation of IL-18 blood levels in stroke patients [8,9] associated with adverse clinical outcome [9] and the profound impact of IL-18 in experimental neonatal stroke [10,11]. However, two recent nested case-control studies have not confirmed an association of IL-18 with increased risk of stroke in older people [21] or with recurrent stroke [22]. Moreover, our data are in accordance with a previous study by Wheeler et al. showing no differences in infarct size at 24 hours between wild-type and IL-18 -/- mice on a C57BL/6 background subjected to 15 and 30 minutes of tMCAO which leads to smaller infarcts than 1 hour occlusion time [23]. We extend this previous study by applying a longer tMCAO time (1 hour) leading to infarcts involving the entire MCA territory, by assessing functional outcome and by following infarct development up to 48 hours. Recently, IL-18 expression and activation has been described already at 24 hours in a thromboembolic murine stroke model [24]. In contrast, in the rat, IL-18 mRNA expression was increased later at 48 hours, and peaked between 7 and 14 days [25]. IL-18 has been localized to microglia/macrophages within ischemic lesions [25]. The structurally similar interleukin-1β reached a peak already within 16 hours and was rapidly downregulated subsequently [25]. Thus, unlike interleukin-1β, IL-18 seems to be associated with the mid-stage inflammatory response to ischemic brain lesions. Accordingly, similar findings were reported for traumatic brain injury in mice: IL-18 was significantly elevated at 7 days, but not within 4 hours to 24 hours, after experimental closed head injury as compared to sham treatment [26]. Inhibition of IL-18 by IL-18-binding protein resulted in improved neurological recovery by 7 days, while brain edema at 24 hours was not reduced [26]. The cytokine response in neonatal rodents subjected to hypoxic-ischemic brain injury may differ.

fter experimental closed head injury as compared to sham treatment [26]. Inhibition of IL-18 by IL-18-binding protein resulted in improved neurological recovery by 7 days, while brain edema at 24 hours was not reduced [26]. The cytokine response in neonatal rodents subjected to hypoxic-ischemic brain injury may differ. Here, a significant mRNA elevation for IL-18 has been reported to occur already at 3 hours, but also progressively increased until day 14 [10]. In summary, our findings in adult IL-18 knock-out mice support the notion that IL-18 is not functionally relevant for early stroke development, but may play a role in late-stage neuroinflammation after stroke which awaits further elucidation. Competing interests The authors declare that they have no competing interests. Authors' contributions SB and CK wrote the paper and conducted the experiments. GS designed the study and reviewed the paper. All authors read and approved the final manuscript. Acknowledgements IL-18 knock-out mice and corresponding wild-type mice were kindly provided by Prof. Liew, Glasgow, Scotland. We thank Gabi Koellner for excellent technical assistance.

Background Although ischemic stroke represents a major health care problem with a high rate of permanent disability or even death, the underlying molecular mechanisms leading to neuronal death are still poorly understood [1]. However, ion channels which can influence basal cellular parameters are thought to play a major role within this context. Activation of potassium channels results in membrane hyperpolarization thereby decreasing neuronal activity and cell death under pathophysiological conditions. Additionally, K+ channels (e.g. large conductance Ca2+-activated K+ channels and ATP-sensitive K+ channels [2,3]) might be neuroprotective as they counterbalance a prolonged harmful influx of Ca2+ ions via different pathways including a reversal of the Na+/Ca2+ antiporter and voltage-dependent Ca2+ channels. Furthermore, an enhancement of the Mg2+ block of NMDA receptors (N-methyl D-aspartate) in postsynaptic neurons [4] is thought to protect against glutamate excitotoxicity [5,6].

olonged harmful influx of Ca2+ ions via different pathways including a reversal of the Na+/Ca2+ antiporter and voltage-dependent Ca2+ channels. Furthermore, an enhancement of the Mg2+ block of NMDA receptors (N-methyl D-aspartate) in postsynaptic neurons [4] is thought to protect against glutamate excitotoxicity [5,6]. Concerning the recently identified family of two-pore domain potassium channels (K2P channels), several members have been shown to play a major role in critical conditions leading to cerebral ischemia. K2P2.1-/- mice displayed significantly less neuronal survival rates in a model of cerebral ischemia [7]. These data were confirmed by the neuroprotective effect of several K2P2.1 channel activators (e.g. alpha linelonic acid or riluzole [8-10]). On the other hand, genetic depletion of another family member, namely K2P3.1, resulted in increased infarct volumes following transient or permanent middle cerebral artery occlusion (MCAO) [11,12]. Based on sequence homologies and similar biophysical properties, it was suggested that related channel family members might also be of importance under these circumstances. We challenged the role of K2P9.1 (TASK3; KCNK9) in a tMCAO model using previously described K2P9.1-/- mice [13].

l artery occlusion (MCAO) [11,12]. Based on sequence homologies and similar biophysical properties, it was suggested that related channel family members might also be of importance under these circumstances. We challenged the role of K2P9.1 (TASK3; KCNK9) in a tMCAO model using previously described K2P9.1-/- mice [13]. Methods Slice preparation Thalamic tissue slices including the dorsal lateral geniculate nucleus (dLGN) were prepared from 14 - 22 days old male C57BL/6 or K2P9.1-/- mice [13] as described earlier [14]. Coronal sections were cut on a vibratome (Vibratome®, Series 1000 Classic, St. Louis, USA) in an ice-chilled solution containing (mM): Sucrose, 200; PIPES, 20; KCl, 2.5; NaH2PO4, 1.25; MgSO4, 10; CaCl2, 0.5; dextrose, 10; pH 7.35 adjusted with NaOH. Prior to the recording procedure, slices were kept submerged in artificial cerebrospinal fluid (ACSF, mM): NaCl, 125; KCl, 2.5; NaH2PO4, 1.25; NaHCO3, 24; MgSO4, 2; CaCl2, 2; dextrose, 10; pH adjusted to 7.35 by bubbling with a mixture of 95% O2 and 5% CO2.

SO4, 10; CaCl2, 0.5; dextrose, 10; pH 7.35 adjusted with NaOH. Prior to the recording procedure, slices were kept submerged in artificial cerebrospinal fluid (ACSF, mM): NaCl, 125; KCl, 2.5; NaH2PO4, 1.25; NaHCO3, 24; MgSO4, 2; CaCl2, 2; dextrose, 10; pH adjusted to 7.35 by bubbling with a mixture of 95% O2 and 5% CO2. Electrophysiology Slices were transferred in a recording chamber and thalamic neurons of the dLGN were visualized with a microscope equipped with infrared-differential interference contrast optics [15]. Whole-cell recording pipettes were fabricated from borosilicate glass (GT150T-10, Clark Electromedical Instruments, Pangbourne, UK; typical resistance 2-3 MΩ) and filled with an intracellular solution containing (in mM): K-gluconate, 88; K3-citrate, 20; NaCl, 10; HEPES, 10; MgCl2, 1; CaCl2, 0.5; BAPTA, 3; phosphocreatin, 15; Mg-ATP, 3; Na-GTP, 0.5. The internal solution was set to a pH value of 7.25 using KOH and an osmolarity of 295 mOsm/kg. Slices were continuously superfused with a solution containing NaCl 125 mM, KCl 2.5 mM, NaH2PO4 1.25 mM, HEPES 30 mM, MgSO4 2 mM, CaCl2 2 mM and dextrose 10 mM. Whole-cell patch-clamp recordings were measured from relay neurons of the dLGN with an EPC-10 amplifier (HEKA Elektronik, Lamprecht, Germany) and digitally analyzed using Pulse software (HEKA Elektronik; [16]). pH was adjusted to 7.35 or 6.0 with HCl. For divalent-cation-free conditions we switched from control solution to a solution containing 0 mM Mg2+ and 0 mM Ca2+; the osmolality was kept constant at 305 mosmol kg-1 by adding 4 mM NaCl [17]. All cells had a resting membrane potential negative to -65 mV, the access resistance was in the range of 5-15 MΩ and series resistance compensation of more than 40% was routinely used.

olution to a solution containing 0 mM Mg2+ and 0 mM Ca2+; the osmolality was kept constant at 305 mosmol kg-1 by adding 4 mM NaCl [17]. All cells had a resting membrane potential negative to -65 mV, the access resistance was in the range of 5-15 MΩ and series resistance compensation of more than 40% was routinely used. Induction of cerebral ischemia Animal experiments were approved by governmental agencies for animal research and conducted according to the recommendations for research in mechanism-driven basic stroke studies [18]. Focal cerebral ischemia was induced in 6-8 weeks old male C57BL/6 and K2P9.1-/- mice [13] weighing 20-25 g by transient middle cerebral artery occlusion (tMCAO) as described previously [19,20]. Briefly, mice were anesthetized with 2.5% isoflurane (Abbott, Wiesbaden, Germany) in a 70% N2O/30% O2 mixture. Core body temperature was maintained at 37°C throughout surgery using a feedback-controlled heating device. Following a midline skin incision in the neck, the proximal common carotid artery and the external carotid artery were ligated and a standardized silicon rubber-coated 6.0 nylon monofilament (6021; Doccol Corp., CA, USA) was inserted and advanced via the right internal carotid artery to occlude the origin of the right MCA. The intraluminal suture was left in situ for 1 hour, respectively. Then animals were re-anesthetized and the occluding monofilament was withdrawn to allow reperfusion. After 24 hours neurological deficits were scored by two blinded investigators and quantified according to Bederson [21]: 0, no deficit; 1, forelimb flexion; 2, as for 1, plus decreased resistance to lateral push; 3, unidirectional circling; 4, longitudinal spinning; 5, no movement. For the gript test, the mouse was placed midway on a string between two supports and rated as follows: 0, falls off; 1, hangs onto string by one or both forepaws; 2, as for 1, and attempts to climb onto string; 3, hangs onto string by one or both forepaws plus one or both hindpaws; 4, hangs onto string by fore- and hindpaws plus tail wrapped around string; 5, escape (to the supports).

two supports and rated as follows: 0, falls off; 1, hangs onto string by one or both forepaws; 2, as for 1, and attempts to climb onto string; 3, hangs onto string by one or both forepaws plus one or both hindpaws; 4, hangs onto string by fore- and hindpaws plus tail wrapped around string; 5, escape (to the supports). Laser doppler flowmetry (Moor Instruments, Axminster, United Kingdom) was used to monitor cerebral blood flow [22] in wildtype, K2P9.1-/- and sham-treated animals (n = 4/group) before surgery (baseline), immediately after MCA occlusion, and 5 minutes after removal of the occluding monofilament (reperfusion). Cerebral perfusion did not differ between the groups at any time point (Additional File 1, Figure S1). Determination of infarct size Mice were sacrificed 24 hours after tMCAO, respectively. Brains were quickly removed and cut in 2 mm thick coronal sections using a mouse brain slice matrix. The slices were stained with 2% 2,3,5-triphenyltetrazolium chloride (TTC; Sigma-Aldrich, St. Louis, MO) in PBS to visualize the infarctions. Planimetric measurements (ImageJ software, National Institutes of Health, Bethesda, MD) blinded to the treatment groups were used to calculate lesion volumes, which were corrected for brain edema as described [23,24].

triphenyltetrazolium chloride (TTC; Sigma-Aldrich, St. Louis, MO) in PBS to visualize the infarctions. Planimetric measurements (ImageJ software, National Institutes of Health, Bethesda, MD) blinded to the treatment groups were used to calculate lesion volumes, which were corrected for brain edema as described [23,24]. Statistical analysis Electrophysiological data and results from the animal experiments were analyzed by a modified student's t test for small samples [25] or by a Bonferroni-corrected One-way ANOVA in case of multiple comparisons using PrismGraph 4.0 software (GraphPad Software, San Diego, CA) or Origin® (Microcal). P-values < 0.05 were considered statistically significant.

ts from the animal experiments were analyzed by a modified student's t test for small samples [25] or by a Bonferroni-corrected One-way ANOVA in case of multiple comparisons using PrismGraph 4.0 software (GraphPad Software, San Diego, CA) or Origin® (Microcal). P-values < 0.05 were considered statistically significant. Results Thalamic relay neurons as a model system of central nervous system neurons display electrophysiological properties indicative of currents through K2P9.1 channels K2P9.1-like currents have been demonstrated in a number of different central nervous system neurons [14,26-28]. As highly specific inhibitors for K2P channel subtypes are not available, different semi-selective blockers and experimental strategies to distinguish between these channels were established. Among them, extracellular reduction of divalent cations was introduced to increase potassium outward currents through K2P9.1 channels [17]. Current-voltage relationships (I/V) of the standing outward current of wildtype and K2P9.1-/- mice were investigated by ramping the membrane potential from -35 mV to -125 mV over 800 ms (Fig. 1A, inset; [29,30]). Under control conditions a standing outward current (ISO) of 322.33 ± 30.20 pA was measured at -35 mV (Fig. 1A). Application of hyperpolarizing voltage ramps induced a complex current response. The wave form of this response was indicative for the contribution of current through outwardly rectifying TASK channels as well as inwardly rectifying K+ channels (Fig. 1A, black trace). Removal of extracellular divalent cations resulted in a significant increase of ISO by 35.47 ± 9.59% compared to control conditions (n = 6, p = 0.007; Fig. 1A). Ramp responses revealed a clear increase in the outwardly rectifying component (Fig. 1A, gray trace). The current sensitive to administering cation-free conditions was calculated by numerical subtraction of control currents from currents recorded under cation-free conditions [14]. The I/V relationship of the cation-sensitive current was typical of TASK channels with a strong outward rectification and a reversal potential close to the expected potassium equilibrium potential (Fig. 1B; EK = -104 mV). These findings indicate a strong contribution of K2P9.1 channels to the ISO of thalamocortical (TC) neurons in wildtype mice.

cation-sensitive current was typical of TASK channels with a strong outward rectification and a reversal potential close to the expected potassium equilibrium potential (Fig. 1B; EK = -104 mV). These findings indicate a strong contribution of K2P9.1 channels to the ISO of thalamocortical (TC) neurons in wildtype mice. Figure 1 Whole cell currents recorded from relay neurons in thalamic slice preparations show characteristics indicative for K2P9.1 channels. (A) Mean ramp currents under control recording conditions (ctrl) and after removal of extracellular divalent cations (div. cation free). Inset: Currents were evoked by ramping the membrane potential from -35 mV to -125 mV over 800 ms. (B) I-V relationship of the divalent cation-sensitive current component shows characteristics indicative of TASK channels. I = current; ISO = standing outward current; VM = membrane potential. Neurons from K2P9.1-/- and wildtype animals show no significant differences upon extracellular acidification Sensitivity to extracellular acidification is a hallmark of TASK channels and a reduction of the extracellular pH value can be typically observed under ischemic conditions. In a next experimental step we therefore mimicked cerebral ischemia by lowering the extracellular pH from control conditions (7.35) to 6.0. This maneuver resulted in a significant (p < 0.05) reduction of ISO amplitudes in both genotypes (Fig. 2A). The degree of ISO reduction was not different in wildtype (68.31 ± 9.80%) and K2P9.1-/- neurons (69.92 ± 11.65%; n = 5; p = 0.91; Fig. 2B).

by lowering the extracellular pH from control conditions (7.35) to 6.0. This maneuver resulted in a significant (p < 0.05) reduction of ISO amplitudes in both genotypes (Fig. 2A). The degree of ISO reduction was not different in wildtype (68.31 ± 9.80%) and K2P9.1-/- neurons (69.92 ± 11.65%; n = 5; p = 0.91; Fig. 2B). Figure 2 pH dependence of the standing outward currents in K2P9.1-/- neurons shows no difference compared to wildtype mice (pH 7.35 → pH 6.0). (A) Ramp currents under control conditions (pH 7.35) and after extracellular acidification (pH 6.0) in wildtype (WT; left panel) and K2P9.1-/- mice (right panel). (B) Bar graph representation of current reduction after extracellular pH lowering in wildtype (WT) and K2P9.1-/- mice. Genetic ablation of K2P9.1 channels trends to result in a not significant reduction of stroke development after tMCAO Stroke volumes of wildtype and K2P9.1-/- mice were determined 24 hours after animals subjected to 60 min of tMCAO. Wildtype animals showed stroke volumes of 60.50 ± 17.31 mm3 while K2P9.1-/- mice displayed infarct areas of 47.10 ± 19.26 mm3 (n = 10 and 8; p = 0.23; Fig. 3A and 3B). In accordance with this tendency towards none significantly smaller infarct sizes in K2P9.1-/-, no functionally relevant differences could be found for the Bederson score (WT: 1.83 ± 0.98; K2P9.1-/-: 2.14 ± 0.80; n = 6; p = 0.55; Fig. 3C) and the grip test (WT: 3.17 ± 1.13; K2P9.1-/-: 4.29 ± 0.64; n = 6; p = 0.09; Fig. 3D).

cordance with this tendency towards none significantly smaller infarct sizes in K2P9.1-/-, no functionally relevant differences could be found for the Bederson score (WT: 1.83 ± 0.98; K2P9.1-/-: 2.14 ± 0.80; n = 6; p = 0.55; Fig. 3C) and the grip test (WT: 3.17 ± 1.13; K2P9.1-/-: 4.29 ± 0.64; n = 6; p = 0.09; Fig. 3D). Figure 3 Infarct volumes 24 h after 60 min MCA occlusion in wildtype and K2P9.1-/- mice. (A) Representative TTC-stained images of three corresponding coronal sections of control animals (WT) and K2P9.1-/- mice. (B) Mean brain infarct volumes calculated from (A) (control group: n = 10; K2P9.1-/- mice: n = 8). (C) Mean Bederson score and (D) grip test from the animals shown in (B). ns = not significant.

ice. (A) Representative TTC-stained images of three corresponding coronal sections of control animals (WT) and K2P9.1-/- mice. (B) Mean brain infarct volumes calculated from (A) (control group: n = 10; K2P9.1-/- mice: n = 8). (C) Mean Bederson score and (D) grip test from the animals shown in (B). ns = not significant. Discussion The results of the present study can be summarized as follows: (1) A pH- and divalent cation-sensitive ISO is present in TC neurons of the dLGN. (2) The divalent cation-sensitive component is characterized by outward rectification and a reversal potential close to the potassium equilibrium potential. (3) The ISO of neurons recorded from brain slices of K2P9.1-/- mice and wildtype mice showed comparable pH-sensitivity during extracellular pH changes from 7.35 to 6.0. (4) In a model of cerebral ischemia, K2P9.1-/- animals showed a tendency to reduced infarct volumes 24 hours after undergoing 60 min of tMCAO compared to wildtype controls although these results were not statistically significant. (5) It is concluded that K2P9.1-containing homodimeric and heterodimeric channels significantly contribute to ISO in TC neurons from wildtype mice and that K2P9.1 channels have only a minor impact on infarct volume and motor function following tMCAO compared to other members of the K2P channel family.

ally significant. (5) It is concluded that K2P9.1-containing homodimeric and heterodimeric channels significantly contribute to ISO in TC neurons from wildtype mice and that K2P9.1 channels have only a minor impact on infarct volume and motor function following tMCAO compared to other members of the K2P channel family. Contribution of TASK channel subtypes to ISO in TC neurons During development, the mouse thalamus is characterized by high K2P3.1 gene expression at P0 and displays moderate expression levels throughout postnatal stages [31]. K2P9.1 expression in many thalamic nuclei is rather moderate for all developmental stages but is strong in dLGN from P14 to adult stages. Functional TASK channels can be K2P3.1 homodimers, K2P9.1 homodimers, and K2P3.1/K2P9.1 heterodimers [32-35]. Although K2P3.1 and K2P9.1 show high sequence homology, they differ in their sensitivity to extracellular divalent cations (Mg2+, Ca2+) based on the presence of a glutamate residue at position 70 in K2P9.1 channels [17]. While the conductance of K2P3.1 homodimeric channels is unaffected, the conductance of K2P9.1 homodimeric and K2P3.1/K2P9.1 heterodimeric channels is strongly reduced in the presence of divalent cations [17,33]. Therefore the increase in ISO following removal of extracellular divalent cations which was found in cells from different rodent strain (Long Evans rats, wildtype mice, K2P3.1-/- mice) point to the functional expression of K2P9.1 homodimeric and K2P3.1/K2P9.1 heterodimeric channels in TC neurons.

alent cations [17,33]. Therefore the increase in ISO following removal of extracellular divalent cations which was found in cells from different rodent strain (Long Evans rats, wildtype mice, K2P3.1-/- mice) point to the functional expression of K2P9.1 homodimeric and K2P3.1/K2P9.1 heterodimeric channels in TC neurons. Homodimeric and heterodimeric TASK channels also differ in their pH-sensitivity. While K2P3.1/K2P9.1 channel constructs have a pH-sensitivity (pK approximately 7.3) in the physiological range which is closer to that of K2P3.1 channels (pK approximately 7.5) than K2P9.1 channels (pK approximately 6.8) [34]. In the present study no significant difference was found for the decrease in ISO amplitude when the pH was shifted to a value of 6.0. Therefore the pH- and divalent cation- sensitivities of native TASK-like currents in TC neurons is best represented by K2P3.1/K2P9.1 heterodimeric channels. However, additional modulators (isoflurane, Zn2+, ruthenium red) have to be tested to get more indications for the ratio of homodimeric to heterodimeric TASK channels in these neurons.

divalent cation- sensitivities of native TASK-like currents in TC neurons is best represented by K2P3.1/K2P9.1 heterodimeric channels. However, additional modulators (isoflurane, Zn2+, ruthenium red) have to be tested to get more indications for the ratio of homodimeric to heterodimeric TASK channels in these neurons. The role of TASK channel subtypes in ischemic insults It has been shown before that K2P3.1-/- mice reveal larger tMCAO lesions in comparison to wildtype mice probably through a combination of direct neuronal effects and due to blood pressure/aldosterone effects [11,12]. Based on its physiogical properties and expression pattern, it seemed reasonable to expect an at least similar phenotype of K2P9.1-/- mice compared to K2P3.1-/- mice. The reason for the unexpected results presented here remains unclear but may involve one or more of the following considerations: (1) The cell type-specific expression and the exact conditions of the cellular environment of TASK channels have to be taken into account [36]. (2) Compensatory mechanisms, e.g. upregulation of other K2P channel family members, differences in oxygen sensitivity or yet unknown K2P channel properties may play a role. (2) In GABAergic interneurons of the entorhinal cortex membrane depolarization mediated by inhibition of K2P9.1 channels induce an increase in action potential firing [37]. In consequence, an increase in the release of GABA by interneurons results in a decrease in pyramidal cell activity thereby limiting the injurious effects of ischemia. Assuming that this type of network interaction is found in brain regions affected by tMCAO, the neuroprotection by K2P9.1-/- channels is missing in knock out mice. (4) Gender differences should be taken into account [12]. (5) It should also be kept in mind that ischemic conditions may also affect a variety of other target structures including several ion channel, e.g. TRPV1 or ASICs as well as NMDA receptors [36].

europrotection by K2P9.1-/- channels is missing in knock out mice. (4) Gender differences should be taken into account [12]. (5) It should also be kept in mind that ischemic conditions may also affect a variety of other target structures including several ion channel, e.g. TRPV1 or ASICs as well as NMDA receptors [36]. To unravel the complex scenario of cerebral ischemia and to define the exact functional contribution of a particular K2P channel family member, further pharmacological and genetic tools are warranted, e.g. cell-type specific or conditional K2P3.1-/-, K2P9.1-/- or K2P10.1-/- mice. Especially the development of highly specific channel inhibitors or activators might open up the opportunity to procede these research efforts. Taken together, results from K2P2.1-/- (enhancement of ischemic damage [7]), K2P3.1-/- (increase in infarct volumes [11,12]) and K2P9.1-/- (no significant change ([12] or a tendency towards none significant reduced infarkt volumes: this work)) mice underline the fact that there are differential effects of different K2P channel subtypes on cerebral ischemia, not allowing to reason a uniform influence of this intriguing channel family on stroke formation. Competing interests The authors declare that they have no competing interests.

Taken together, results from K2P2.1-/- (enhancement of ischemic damage [7]), K2P3.1-/- (increase in infarct volumes [11,12]) and K2P9.1-/- (no significant change ([12] or a tendency towards none significant reduced infarkt volumes: this work)) mice underline the fact that there are differential effects of different K2P channel subtypes on cerebral ischemia, not allowing to reason a uniform influence of this intriguing channel family on stroke formation. Competing interests The authors declare that they have no competing interests. Authors' contributions All authors have read and approved the final manuscript. PE, SB, NB and TB performed and analyzed the electrophysiological recordings. CK and TS operated the animals, assessed the functional scores and interpreted the data. HW, CK, TB, SB and SGM conceived the experiments, analyzed data, funded the project and wrote the manuscript. Supplementary Material Additional file 1 Figure S1 - rCBF does not differ between wildtype mice and K2P9.1-/- mice. Determination of regional cerebral blood flow (rCBF) using Laser Doppler flowmetry before the occlusion of the middle cerebral artery (baseline), 10 min after the occlusion (ischemia) and again 10 min after the removal of the filament (reperfusion) in wildtype mice and K2P9.1-/- mice (n = 3/group). No significant differences in rCBF were observed between the two groups. One-way ANOVA, Bonferroni post hoc test. Click here for file

Supplementary Material Additional file 1 Figure S1 - rCBF does not differ between wildtype mice and K2P9.1-/- mice. Determination of regional cerebral blood flow (rCBF) using Laser Doppler flowmetry before the occlusion of the middle cerebral artery (baseline), 10 min after the occlusion (ischemia) and again 10 min after the removal of the filament (reperfusion) in wildtype mice and K2P9.1-/- mice (n = 3/group). No significant differences in rCBF were observed between the two groups. One-way ANOVA, Bonferroni post hoc test. Click here for file Acknowledgements We are grateful to Ms. Melanie Glaser for excellent technical assistance. We are grateful to Douglas A. Bayliss for providing K2P9.1-/- animals. This work was supported by Interdisziplinäres Zentrum für klinische Forschung (IZKF A-54-1) to SGM and HW, the DFG (SFB 581, TP: A10) to SGM and DFG K2P-Forschergruppe 1086 (BU1019/9-1 to TB and ME3283/1-1 to SGM). PE was a member of the Otto-Creutzfeldt-Center for Cognitive and Behavioral Neuroscience Muenster (OCC).

Introduction Prostaglandin E2 (PGE2), formed from arachidonic acid by the action of cyclooxygenases and PGE2 synthases, was previously considered to be a pro-inflammatory prostaglandin, but it may have a much more complex role. PGE2 executes its functions by binding mainly to four membrane-bound G-protein-coupled receptors known as E-prostanoid (EP) 1, 2, 3, and 4. These EP receptors have varied effects on cyclic adenosine monophosphate (cAMP) production, phosphoinositol turnover, and intracellular calcium level regulation [1]. The EP1 receptor increases levels of intracellular calcium [2]. The EP3 receptor, which has several isoforms and consequently couples to several different G-proteins, elicits varied signaling pathways that lead to changes in cAMP levels, calcium mobilization, and activation of phospholipase C [3]. The EP2 and EP4 receptors, which increase the intracellular levels of cAMP, work via a G-protein-coupled mechanism that stimulates adenylyl cyclase [4-6].

to several different G-proteins, elicits varied signaling pathways that lead to changes in cAMP levels, calcium mobilization, and activation of phospholipase C [3]. The EP2 and EP4 receptors, which increase the intracellular levels of cAMP, work via a G-protein-coupled mechanism that stimulates adenylyl cyclase [4-6]. In mice, the EP2 receptor has been reported to be highly expressed in cerebral cortex, striatum, and hippocampus [7,8]. Genetic knockout of this receptor significantly increases lesion volume at the 24-h time point in mice subjected to ischemic paradigms, with no apparent change in behavior [9,10]. It is important to understand the etiopathology of stroke damage and especially its inflammatory cascade over time. Therefore, our goal was to determine whether the anatomical protective effects of EP2 activation are sustained over time (and are not only transient or delayed) and whether such changes correlate with neurologic improvements.

derstand the etiopathology of stroke damage and especially its inflammatory cascade over time. Therefore, our goal was to determine whether the anatomical protective effects of EP2 activation are sustained over time (and are not only transient or delayed) and whether such changes correlate with neurologic improvements. Furthermore, even though previous studies that used EP2 knockout (EP2-/-) mice reported that the presence of EP2 is beneficial in ischemic stroke in vivo, none has demonstrated that EP2 stimulation indeed limits infarct damage. To test this paradigm, we investigated whether the highly selective EP2 agonist ONO-AE1-259-01 could reduce ischemic brain damage in C57BL/6 WT mice. ONO-AE1-259-01 has no detectable affinity to any other prostaglandin receptor [11] and binds with higher affinity to the EP2 receptor (Ki = 3 nM) than does PGE1, 16,16-dimethyl-PGE2, 11-deoxy-PGE1, butaprost, or AH-6809 (Ki = 10, 17, 45, 110, and 350 nM, respectively) [5,6,12]. Materials and methods Animals and Treatments These studies were carried out in male C57BL/6 mice (25 to 30 g) purchased from Charles River Laboratories, Inc (Wilmington, MA). The EP2-/- mouse colony was maintained in the Johns Hopkins animal facility. Animal protocols for these studies were approved by the Johns Hopkins University Animal Care and Use Committee. The animals were allowed free access to water and food before and after surgery. ONO-AE1-259-01 [(16S)-9-deoxy-9beta-chloro-15-deoxy-16-hydroxy-17,17-trimethylene-19,20-didehydro-PGE2 sodium salt] was provided by Ono Pharmaceutical Co. Ltd.

es were approved by the Johns Hopkins University Animal Care and Use Committee. The animals were allowed free access to water and food before and after surgery. ONO-AE1-259-01 [(16S)-9-deoxy-9beta-chloro-15-deoxy-16-hydroxy-17,17-trimethylene-19,20-didehydro-PGE2 sodium salt] was provided by Ono Pharmaceutical Co. Ltd. Cerebral Vessel Diameter and Anatomy To determine the large cerebral vessel gross anatomy in WT and EP2-/- mice, three naïve mice of each genotype were anesthetized deeply and perfused via the heart left ventricle with 5 mL of ice-cold saline followed by 1 mL of black latex paint. Then the mice were decapitated and their brains removed with the circle of Willis intact. The brains were placed in 10% formalin for 24 h before examination with MetaVue software (Meta Imaging Series Software, Downingtown, PA).

the heart left ventricle with 5 mL of ice-cold saline followed by 1 mL of black latex paint. Then the mice were decapitated and their brains removed with the circle of Willis intact. The brains were placed in 10% formalin for 24 h before examination with MetaVue software (Meta Imaging Series Software, Downingtown, PA). Experimental Design and Drug Injection In this study, three sets of experiments were performed. In the first experiment, WT (n = 9) and EP2-/- (n = 14) mice were subjected to 90 min of transient middle cerebral artery occlusion (tMCAO) and 96 h of reperfusion. In the second experiment, WT mice (n = 9/group) were given intracerebroventricular injections of ONO-AE1-259-01 (0.5, 1.0, 2.0 nmol) or vehicle (water) 45-50 min before tMCAO. Briefly, mice were anesthetized and mounted on a stereotaxic frame, the skull was exposed under aseptic conditions, and a hole was drilled according to the coordinates: anteroposterior, 0.5 mm; lateral, 1.0 mm from the bregma; and ventral, 2.5 mm relative to the dura. ONO-AE1-259-01 or vehicle was injected in a volume of 0.2 μL into the right lateral ventricle; the needle was left in place for 10 min before being slowly retracted. Finally, the hole was blocked, and the skin overlying the skull was sutured. Mice were then prepared for tMCAO surgery as described below. Physiologic studies were carried out in a separate cohort of correspondingly treated EP2-/- and ONO-AE1-259-01-treated mice. In the third experiment, WT (n = 8) and EP2-/- (n = 7) mice were subjected to distal permanent middle cerebral artery occlusion (pMCAO).

e were then prepared for tMCAO surgery as described below. Physiologic studies were carried out in a separate cohort of correspondingly treated EP2-/- and ONO-AE1-259-01-treated mice. In the third experiment, WT (n = 8) and EP2-/- (n = 7) mice were subjected to distal permanent middle cerebral artery occlusion (pMCAO). Transient Focal Cerebral Ischemia (tMCAO) and Reperfusion Transient focal cerebral ischemia was induced by occlusion of the middle cerebral artery (MCA) with an intraluminal filament, as described previously [13]. Each mouse was maintained with continuous-flow 1.0-1.5% halothane (after induction with 3.0% halothane) in oxygen-enriched air via a nose cone. The core body temperature (rectal) was maintained at 37.0 ± 0.5°C by a heating pad. No differences in rectal temperature between genotypes were noted before, during, or immediately after ischemia. Relative cerebral blood flow (CBF) was measured by laser-Doppler flowmetry (Moor Instruments, Devon, England) with a flexible fiberoptic probe affixed to the skull over the parietal cortex supplied by the MCA (2 mm posterior and 6 mm lateral to the bregma). Under aseptic conditions, the neck and carotid bifurcation were dissected, and the common carotid artery was temporarily ligated. A 7-0 Ethilon nylon monofilament (Ethicon, Inc., Somerville, NJ) coated with flexible silicone (Cutter Sil light universal hardener, Heraeus Kulzer GmbH, Hanau, Germany) was inserted to occlude the MCA. The filament was advanced through an incision in the external carotid artery stump, through the internal carotid artery to the origin of the MCA; successful occlusion was documented by a decrease in laser-Doppler signal of at least 80%. The filament was left in position for 90 min. During occlusion, the neck was closed with sutures, anesthesia was discontinued, and the animals were transferred to a temperature-controlled chamber to maintain the body temperature at 37.0 ± 0.5°C. At 90 min of occlusion, the mouse was briefly re-anesthetized with halothane, and reperfusion was achieved by slowly withdrawing the filament. After its neck was sutured, the mouse was again placed in the temperature-controlled chamber for 2 h and then returned to its home cage for 4 days. The 4-day time point was selected because it allows maximal survival following the tMCAO.

d with halothane, and reperfusion was achieved by slowly withdrawing the filament. After its neck was sutured, the mouse was again placed in the temperature-controlled chamber for 2 h and then returned to its home cage for 4 days. The 4-day time point was selected because it allows maximal survival following the tMCAO. Permanent Distal Middle Cerebral Artery Occlusion (pMCAO) Permanent MCAO studies were carried out by the method of Majid et al. [14,15] with minor modifications. This permanent distal ischemic protocol was selected because it is highly reproducible and it affects mainly the cortical region, whereas the transient ischemic model causes striatal damage that then extends to the cortical region. The permanent model also allows the study of more distal time points (as described below). Briefly, mice were anesthetized with halothane, and a 1.0-cm vertical skin incision was made between the right eye and ear. The temporal muscle was moved aside to expose the temporal bone. Under a surgical microscope, a 2.0-mm burr hole was drilled over the MCA, transparently visible through the temporal bone. The distal part of the MCA was occluded with a bipolar coagulator, and complete interruption of blood flow at the occlusion site was confirmed by severance of the occlusion site of the MCA. Core body temperature was maintained between 36.5 and 37.5°C during and after the procedure. Animals not circling toward the paretic side after the onset of ischemia and those that developed subarachnoid hemorrhage were eliminated from the study. The successful occlusion was confirmed by placing the laser-Doppler probe above the temporal ridge to establish that blood flow into the region was stopped. At 7 days, mice were euthanized and the brains sectioned. The 7-day time point was selected rather than the 24-h time point because it allows time for extended permanent ischemic damage and can be achieved with 100% survival.

ppler probe above the temporal ridge to establish that blood flow into the region was stopped. At 7 days, mice were euthanized and the brains sectioned. The 7-day time point was selected rather than the 24-h time point because it allows time for extended permanent ischemic damage and can be achieved with 100% survival. Assessment of Neurologic Function and Physiologic Parameters Mice subjected to tMCAO were evaluated for neurologic deficit via a 5-point scale after 96 h of reperfusion. The scores were recorded as: 0, no deficit; 1, forelimb weakness and torso turning to the ipsilateral side when held by the tail; 2, circling to the affected side; 3, unable to bear weight on the affected side; and 4, no spontaneous locomotor activity, according to the protocol used in our previous studies [13,15-17,19]. In a separate cohort, physiologic parameters were measured before occlusion, during occlusion, and during reperfusion. A catheter was inserted into the femoral artery and attached to an automated blood pressure monitor to measure mean arterial blood pressure (MABP). At regular intervals, blood samples were collected through the catheter for analysis of pH, PaO2, and PaCO2.

red before occlusion, during occlusion, and during reperfusion. A catheter was inserted into the femoral artery and attached to an automated blood pressure monitor to measure mean arterial blood pressure (MABP). At regular intervals, blood samples were collected through the catheter for analysis of pH, PaO2, and PaCO2. Mice subjected to pMCAO were tested for neurologic deficits 7 days after occlusion by an experimenter blinded to the mouse genotype according to a 28-point scoring system [15,20]. Since the pMCAO lesion is distinct from that which occurs with the tMCAO model, we have optimized a neurobehavioral test that enables us to look for subtle and reproducible differences between groups. The tests included assessments of body symmetry, gait, climbing, circling behavior, front limb symmetry, compulsory circling, and whisker response. Each test was graded from 0 to 4, establishing a maximum deficit score of 28. Immediately after the testing, the mice were sacrificed for infarct volume analysis.

n groups. The tests included assessments of body symmetry, gait, climbing, circling behavior, front limb symmetry, compulsory circling, and whisker response. Each test was graded from 0 to 4, establishing a maximum deficit score of 28. Immediately after the testing, the mice were sacrificed for infarct volume analysis. Quantification of Infarct Volumes After tMCAO and pMCAO, mice were deeply anesthetized, and brains were harvested and cooled in a deep freezer. Five coronal sections of 2-mm thickness were cut and then incubated in 2% 2,3,5-triphenyl-tetrazolium chloride (TTC) in saline for 20 min at 37°C. The area of infarct, which remains white, was measured on the rostral and caudal surfaces of all five slices and numerically integrated across the thickness of the slice to obtain an estimate of infarct volume (SigmaScan Pro, SPSS Inc. Port Richmond, CA). Infarct volume was converted by multiplying the measured infarct volume by the ratio of the contralateral structure to the ipsilateral structure [21]. Statistics Data, expressed as mean ± SEM, were analyzed by ANOVA and when appropriate, Newman-Keuls multiple range test. Statistical significance was set at P < 0.05.

Quantification of Infarct Volumes After tMCAO and pMCAO, mice were deeply anesthetized, and brains were harvested and cooled in a deep freezer. Five coronal sections of 2-mm thickness were cut and then incubated in 2% 2,3,5-triphenyl-tetrazolium chloride (TTC) in saline for 20 min at 37°C. The area of infarct, which remains white, was measured on the rostral and caudal surfaces of all five slices and numerically integrated across the thickness of the slice to obtain an estimate of infarct volume (SigmaScan Pro, SPSS Inc. Port Richmond, CA). Infarct volume was converted by multiplying the measured infarct volume by the ratio of the contralateral structure to the ipsilateral structure [21]. Statistics Data, expressed as mean ± SEM, were analyzed by ANOVA and when appropriate, Newman-Keuls multiple range test. Statistical significance was set at P < 0.05. Results Comparison of Cerebrovascular Anatomy in WT and EP2-/- Mice We evaluated the gross cerebrovascular anatomy by measuring the large vessel diameters in the brains of WT and EP2-/- mice and found no significant differences between the two genotypes (Figure 1). This observation suggested that there are no obvious changes in blood vessel diameters of these mice under this experimental design. This knowledge made us confident to pursue the following ischemic stroke paradigms.

brains of WT and EP2-/- mice and found no significant differences between the two genotypes (Figure 1). This observation suggested that there are no obvious changes in blood vessel diameters of these mice under this experimental design. This knowledge made us confident to pursue the following ischemic stroke paradigms. Figure 1 Genetic deletion of the EP2 receptor does not significantly alter gross vascular anatomy of the brain. Macroscopic analysis of cerebral arterial vasculature revealed no differences in the circle of Willis or major cerebral arteries between EP2-/- and WT mice (n = 3/group). Effect of EP2 Receptor Genetic Deletion on Monitored Physiologic Parameters of Mice Subjected to tMCAO No significant differences in blood gases (PaO2, PaCO2, pH) were noted between WT and EP2-/- mice at baseline, 1 h after tMCAO, or 1 h after reperfusion (Table 1). The relative CBF dropped more than 80% from baseline after tMCAO and returned to near baseline after reperfusion in WT and EP2-/- mice (Figure 2). Our data suggest an immediate reperfusion after removal of the filament, whereas a previous report documented a slower recovery that could have potentially affected the stroke outcomes [9]. Overall, no significant differences in CBF, body temperature, or MABP were observed between the two genotypes before, during, or after tMCAO. Table 1 Physiologic Parameters in WT and EP2-/- Mice Parameter WT Mice EP2-/- Mice

Effect of EP2 Receptor Genetic Deletion on Monitored Physiologic Parameters of Mice Subjected to tMCAO No significant differences in blood gases (PaO2, PaCO2, pH) were noted between WT and EP2-/- mice at baseline, 1 h after tMCAO, or 1 h after reperfusion (Table 1). The relative CBF dropped more than 80% from baseline after tMCAO and returned to near baseline after reperfusion in WT and EP2-/- mice (Figure 2). Our data suggest an immediate reperfusion after removal of the filament, whereas a previous report documented a slower recovery that could have potentially affected the stroke outcomes [9]. Overall, no significant differences in CBF, body temperature, or MABP were observed between the two genotypes before, during, or after tMCAO. Table 1 Physiologic Parameters in WT and EP2-/- Mice Parameter WT Mice EP2-/- Mice Baseline 1 h MCAO 1 h Reperfusion Baseline 1 h MCAO 1 h Reperfusion pH 7.37 ± 0.02 7.34 ± 0.02 7.35 ± 0.01 7.32 ± 0.01 7.33 ± 0.01 7.33 ± 0.01 PaCO2 38.2 ± 1.4 40.0 ± 1.5 38.7 ± 1.5 42.5 ± 1.3 41.2 ± 2.2 41.4 ± 1.0 PaO2 113 ± 6 131 ± 3 121 ± 9 127 ± 3 124 ± 5 120 ± 4 Measurements were made in arterial blood samples obtained via femoral catheter. WT, n = 4; EP2-/-, n = 5.

h Reperfusion pH 7.37 ± 0.02 7.34 ± 0.02 7.35 ± 0.01 7.32 ± 0.01 7.33 ± 0.01 7.33 ± 0.01 PaCO2 38.2 ± 1.4 40.0 ± 1.5 38.7 ± 1.5 42.5 ± 1.3 41.2 ± 2.2 41.4 ± 1.0 PaO2 113 ± 6 131 ± 3 121 ± 9 127 ± 3 124 ± 5 120 ± 4 Measurements were made in arterial blood samples obtained via femoral catheter. WT, n = 4; EP2-/-, n = 5. Figure 2 Genetic deletion of the EP2 receptor does not affect physiologic parameters. Relative cerebral blood flow (CBF, A), core body temperature (B), and mean arterial blood pressure (MABP, C) were recorded at baseline, at induction of ischemia, and at 15-min intervals during ischemia and 1 h of reperfusion in WT and EP2-/- mice (n = 4 WT and 5 EP2-/-). The change in CBF was recorded as a percent of baseline. Effect of EP2 Receptor Genetic Deletion on Neurologic Scores and Infarct Volume after tMCAO The neurologic deficit scores of EP2-/- mice were significantly (P < 0.01) higher than those of WT mice after 90-min tMCAO and 4-day reperfusion (Figure 3). Furthermore, according to the TTC-staining method, EP2-/- mice had a significantly larger mean infarct size (P < 0.01) than that of their WT counterparts after tMCAO (Figure 3). The percent mortality was estimated at 36% in EP2-/- mice (14 of 22 survived) and 25% in WT mice (9 of 12 survived).

y reperfusion (Figure 3). Furthermore, according to the TTC-staining method, EP2-/- mice had a significantly larger mean infarct size (P < 0.01) than that of their WT counterparts after tMCAO (Figure 3). The percent mortality was estimated at 36% in EP2-/- mice (14 of 22 survived) and 25% in WT mice (9 of 12 survived). Figure 3 EP2 receptor deletion increases neurologic deficit scores and infarct volume in mice subjected to tMCAO. Mice were subjected to 90-min tMCAO and evaluated for neurologic deficits at 4 days. After being scored, the mice were sacrificed, and brain infarction was estimated by TTC staining. (A) Representative photographs show infarcted brain slices from WT (left) and EP2-/- (right) mice. (B) The bar graph shows the corrected cortical, striatal, and hemispheric infarct volumes of WT and EP2-/- mice. The infarct size was significantly larger in EP2-/- (n = 14) than in WT mice (n = 9). (C) Neurologic deficit scores at 4 days after ischemia were significantly higher in EP2-/- mice than in WT mice, indicating more neurologic dysfunction. **P < 0.01.

cal, striatal, and hemispheric infarct volumes of WT and EP2-/- mice. The infarct size was significantly larger in EP2-/- (n = 14) than in WT mice (n = 9). (C) Neurologic deficit scores at 4 days after ischemia were significantly higher in EP2-/- mice than in WT mice, indicating more neurologic dysfunction. **P < 0.01. Effect of the EP2-selective Agonist ONO-AE1-259-01 on Physiologic Parameters of Mice Subjected to tMCAO The blood gas (pH, PaCO2, PaO2) concentrations in vehicle- and ONO-AE1-259-01-treated groups (n = 5/group) remained within normal physiologic ranges, and no significant differences were measured between the groups. Similarly, the MABP did not differ significantly between the groups (Table 2). As estimated by laser-Doppler flowmetry, the relative CBF rapidly decreased to more than 80% below baseline in both vehicle- and ONO-AE1-259-01-treated groups. The percent reduction in CBF did not differ significantly between the groups. Table 2 Effect of ONO-AE1-259-01 on Physiologic Parameters

Effect of the EP2-selective Agonist ONO-AE1-259-01 on Physiologic Parameters of Mice Subjected to tMCAO The blood gas (pH, PaCO2, PaO2) concentrations in vehicle- and ONO-AE1-259-01-treated groups (n = 5/group) remained within normal physiologic ranges, and no significant differences were measured between the groups. Similarly, the MABP did not differ significantly between the groups (Table 2). As estimated by laser-Doppler flowmetry, the relative CBF rapidly decreased to more than 80% below baseline in both vehicle- and ONO-AE1-259-01-treated groups. The percent reduction in CBF did not differ significantly between the groups. Table 2 Effect of ONO-AE1-259-01 on Physiologic Parameters Parameter Vehicle 0.5 nmol 1 nmol 2 nmol Pre-ischemia pH 7.31 ± 0.02 7.28 ± 0.47 7.30 ± 0.46 7.25 ± 0.23 PaCO2 43.2 ± 1.8 44.4 ± 2.0 45.2 ± 1.6 43.0 ± 1.3 PaO2 133 ± 5 139 ± 5 130 ± 8 134 ± 9 MABP 78.6 ± 3.9 75.6 ± 1.7 77.4 ± 2.8 72.8 ± 2.5 Ischemia pH 7.28 ± 0.02 7.29 ± 0.64 7.26 ± 0.03 7.33 ± 0.02 PaCO2 44.0 ± 1.8 43.2 ± 1.5 45.6 ± 2.6 43.16 ± 1.6 PaO2 129 ± 5 133 ± 4 124 ± 6 133 ± 5 MABP 74.4 ± 1.7 76.6 ± 2.5 74.4 ± 1.8 74.8 ± 1.7 Reperfusion pH 7.26 ± 0.03 7.28 ± 0.05 7.29 ± 0.40 7.28 ± 0.25 PaCO2 43.0 ± 1.8 45.8 ± 1.4 47.2 ± 1.3 44.8 ± 2.7 PaO2 135 ± 9 129 ± 6 126 ± 7 138 ± 9 MABP 74.6 ± 1.3 73.8 ± 1.7 76.0 ± 1.6 74.0 ± 1.6 Measurements were made in arterial blood samples obtained via femoral catheter (n = 5/group).

± 2.5 74.4 ± 1.8 74.8 ± 1.7 Reperfusion pH 7.26 ± 0.03 7.28 ± 0.05 7.29 ± 0.40 7.28 ± 0.25 PaCO2 43.0 ± 1.8 45.8 ± 1.4 47.2 ± 1.3 44.8 ± 2.7 PaO2 135 ± 9 129 ± 6 126 ± 7 138 ± 9 MABP 74.6 ± 1.3 73.8 ± 1.7 76.0 ± 1.6 74.0 ± 1.6 Measurements were made in arterial blood samples obtained via femoral catheter (n = 5/group). ONO-AE1-259-01 Attenuates Neurologic Dysfunction and Infarct Volume in Mice Subjected to tMCAO Although we and others have shown that deletion of EP2 is detrimental to stroke outcome, here we wanted to test whether the use of a selective agonist would provide protection. The neurologic deficits in the groups administered 1.0 and 2.0 nmol of ONO-AE1-259-01 before tMCAO were significantly less severe than those of the vehicle-treated group; no difference was observed in the group given the lowest dose of 0.5 nmol (Figure 4). Furthermore, the ONO-AE1-259-01-treated groups had significantly (P < 0.05) attenuated hemispheric infarct volumes at the 1.0- and 2.0-nmol doses compared with the vehicle-treated group, but the lowest dose did not reach significance (Figure 4). In mice treated with ONO-AE1-259-01, the percent mortality was estimated at 30% at the dose of 0.05 nmol (9 of 13 survived), 30% at 1.0 nmol (9 of 13 survived), and 25% at 2.0 nmol (9 of 12 survived) compared with 36% in the vehicle-treated group (9 of 14 survived).

he lowest dose did not reach significance (Figure 4). In mice treated with ONO-AE1-259-01, the percent mortality was estimated at 30% at the dose of 0.05 nmol (9 of 13 survived), 30% at 1.0 nmol (9 of 13 survived), and 25% at 2.0 nmol (9 of 12 survived) compared with 36% in the vehicle-treated group (9 of 14 survived). Figure 4 Effect of pretreatment with ONO-AE1-259-01 on neurologic deficit scores and infarct volume after 4 days of reperfusion. Mice were pretreated with ONO-AE1-259-01 before being subjected to 90-min MCAO and 4 days of reperfusion. (A) Images of representative TTC-stained brain sections. (B) Percent corrected cortical, striatal, and hemispheric infarct volumes. (C) Neurologic deficit scores were significantly lower in mice treated with 1.0 and 2.0 nmol ONO-AE1-259-01 than in vehicle-treated mice. n = 9/group; *P < 0.05 compared with the vehicle-treated group.

representative TTC-stained brain sections. (B) Percent corrected cortical, striatal, and hemispheric infarct volumes. (C) Neurologic deficit scores were significantly lower in mice treated with 1.0 and 2.0 nmol ONO-AE1-259-01 than in vehicle-treated mice. n = 9/group; *P < 0.05 compared with the vehicle-treated group. Effect of EP2 Receptor Genetic Deletion on Neurologic Deficits and Infarct Volume in Mice Subjected to pMCAO To address whether ischemic outcomes could differ in different ischemic stroke models, we compared outcomes obtained from the transient ischemic reperfusion model (as represented in Figure 3) with those of the permanent ischemic model. The transient model is characterized by an ischemic-reperfusion injury that results consistently in damage that begins mostly in the striatum and then spreads especially to the surrounding cortical region. In contrast, the permanent distal occlusion model does not have the reperfusion injury component, and it results in a brain lesion mainly constrained within the cortex. Thus, these two models are complementary. In addition to addressing different ischemic stroke paradigms, their outcomes are also regionally dependent. Finally, whereas others have studied early time points after stroke, namely 24 h [10], we assessed outcomes at 7 days after pMCAO. We found that at 7 days after pMCAO, EP2-/- mice suffered significantly greater neurologic deficits than did WT mice (P < 0.04) and had significantly larger infarct volumes than did their WT counterparts (P < 0.008; Figure 5). No mortality was observed after distal permanent focal cerebral ischemia in either genotype.

We found that at 7 days after pMCAO, EP2-/- mice suffered significantly greater neurologic deficits than did WT mice (P < 0.04) and had significantly larger infarct volumes than did their WT counterparts (P < 0.008; Figure 5). No mortality was observed after distal permanent focal cerebral ischemia in either genotype. Figure 5 EP2 receptor deletion increases neurologic deficit scores and infarct volume of mice subjected to pMCAO. Seven days after being subjected to pMCAO, mice were assessed for neurologic dysfunction and then sacrificed. (A) Representative images show TTC-stained infarcted brain slices from WT (left) and EP2-/- (right) mice. (B) The bar graph shows the corrected cortical infarct volumes of WT and EP2-/- mice. (C) Neurologic deficit scores at 7 days after ischemia were significantly higher in EP2-/- mice (n = 7) than in WT mice (n = 8). **P < 0.008.

ative images show TTC-stained infarcted brain slices from WT (left) and EP2-/- (right) mice. (B) The bar graph shows the corrected cortical infarct volumes of WT and EP2-/- mice. (C) Neurologic deficit scores at 7 days after ischemia were significantly higher in EP2-/- mice (n = 7) than in WT mice (n = 8). **P < 0.008. Discussion This study was designed to further ascertain the unique neuroprotective properties of the PGE2 EP2 receptor in ischemic stroke. Using EP2-/- mice, we showed that the EP2 receptor is protective in cortical and striatal brain regions affected by tMCAO and in the cortex after pMCAO but that its deletion does not affect the gross cerebrovascular anatomy or the physiologic parameters of blood pH, PaO2, PaCO2, or MABP. We also demonstrated that this protective action of EP2 receptors is not simply transient or acute, as previously suggested, but lasts for days. Furthermore, pretreatment of mice with 1.0- and 2.0-nmol doses of the highly selective EP2 receptor agonist ONO-AE1-259-01 significantly reduced the infarct volume induced by transient ischemia, again without affecting the physiologic parameters monitored. These results clearly demonstrate that pharmacologic stimulation of the EP2 receptors in the brain attenuates brain damage caused by cerebral ischemia and support the hypothesis that EP2 receptor stimulation is important throughout an extended duration of the pathophysiologic response to brain ischemic damage. Measurements of relative CBF showed that tMCAO caused a reduction in cortical perfusion throughout the ischemic period that was similar in the EP2-/- and drug-treated mice to that in their corresponding controls, suggesting that neither genetic deletion nor pharmacologic activation of the EP2 receptor affected the severity of the ischemic insult. Thus, our findings indicate a protective role for EP2 by mechanisms that are likely other than those involving cerebrovascular effects. Moreover, no significant differences were observed in body temperature or MABP.

deletion nor pharmacologic activation of the EP2 receptor affected the severity of the ischemic insult. Thus, our findings indicate a protective role for EP2 by mechanisms that are likely other than those involving cerebrovascular effects. Moreover, no significant differences were observed in body temperature or MABP. EP2-/- mice that underwent tMCAO or pMCAO had greater neurologic disability and infarct size than did WT mice. Although compensatory pathways might occur in the knockout animals, we show here that WT mice that received the two higher doses of the EP2 agonist ONO-AE1-259-01 before tMCAO had significantly less severe neurologic deficits and less infarct damage, supporting observations that EP2 receptors are beneficial in excitotoxicity [8] and in ischemic stroke [9].

ccur in the knockout animals, we show here that WT mice that received the two higher doses of the EP2 agonist ONO-AE1-259-01 before tMCAO had significantly less severe neurologic deficits and less infarct damage, supporting observations that EP2 receptors are beneficial in excitotoxicity [8] and in ischemic stroke [9]. Previous reports have shown that the EP2 receptor elicits neuroprotective effects under conditions of excitotoxicity and oxygen-glucose deprivation by increasing intracellular levels of cAMP and activating PKA signaling [9,10]. We have also previously reported that pharmacologic activation of EP2 receptors leads to neuroprotection via the cAMP-PKA pathway [22]. Moreover, the pharmacologic stimulation of EP4/EP3 receptors affords protection by increasing intracellular levels of cAMP and through activation of the ERK pathway [13]. These cAMP cascades can provide protection for example by: (1) reducing the release of endoplasmic Ca2+ through the inositol triphosphate receptor [23], (2) inhibiting the expression of adhesion molecules [24], (3) suppressing the activity of neuronal nitric oxide synthase [25], (4) stimulating cAMP response element binding (CREB) [26], and (5) stimulating the high-affinity glutamate transporter [27]. Interestingly, ONO-AE1-259-01 has been suggested to elevate cAMP and inhibit expression of inflammatory molecules such as ICAM-1 and B7.2 (CD86) [28,29], and such inflammatory molecules affect stroke outcomes [30,31]. The exact cascade in vivo is likely to be much more elaborate than what we can begin to address in isolated neuronal cultures, especially considering the complexity of the different cells and interactions between them and their environment (blood flow, oxygenation, inflammation, etc) over time. The signaling pathways of these molecules and kinases are continually expanding (kinome-phosphorylome projects [32]), with many branches that link to other pathways. Therefore, the complete cascade of events that takes place within any given cell based on its location with respect to the infarct is likely to differ substantially. Indeed, it is highly probable that the neuroprotection provided by EP2 stimulation results from a combination of pathways rather than a single one. For these reasons, we have first focused here on demonstrating that in the brain, stimulating the EP2 receptors leads to anatomical and behavioral protection.

ubstantially. Indeed, it is highly probable that the neuroprotection provided by EP2 stimulation results from a combination of pathways rather than a single one. For these reasons, we have first focused here on demonstrating that in the brain, stimulating the EP2 receptors leads to anatomical and behavioral protection. To build on our work with the tMCAO model, we are now endeavoring to confirm that the EP2-selective agonist is also protective in the pMCAO model and to address potential targets/biosystems that could begin to explain some of the steps leading to neuroprotection.

ubstantially. Indeed, it is highly probable that the neuroprotection provided by EP2 stimulation results from a combination of pathways rather than a single one. For these reasons, we have first focused here on demonstrating that in the brain, stimulating the EP2 receptors leads to anatomical and behavioral protection. To build on our work with the tMCAO model, we are now endeavoring to confirm that the EP2-selective agonist is also protective in the pMCAO model and to address potential targets/biosystems that could begin to explain some of the steps leading to neuroprotection. Two previous studies have indicated that activation of the PGE2 EP2 receptor can protect against excitotoxic and anoxic injury [9,10]. In one of those studies, the tMCAO was followed by only 22.5 h of reperfusion, and no behavioral outcomes were reported. It is important to document that such ischemic-reperfusion-related change in the knockout animal is not transient or delayed, and that it is indeed maintained at later time points. When a brief ischemic event is followed by reperfusion, a second phase of injury occurs (potentially mediated by a surge in inflammatory markers). What's more, over time, endogenous repair pathways can be activated. Although it is more challenging to keep a mouse alive for 96 h after transient ischemia, we selected this extended reperfusion time because it enables us to document that the protective role of EP2 is sustained during the entire period. In the other study, Liu et al. first tested the stroke outcomes (without behavioral outcomes) in mice subjected to pMCAO after only 24 h of survival [10] and then a subsequent study suggested that misoprostol (which is a poorly selective mouse EP2 receptor agonist) has a protective effect against MCAO injury [33]. For reasons similar to those described above, it is important to determine the potential contributions of EP2 to neurobehavioral and ischemic outcomes at later time points; that is why in our study we selected 7 days. Both of these previous studies led to the suggestion that activation of the PGE2 EP2 receptor can protect against ischemic injury, although no data were provided to support this hypothesis. To make such a conclusion, one needs to test whether selectively targeting EP2 would result in neuroprotection. Furthermore, it has been suggested that compensatory mechanisms can occur in knockout animals. Therefore, we chose to directly activate the EP2 receptor with a selective agonist to complement the findings in knockout mice.

conclusion, one needs to test whether selectively targeting EP2 would result in neuroprotection. Furthermore, it has been suggested that compensatory mechanisms can occur in knockout animals. Therefore, we chose to directly activate the EP2 receptor with a selective agonist to complement the findings in knockout mice. Based on these novel observations, we can conclude that pharmacologic stimulation of the EP2 receptor with a selective pharmacologic agent could potentially be used therapeutically in translational medicine (most likely in combination with other standard treatments) to limit brain damage following ischemic stroke. Competing interests The authors declare that they have no competing interests. Authors' contributions SD conceived, designed and coordinated the study; MA and SS participated in the design, performed the experiments and analyzed the data; ZA assisted with the preclinical model; TM and SN contributed the animals, the drugs and reviewed the paper. All authors read and approved the manuscript. Acknowledgements This study was supported by NIH grants (NS046400, AG022971). We thank Claire Levine for assistance in the preparation of the manuscript and all Doré lab team members for their active participation.

Introduction The implementation of transgenic mice has also revolutionized the field of experimental stroke research in that the effects of distinct genes on stroke outcome can be easily assessed. Most of transgenic mice originate from C57Bl/6 or Sv/129 inbred strains. Moreover, these strains are also commonly used as "wild-type controls". However, there are considerable strain-related differences in the susceptibility to cerebral ischemia. Variations in cerebrovascular anatomy and hemodynamics as well as sensitivity to excitotoxicity have been identified as underlying reasons [1-7]. This may become of particular relevance when transgenic mice with mixed genetic background are compared to purebred controls. C57Bl/6 mice have been shown to be more susceptible to ischemic injury compared to Sv/129 mice in models of global (forebrain) ischemia [8] and to develop larger brain infarctions in permanent middle cerebral artery occlusion (pMCAO) [2,9]. Whether these findings also extend to the most frequently used murine stroke model, i.e. transient middle cerebral artery occlusion (tMCAO) with an intraluminal thread, is not known. Furthermore, important functional parameters of acute ischemic brain damage such as cerebral perfusion or cytotoxic edema formation have not been systematically compared between these two popular strains until now.

model, i.e. transient middle cerebral artery occlusion (tMCAO) with an intraluminal thread, is not known. Furthermore, important functional parameters of acute ischemic brain damage such as cerebral perfusion or cytotoxic edema formation have not been systematically compared between these two popular strains until now. We recently introduced multimodal magnetic resonance imaging (MRI) at high-field strength of 17.6 Tesla as a powerful tool to non-invasively analyze the early phase of cerebral ischemia and the evolution of infarctions in individual mice over time [10]. In the present study, C57Bl/6 and Sv/129 mice were subjected to 60 min tMCAO. MRI outcome measures were chosen to characterize cerebral perfusion by continuous arterial spin labeling (CASL), hypoxic diffusion restriction by diffusion-weighted imaging (DWI), and irreversible infarction by lesion extent on T2-w and quantitative T2 relaxometric images. Methods Experimental design and animal stroke model All procedures and animal studies were approved by the appropriate local authorities (Regierung von Unterfranken, Wuerzburg, Germany) and conducted in accordance with recommendations for the performance of basic experimental stroke studies [11]. The experimental groups consisted of 8 6-8-weeks-old male C57Bl/6 mice (Charles River, Sulzfeld, Germany) and 8 6-8-weeks-old male Sv/129 mice (Harlan-Winkelmann, Borchen, Germany) weighing 20-25 g. Animals of each group underwent tMCAO with 60 min occlusion time.

performance of basic experimental stroke studies [11]. The experimental groups consisted of 8 6-8-weeks-old male C57Bl/6 mice (Charles River, Sulzfeld, Germany) and 8 6-8-weeks-old male Sv/129 mice (Harlan-Winkelmann, Borchen, Germany) weighing 20-25 g. Animals of each group underwent tMCAO with 60 min occlusion time. tMCAO was performed as described in detail previously [12-14]. Briefly, a standardized suture coated with silicon rubber (6021PK10; Doccol Corporation, Redlands, CA, USA) was introduced into the right common carotid artery and advanced via the internal carotid artery to the origin of the middle cerebral artery (MCA). The suture was fixed and left in situ and animals were allowed to recover. Operation time per animal did not exceed 15 min. After 60 min, animals were re-anesthetized and the suture was withdrawn to allow tissue reperfusion (tMCAO). The operations were performed under inhalation anesthesia (2.0% isoflurane in a 70%/30% N2O/O2 mixture) and the body temperature was maintained at 37°C using a servo-controlled heating pad. All animals were subsequently followed in-vivo by serial multimodal ultra-high field MRI at 2 h and 24 h.

sue reperfusion (tMCAO). The operations were performed under inhalation anesthesia (2.0% isoflurane in a 70%/30% N2O/O2 mixture) and the body temperature was maintained at 37°C using a servo-controlled heating pad. All animals were subsequently followed in-vivo by serial multimodal ultra-high field MRI at 2 h and 24 h. Multimodal ultra-high field MRI of experimental cerebral ischemia in-vivo The detailed description of the imaging protocol can be found elsewhere [10]. A short summary of relevant parameters of the employed pulse sequences is given here. Cerebral perfusion was measured using a modified arterial spin labeling (CASL) method [15-17]. To benefit especially from increased longitudinal magnetization and the elevation of the T1 relaxation time for detailed anatomical mapping of CBF and group analysis, all measurements were performed at ultra-high field strength (17.6 T, 750 Hz, Biospin, Bruker BioSpin GmbH, Ettlingen, Germany). Image maps of cerebral perfusion were calculated on a pixel-by-pixel basis according to Detre et al. [15]. The degree of the inversion efficiency was assumed to be alpha = 0.7 [18,19], and the brain-blood partition coefficient value for water lambda = 0.95 mL/g [20]. Parameters for the fast spin-echo imaging sequence (RARE) were: echo train length (ETL) = 8, effective echo time TEeff = 17.2 ms, repetition time TR = 1 s, slice thickness 1.5 cm, FOV 2.5 × 2.5 cm, matrix of 64 × 64 voxels. The signal was averaged over 12 repetitions resulting in a total acquisition time of 9.5 min.

]. Parameters for the fast spin-echo imaging sequence (RARE) were: echo train length (ETL) = 8, effective echo time TEeff = 17.2 ms, repetition time TR = 1 s, slice thickness 1.5 cm, FOV 2.5 × 2.5 cm, matrix of 64 × 64 voxels. The signal was averaged over 12 repetitions resulting in a total acquisition time of 9.5 min. DWI was performed with a pulsed-field gradient Setjskal-Tanner-like multislice spin echo sequence with diffusion sensitization along the slice direction [21]. Images with different b-values, 0 and 800 s/mm2, were acquired to allow for the calculation of apparent diffusion coefficient (ADC) maps of brain water. Whole brain coverage was achieved by thirteen coronal slices acquired with a matrix size of 64 × 64, FOV 2.5 × 2.5 cm, in plane resolution 282 × 282 μm, slice thickness = 0.5 mm, interslice distance = 1 mm, TE/TR = 22.3/2000 ms. Repeated measurement of the b = 800 s/mm2 DWI experiments (number of repetition NR = 3) resulted at an overall acquisition time for diffusion weighted experiments of 8 min. ADC maps were calculated by applying the common equation ADC = -0.00125 × ln (SIB800/SIB0).

5 mm, interslice distance = 1 mm, TE/TR = 22.3/2000 ms. Repeated measurement of the b = 800 s/mm2 DWI experiments (number of repetition NR = 3) resulted at an overall acquisition time for diffusion weighted experiments of 8 min. ADC maps were calculated by applying the common equation ADC = -0.00125 × ln (SIB800/SIB0). T2 relaxometric mapping was performed for the in-vivo delineation of infarcted brain tissue at 24 h. Single slice T2-w imaging was performed using a Carr-Purcell-Meiboom-Gill (CPMG) multi-spin echo sequence collecting 32 echoes at TR/TE = 4.2/2000 ms. T2 relaxation times constants were calculated voxel-wise by fitting the intensities of the 20 first echoes to a monoexponential model. In addition, a strongly T2 weighted 2D turbo spin-echo sequence was acquired for high resolution anatomical imaging (RARE factor 16, TR = 8 s, effective TE = 56.44 ms, 2 averages, 13 coronal slices with an image matrix of 128 × 128 were acquired, FOV = 2.5 cm × 2.5 cm, slice thickness = 0.5 mm, interslice distance = 1 mm, overall acquisition time of 1 min). The center position of the 1.5 mm slab for measurement of CBF, T1 and T2 relaxometric maps each with the exact same geometry was centered at the bregma as the operational definition of the central MCA territory. At the host console measurements and data processing were performed with the ParaVision software (version 3.02, Bruker BioSpin GmbH, Ettlingen, Germany). Further image calculation and fitting procedures were done using MATLAB® (The Mathworks Inc., Natick, MA, USA).

T2 relaxometric mapping was performed for the in-vivo delineation of infarcted brain tissue at 24 h. Single slice T2-w imaging was performed using a Carr-Purcell-Meiboom-Gill (CPMG) multi-spin echo sequence collecting 32 echoes at TR/TE = 4.2/2000 ms. T2 relaxation times constants were calculated voxel-wise by fitting the intensities of the 20 first echoes to a monoexponential model. In addition, a strongly T2 weighted 2D turbo spin-echo sequence was acquired for high resolution anatomical imaging (RARE factor 16, TR = 8 s, effective TE = 56.44 ms, 2 averages, 13 coronal slices with an image matrix of 128 × 128 were acquired, FOV = 2.5 cm × 2.5 cm, slice thickness = 0.5 mm, interslice distance = 1 mm, overall acquisition time of 1 min). The center position of the 1.5 mm slab for measurement of CBF, T1 and T2 relaxometric maps each with the exact same geometry was centered at the bregma as the operational definition of the central MCA territory. At the host console measurements and data processing were performed with the ParaVision software (version 3.02, Bruker BioSpin GmbH, Ettlingen, Germany). Further image calculation and fitting procedures were done using MATLAB® (The Mathworks Inc., Natick, MA, USA). During MRI measurements, mice were anesthesized by 2% isoflurane in medical air (21%). The respiratory rate was monitored using an air-balloon positioned ventrally underneath the mouse body. The body temperature was constantly measured on the body surface and actively maintained at 37°C.

At the host console measurements and data processing were performed with the ParaVision software (version 3.02, Bruker BioSpin GmbH, Ettlingen, Germany). Further image calculation and fitting procedures were done using MATLAB® (The Mathworks Inc., Natick, MA, USA). During MRI measurements, mice were anesthesized by 2% isoflurane in medical air (21%). The respiratory rate was monitored using an air-balloon positioned ventrally underneath the mouse body. The body temperature was constantly measured on the body surface and actively maintained at 37°C. Statistical and image analysis The extraction of brain tissue from the scalp and skull was done by manual segmentation for each subject and time point. Packages from the FMRIB Software Library FSL (version 4.1) [22] were used for motion correction, registration (FLIRT) [23], and statistical image analysis. Intra-subject linear alignment and registration to a common standard template [24] was achieved by a step-wise affine procedure with six degrees of freedom.

point. Packages from the FMRIB Software Library FSL (version 4.1) [22] were used for motion correction, registration (FLIRT) [23], and statistical image analysis. Intra-subject linear alignment and registration to a common standard template [24] was achieved by a step-wise affine procedure with six degrees of freedom. For quantitative group comparisons, selected regions of interest (ROIs) were delineated in atlas space: 1) the cerebral cortex in the center of the MCA territory and 2) the subcortex including the ipsilateral caudoputamen and pyramidal tract (Figure 1). Statistical analysis of ROIs was done by non-parametric pairwise comparisons using the Wilcoxon matched-pairs signed-rank test. The risk of cerebral infarction was determined on within-group probability maps calculated by averaging the binary segmentations of healthy versus infarcted brain tissue within-subject. Automated binary segmentation of cerebral infarction was performed by applying a threshold of 34 ms T2 relaxation time on the T2 relaxometric maps at 24 h as demonstrated previously [10]. Manual input was given only for the removal of intraventricular cerebrospinal fluid. Finally, whole-brain volumetric analysis of infarction was done by manual segmentation on the images of the T2-w RARE sequence. Values are always given as mean ± standard error of the mean (SEM).

t 24 h as demonstrated previously [10]. Manual input was given only for the removal of intraventricular cerebrospinal fluid. Finally, whole-brain volumetric analysis of infarction was done by manual segmentation on the images of the T2-w RARE sequence. Values are always given as mean ± standard error of the mean (SEM). Figure 1 Region-of-interest masks. On the left the central cortical territory of the middle cerebral artery (MCA) is delineated by the white overlay region according to the corresponding coordinates in atlas space, and the deep subcortical territory of the MCA is displayed on the right. Results Manual whole-brain segmentation of infarction Volumetric extents of cerebral infarction as delineated manually on T2-w RARE imaging are given as volume ratios (infarction/hemisphere) and were as follows for the C57Bl/6 mice: at 2 h 0.1 ± 0.03 and at 24 h 0.43 ± 0.02. For the Sv/129 mice, values were 0.06 ± 0.03 at 2 h and 0.37 ± 0.03 at 24 h. No statistical differences were found when comparing pairwise-matched groups (Figure 2). Figure 2 Group means of infarct volume ratios (infarction/hemisphere). Pairwise comparisons between 8 C57Bl/6 and Sv/129 mice did not reveal any statistical differences for the 2 h (p = 0.204) or 24 h time point (p = 0.172) after tMCAO. Error bars denote standard errors of the mean.

Results Manual whole-brain segmentation of infarction Volumetric extents of cerebral infarction as delineated manually on T2-w RARE imaging are given as volume ratios (infarction/hemisphere) and were as follows for the C57Bl/6 mice: at 2 h 0.1 ± 0.03 and at 24 h 0.43 ± 0.02. For the Sv/129 mice, values were 0.06 ± 0.03 at 2 h and 0.37 ± 0.03 at 24 h. No statistical differences were found when comparing pairwise-matched groups (Figure 2). Figure 2 Group means of infarct volume ratios (infarction/hemisphere). Pairwise comparisons between 8 C57Bl/6 and Sv/129 mice did not reveal any statistical differences for the 2 h (p = 0.204) or 24 h time point (p = 0.172) after tMCAO. Error bars denote standard errors of the mean. Probability maps of infarction by quantitative T2 thresholds Automated segmentation of infarction for each individual animal was performed with a threshold of a quantitative T2 value of 34 ms as described previously [10]. The probabilities of infarction in atlas space are given in Table 1 for both groups, time points and corresponding subcortical and cortical ROIs. There was no significant difference in the probability of infarction between both groups at 2 h (p = 0.4) or 24 h (p = 0.93). The spatial distribution of the different probabilities of infarction is demonstrated on color-coded probability maps and relaxometric T2 maps (Figure 3). Interestingly, subcortical infarctions (basal ganglia) were present already 2 h after stroke onset corresponding to the insufficient collateral blood supply in these areas. After 24 h infarction has extended to the whole MCA territory including the cortex.

lor-coded probability maps and relaxometric T2 maps (Figure 3). Interestingly, subcortical infarctions (basal ganglia) were present already 2 h after stroke onset corresponding to the insufficient collateral blood supply in these areas. After 24 h infarction has extended to the whole MCA territory including the cortex. Figure 3 Color coded group means of probability of infarction and quantitative T2 values. For both groups probability of cortical and subcortical infarction was similar. Note that subcortical infarction is evident on T2 maps already 2 h after tMCAO. The segmentation threshold for infarction was set to 34 ms. Table 1 Group means of infarct probabilities (%). 129/Sv C57BL/6 2 h 24 h 2 h 24 h Cortex 53.3 ± 10.1 92.1 ± 2.5 60.7 ± 10.4 95.1 ± 3.1 Subcortex 67.8 ± 12.8 100 79.1 ± 11.3 100 Values are given for both cortical and subcortical ROIs and both time points after tMCAo. Results of statistical pairwise comparisons are given in the text and did not significantly differ between groups for any time point. Standard errors of the mean are denoted by ±. Analysis of functional ischemic outcome parameters Pairwise comparisons between group means of quantitative CBF values were not significant for any time point or region of interest. Two-sided p-values for matched pairwise comparisons between C57BL/6 and Sv/129 mice were: subcortical ROI at 2 h p = 0.7 and p = 0.53 at 24 h; cortical ROI at 2 h p = 0.64 and p = 0.86 at 24 h.

wise comparisons between group means of quantitative CBF values were not significant for any time point or region of interest. Two-sided p-values for matched pairwise comparisons between C57BL/6 and Sv/129 mice were: subcortical ROI at 2 h p = 0.7 and p = 0.53 at 24 h; cortical ROI at 2 h p = 0.64 and p = 0.86 at 24 h. Similarly, ADC values in the same ROIs showed no statistical differences between group means. Two-sided p-values for matched pairwise comparisons between C57BL/6 and Sv/129 mice were: subcortical ROI at 2 h p = 0.34 and at 24 h p = 0.91; cortical ROI at 2 h p = 0.28 and at 24 h p = 0.63. Quantitative mean values for CBF and ADC are given in Table 2. Figure 4 displays the spatial distribution of functional outcome measures on color maps. Again, infarctions in both groups evolved from the subcortical to the cortical areas over time. Figure 4 Color maps of group means for CBF and ADC. Both are functional parameters cerebral ischemia indicating cerebral perfusion and cytotoxic diffusion restriction. Severe cortical and subcortical hypoperfusion (CBF maps) over time was observed in both groups. ADC maps show the transition from imminent to definite cerebral infarction involving cortical and subcortical territories. No significant differences between the strains were observed. Error bars denote standard errors of the mean. Table 2 Group means of functional ischemic outcome measures: CBF and ADC. Sv/129 C57Bl/6 CBF (ml/100g/min) 2 h vs. 24 h Cortex 34.1 ± 4.7 vs. 22.3 ± 2.9 38.7 ± 4.8 vs. 25.2 ± 3.7 Subcortex 30.4 ± 4.5 vs.20.8 ± 2.7 32.8 ± 4.6 vs. 23.4 ± 3.5

Figure 4 Color maps of group means for CBF and ADC. Both are functional parameters cerebral ischemia indicating cerebral perfusion and cytotoxic diffusion restriction. Severe cortical and subcortical hypoperfusion (CBF maps) over time was observed in both groups. ADC maps show the transition from imminent to definite cerebral infarction involving cortical and subcortical territories. No significant differences between the strains were observed. Error bars denote standard errors of the mean. Table 2 Group means of functional ischemic outcome measures: CBF and ADC. Sv/129 C57Bl/6 CBF (ml/100g/min) 2 h vs. 24 h Cortex 34.1 ± 4.7 vs. 22.3 ± 2.9 38.7 ± 4.8 vs. 25.2 ± 3.7 Subcortex 30.4 ± 4.5 vs.20.8 ± 2.7 32.8 ± 4.6 vs. 23.4 ± 3.5 ADC (mm2/s*10-4) 2 h vs. 24 h Cortex 7.8 ± 1.0 vs. 5.8 ± 0.5 7.8 ± 1.0 vs. 5.7 ± 0.6 Subcortex 7.3 ± 1.0 vs. 5.4 ± 0.5 7.2 ± 1.1 vs. 5.3 ± 0.4 Values are given for both cortical and subcortical ROIs and compared between both time points. P-values of statistical comparisons for CBF and ADC between groups were not significant and are given in the text. Standard errors of the mean are denoted by ±.

± 0.6 Subcortex 7.3 ± 1.0 vs. 5.4 ± 0.5 7.2 ± 1.1 vs. 5.3 ± 0.4 Values are given for both cortical and subcortical ROIs and compared between both time points. P-values of statistical comparisons for CBF and ADC between groups were not significant and are given in the text. Standard errors of the mean are denoted by ±. Discussion As principle finding, we here show that C57Bl/6 and Sv/129 mice, the two most frequently utilized strains in experimental stroke research, behave similar in the tMCAO model regarding critical functional and structural parameters of infarct development. Volumetric extents and probability of cerebral infarctions did not significantly differ between the two groups as assessed by multimodal ultra-high field MRI as did cerebral perfusion and diffusion restriction of free water.

the tMCAO model regarding critical functional and structural parameters of infarct development. Volumetric extents and probability of cerebral infarctions did not significantly differ between the two groups as assessed by multimodal ultra-high field MRI as did cerebral perfusion and diffusion restriction of free water. Strain-related differences between C57Bl/6 and Sv/129 mice have already been investigated in models of global and permanent focal cerebral ischemia. C57Bl/6 mice challenged by transient bilateral common carotid artery occlusion in the presence or absence of systemic hypotension developed more severe global forebrain ischemia than Sv/129 mice [7,8]. Results after permanent MCAO have been inconclusive. While some reports described larger infarctions in C57Bl/6 mice in direct comparison to the Sv/129 strain [2,9], others could not confirm this observation [4]. We further extend these findings to the most frequently used model of focal cerebral ischemia, i.e. tMCAO with an intaluminal thread. Here, 60 min of focal ischemia had no differential effect on definite infarct volumes on day 1. Importantly, diffusion restriction of free water as an early marker of ischemic cell damage and cytoxic edema likewise occurred similar in both strains. This implies that the C57Bl/6 or Sv/129 genetic background is no major confounding factor of infarct evolution in the acute phase after tMCAO. Whether this also applies to later stages of infarct development or different MCA occlusion times, however, still needs to be assessed.

ikewise occurred similar in both strains. This implies that the C57Bl/6 or Sv/129 genetic background is no major confounding factor of infarct evolution in the acute phase after tMCAO. Whether this also applies to later stages of infarct development or different MCA occlusion times, however, still needs to be assessed. Besides infarct volume, regional cerebral blood flow (rCBF) is frequently chosen as readout parameter in experimental stroke studies. Both critically depend on the composition of the cerebral vasculature which is known to vary considerably between different mouse strains [1-3]. The posterior communicating artery is absent or hypoplastic in more than 90% of C57/Bl6 mice, but in less than 50% of Sv/129 mice, which is also supported by in-vivo MRI data [4,25]. Moreover, the MCA territory seems to be larger in C57/Bl6 as compared to Sv/129 mice [26]. In contrast, pial anastomoses between the anterior and middle cerebral artery territories are of similar number and diameter [26]. PWI produced comparable results in C57/Bl6 and Sv/129 mice subjected to tMCAO indicating that the reported anatomical variations in the Circle of Willis do not translate into relevant functional differences in cerebral blood supply. Interestingly, infarction of subcortical areas (e.g. basal ganglia) was present as early as 2 h after tMCAO in both groups as assessed by various structural and functional MRI parameters. Early completed infarction in this subcortical territory of the MCA corresponds to the anatomically limited collateral blood supply of this region [27]. In contrast, severe hypoperfusion was observed in the cortical grey matter of the MCA territory at the same time point (2 h). Irreversible infarction followed only at 24 h determining this region as tissue-at-risk with a prolonged capacity for survival because of a richer collateral network through pial anastomoses. These observations emphasize the need for fast therapeutic interventions in ischemic stroke.

itory at the same time point (2 h). Irreversible infarction followed only at 24 h determining this region as tissue-at-risk with a prolonged capacity for survival because of a richer collateral network through pial anastomoses. These observations emphasize the need for fast therapeutic interventions in ischemic stroke. In summary, the C57/Bl6 and Sv/129 genetic background is no major confounding factors of stroke size and cerebral perfusion in the tMCAO model. Multimodal ultra-high field MRI is a valuable tool to non-invasively assess important structural and functional parameters of brain infarction in mice and can help to increase the validity of experimental stroke studies. Competing interests The authors declare that they have no competing interests. Authors' contributions MP, CK and SB wrote the paper. CK induced cerebral infarctions; MP and XH performed MRI measurements. MB, GS and PJ designed the study and revised the paper. All authors read and approved the final manuscript. Acknowledgements This study was supported by the Deutsche Forschungsgemeinschaft, Bonn (Sonderforschungsbereich 688 B1, A13, and Z-02) and a postdoctoral-fellowship granted to MP by the Medical Faculty of the University of Heidelberg.

WIND e. V. aims to provide an interdisciplinary and interactive platform for young researchers in order to gather and disseminate new knowledge in the fields of clinical and basic neurosciences. NEUROWIND e. V. focuses on three main topics: [1] cerebrovascular diseases, [2] neuroinflammation, and [3] neurodegeneration. Tremendous progress has been made in recent years in understanding the pathophysiology of individual disease conditions such as multiple sclerosis (MS), stroke, or Alzheimer's disease (AD). In spite of this success in unravelling disease mechanisms, the translation of novel experimental therapies into effective treatment for patients has so far been unsatisfying for a number of reasons, with one central problem certainly emanating from insufficient stringency of the translational process from bench to bedside and vice versa. One clear aim of NEUROWIND e. V. is therefore to bring together basic pathological and therapeutic concepts from different neurological disease models which at first glance might appear largely unrelated. However, numerous studies meanwhile have taught us that pathophysiological pathways and effector mechanisms appear to be common to a variety of different diseases. Good examples are the only recently recognized role of inflammation in stroke, M. Parkinson and AD or degenerative processes during the course of MS. Our ultimate goal is to overcome out-dated model barriers and raise the efficacy and quality of translational neurological research by fostering exchange of information and providing an interactive platform that favors fruitful discussions necessary to identify and solve distinct problems.

sses during the course of MS. Our ultimate goal is to overcome out-dated model barriers and raise the efficacy and quality of translational neurological research by fostering exchange of information and providing an interactive platform that favors fruitful discussions necessary to identify and solve distinct problems. To reach these ambitious goals, the first scientific meeting of NEUROWIND e.V. was recently held from Oct. 30'th - Nov. 1'st, 2009 in Mittenwalde/Motzen, Germany. Approximately 60 participants mainly on the level of doctoral students and young postdocs joined the meeting and presented their scientific work in the beautiful and stimulating environment of the Prussian scenery. A brief summary of the most interesting findings from the five sessions is given below and the full program as well as further information is available at http://www.neurowind.de.

young postdocs joined the meeting and presented their scientific work in the beautiful and stimulating environment of the Prussian scenery. A brief summary of the most interesting findings from the five sessions is given below and the full program as well as further information is available at http://www.neurowind.de. Summary of the scientific contributions to the NEUROWIND meeting 2009 Contributions in the fields of neuroimmunology and neurodegeneration Experimental autoimmune encephalomyelitis (EAE) is widely used to investigate the biology of autoreactive T cells in vivo. While EAE is considered an animal model to simulate inflammatory aspects of MS, similar aspects can be studied in the peripheral nervous system in the model of experimental autoimmune neuritis (EAN; work on new therapeutics in EAN presented by Gerd Meyer zu Hörste from the group of Bernd Kieseier, Dept. of Neurology, University of Düsseldorf). A series of myelin antigens have been identified as potential targets of autoreactive T cells and the most frequently used epitope to induce EAE in C57Bl/6 mice is the myelin oligodendrocyte glycoprotein (MOG) peptide 35-55. T cell receptor (TCR) transgenic mice with specificity for MOG35-55 have been generated and are used as a spontaneous model of MS since up to 10 percent of MOG35-55 TCR transgenic mice develop EAE without active immunization. As has recently been found by the group of Hartmut Wekerle and Florian Kurschus (Max Planck Institute for Neurobiology, Munich and Institute of Molecular Medicine, University Medical Center of the Johannes Gutenberg-University Mainz), MOG35-55 specific T cells from TCR transgenic mice also recognize another autoantigen, i. e. neurofilament-M (NF-M) peptide 18-30 [1]. Although NF-M and MOG belong to completely distinct protein families, the NF-M epitope shares essential TCR-contacting residues with MOG35-55 that allow for it to be recognized by MOG35-55 specific T cells when bound to the C57Bl/6 MHC class II complex (I-Ab). From these studies arises the concept of 'cumulative' autoimmunity with a degenerate TCR receptor recognizing several autoantigens that may not be related, thus surpassing the threshold to clinically manifest organ specific autoimmune inflammation. In a variety of EAE models using knockout animals, factors that enhance the pathogenicity of autoreactive T cells were characterized.

nity with a degenerate TCR receptor recognizing several autoantigens that may not be related, thus surpassing the threshold to clinically manifest organ specific autoimmune inflammation. In a variety of EAE models using knockout animals, factors that enhance the pathogenicity of autoreactive T cells were characterized. In the past, IFN-γ production by autoreactive T cells has been regarded as pathogenic hallmark of autoantigen specific T cells, which prompted the idea that organ specific autoimmunity might be a "Th1 disease" [2]. More recently, another phenotype of CD4+ T helper cells, so-called Th17 cells have been implicated in the development of EAE and other autoimmune disorders [3]. However, the plasticity of Th17 cells, which have been named after their signature cytokine IL-17, appears to be greater than that of Th1 cells. Whereas the combination of TGF-β and IL-6 in mouse (and TGF-β plus IL-21/or IL-1 in man) are the differentiation factors for Th17 cells [4-7], IL-23 has a major role in the stabilization of IL-17 production by Th17 cells and thus in stabilization of the functional phenotype of these cells. Very recent work performed in the group of Thomas Korn (presented by Malte Christian Claussen, Dept. of Neurology, Technical University of Munich) points to functions of IL-23 that are extrinsic to classical αβ T cells since non-classical T cells and perhaps even non-T cells express the IL-23R and respond to IL-23 by production of cytokines. Usage of IL-23R reporter mice will provide us with important information on the role of IL-23 activated non-classical T cells in tissue inflammation and autoimmunity. The population dynamics and migration events of both T cells and non-T cells into the CNS during EAE are still under intense investigation, and meticulous flow cytometric analysis has been applied by Karin Steinbach from the groups of Manuel Friese and Eva Tolosa (Center for Molecular Neurobiology, Hamburg) to reveal the temporal pattern of immune cell infiltration into the CNS during EAE. At the onset of disease, Th1 cells and Th17 cells accumulate in the CNS almost simultaneously. However, IL-17 production by T cells in the CNS appears to be sustained in actively induced MOG35-55 EAE. As far as the T cell/target interaction in the CNS is concerned, still little is known about the relative impact of Th1 vs Th17 cells. Indeed, the exact pathogenic role of Th17 cells in vivo is not yet understood.

However, IL-17 production by T cells in the CNS appears to be sustained in actively induced MOG35-55 EAE. As far as the T cell/target interaction in the CNS is concerned, still little is known about the relative impact of Th1 vs Th17 cells. Indeed, the exact pathogenic role of Th17 cells in vivo is not yet understood. Besides secreting cytokines and chemokines to attract other immune cells like neutrophils, Th17 cells may also directly interact with target structures in the CNS like naked axons that have been stripped off their myelin sheaths. Here, exciting in vivo imaging studies using the two photon technique [8] have been presented by Volker Siffrin from the group of Frauke Zipp (Dept. of Neurology, University Clinic of Mainz). These investigations revealed that Th17 cells indeed were able to induce axonal damage in an MHC class II independent manner. Although the mechanism of lesion generation remains to be determined, perforin degranulation may be involved in inducing axonal damage.

uke Zipp (Dept. of Neurology, University Clinic of Mainz). These investigations revealed that Th17 cells indeed were able to induce axonal damage in an MHC class II independent manner. Although the mechanism of lesion generation remains to be determined, perforin degranulation may be involved in inducing axonal damage. While the role of CD4+ T cells for the induction of immunopathology is a proven fact in EAE, it has been difficult to assess the mechanism of how CD8+ cytotoxic T cells contribute to lesion formation in this model. Interestingly, CD8+ T cells are present in human MS lesions and have been shown to be clonally expanded suggesting a pathogenic role of these cells [9]. However, only recently new tools have become available to study the role of CD8+ T cells in an experimental setting [10]. Furthermore, double transgenic mice have been generated that express OVA as a neo-autoantigen in a cell type specific manner in oligodendrocytes under the MBP promoter and at the same time bear a transgenic TCR for a specific OVA-peptide on their CD8+ T cells (OT-I) [11]. In a highly interesting study by Kerstin Göbel and Nico Melzer from the group of Heinz Wiendl (Dept. of Neurology, University Clinic of Würzburg), brain slices from MBP-OVA transgenic mice have been used to investigate the interaction of OT-I cells with oligodendrocytes presenting OVA-peptide [12]. While OT-I cells induced apoptosis in oligodendrocytes via direct attack, neighboring neurons were also affected, revealing the possibility of perforin-driven collateral damage in neurons. Besides immunohistochemical and optical methods to characterize lesion formation on a molecular level in vivo, electrophysiological methods like patch clamp approaches are also being used to further characterize the impact of cytotoxic T lymphocytes (CTLs) on neurons. In vitro, it is possible to measure the break-down of the neuronal membrane potential when these neurons, that have been induced to express MHC class I and have been loaded with OVA peptide, are attacked by OVA-peptide specific CD8+ OT-I T cells. Again it appears that this process of CTL-dependent damage to neurons is perforin mediated since perforin deficient OT-I cells fail to short-circuit axons. In order to approach questions of CTL/target interaction on the molecular level in vivo, viral models of neuroinflammation are helpful tools.

fic CD8+ OT-I T cells. Again it appears that this process of CTL-dependent damage to neurons is perforin mediated since perforin deficient OT-I cells fail to short-circuit axons. In order to approach questions of CTL/target interaction on the molecular level in vivo, viral models of neuroinflammation are helpful tools. Recently, a model of LCMV infection is being explored in which the infecting agent (an RNA virus) can be manipulated by reverse genetics and specific T cell responses against infected neurons can be monitored [13,14].

fic CD8+ OT-I T cells. Again it appears that this process of CTL-dependent damage to neurons is perforin mediated since perforin deficient OT-I cells fail to short-circuit axons. In order to approach questions of CTL/target interaction on the molecular level in vivo, viral models of neuroinflammation are helpful tools. Recently, a model of LCMV infection is being explored in which the infecting agent (an RNA virus) can be manipulated by reverse genetics and specific T cell responses against infected neurons can be monitored [13,14]. In this model which is being established by Mario Kreutzfeldt from the group of Doron Merkler at the Dept. of Neuropathology, University of Göttingen, the avidity of the CTL/peptide/MHC class I interaction can be modified and differential immunopathological responses can be studied in vivo. Lesion generation in neuroinflammation is not only dependent on cytotoxic CD8+ T cells, but monocytes and macrophages are major players in inducing damage to myelin and neurons. While T cell-derived IFN-γ is considered the canonical molecule to activate macrophages, a series of other stimuli can trigger effector functions in these cells. Interestingly, adhesion of monocytes to immobilized platelets results in massive TNF production by monocytes. This phenomenon is now being characterized by Harald Langer in the group of Triantafyllos Chavakis (NIH, Bethesda, USA). The molecular basis for the interaction of platelets and monocytes appears to be a ligand/receptor interaction between GPIb on platelets and Mac-1 (CD11b/CD18) on monocytes. Blockade of this interaction results in diminished secretion of TNF by macrophages, and GPIb deficient mice develop attenuated EAE suggesting that platelet-mediated activation of macrophages might be an important effector mechanism in this disease. Besides immune cell/neuron interactions, several further pathways may play an important role in inflammation-mediated neurodegeneration. Among others, neurotrophic factors and ion channels have recently been in the focus of interest. Neurotrophic factors comprise neutrophins and neurotrophic cytokines which are mainly produced in the nervous system, but also by immune cells. As a prototype mediator, the role of brain derived neurotrophic factor (BDNF) and its receptors trkB and p75NTR have been characterized in MS lesions and more recently also in EAE models [15-17]. Using an experimental approach with bone marrow chimera, Tobias Dallenga from the groups of Stefan Nessler and Christine Stadelmann (Dept. of Neuropathology, University of Göttingen) has studied p75NTR-mediated signaling in immune cells and non-immune cells during EAE.

ions and more recently also in EAE models [15-17]. Using an experimental approach with bone marrow chimera, Tobias Dallenga from the groups of Stefan Nessler and Christine Stadelmann (Dept. of Neuropathology, University of Göttingen) has studied p75NTR-mediated signaling in immune cells and non-immune cells during EAE. These data suggest an important role of immune cell-derived BDNF and p75NTR-mediated signaling pathways for axon protection. In autoimmune inflammation, Nav and Kv channels as well as acid sensing ion channels were all shown to play a role for axon or glial cell function but in part also for regulation of the immune cell response [18,19]. More recently, a new family of genes encoding two-pore domain potassium channels that generate "leak" potassium currents were characterized. The TASK subfamily of channels, notably TASK-1 (KCNK3) and TASK-3 (KCNK9), have now been shown to modulate inflammation and neurodegeneration in EAE and probably also ischemic stroke, thus identifying new potential molecular targets for the therapy of inflammatory and degenerative CNS disorders. These data were presented by Petra Ehling from the group of Thomas Budde, Dept. of Physiology, University of Münster.

to modulate inflammation and neurodegeneration in EAE and probably also ischemic stroke, thus identifying new potential molecular targets for the therapy of inflammatory and degenerative CNS disorders. These data were presented by Petra Ehling from the group of Thomas Budde, Dept. of Physiology, University of Münster. Investigation of immune cell/target cell interactions is certainly a major domain of animal studies. Yet, studies with human immune cells are required to test the relevance of hypotheses that have been raised in animal models. We are now witnessing the first reports on Th17 cells in human MS on a larger scale. Notably, the frequency of Th17 cell clones but not Th1 clones in the peripheral blood and the CSF appears to be correlated with disease activity in relapsing remitting MS [20,21]. In ex vivo analyses performed by Verena Brucklacher-Waldert from the group of Eva Tolosa (Center for Molecular Neurobiology, Hamburg), Th17 cells had a strongly activated phenotype and were relatively resistant to regulatory T cell (Treg)-mediated suppression in comparison with Th1 cells. In contrast to EAE which is clearly dependent on CD4+ T helper cells, human MS is more complex and a plethora of other immune cells are directly or indirectly involved in the generation of MS lesions. NK cells have been of particular interest in this regard since their modulation appears to be the basis of the efficacy of a monoclonal antibody to the IL-2R (daclizumab) that has recently been tested in clinical trials [22]. Brady Messmer from Jan Luenemann's group at the Institute for Experimental Immunology, University of Zuerich, Switzerland studied the role of NK cells in MS patients in more detail. NK have been identified in MS lesions. CD56dim NK cells are equipped with the molecular machinery to kill their target cells and are more prominent in PBMCs as compared with lymph nodes. In contrast, CD56bright NK cells produce cytokines like IFN-γ, TNF, and GM-CSF and are more abundant in lymph nodes than in PBMCs [23]. CD56bright NK cells are divided into CD16- and CD16+ populations. In order to evaluate the role of NK cells in the pathogenesis of MS in more detail, NK cells of MS patients were studied and compared with those of healthy control subjects. In MS patients, CD56brightCD16- NK cells show impaired expansion and produce less IFN-γ in response to IL-12 whereas the lytic function of NK cells (CD56dim) appears to be unchanged.

cells in the pathogenesis of MS in more detail, NK cells of MS patients were studied and compared with those of healthy control subjects. In MS patients, CD56brightCD16- NK cells show impaired expansion and produce less IFN-γ in response to IL-12 whereas the lytic function of NK cells (CD56dim) appears to be unchanged. These data suggest that subsets of NK cells in MS patients display a functionally different phenotype and may be implicated in the disease process.

cells in the pathogenesis of MS in more detail, NK cells of MS patients were studied and compared with those of healthy control subjects. In MS patients, CD56brightCD16- NK cells show impaired expansion and produce less IFN-γ in response to IL-12 whereas the lytic function of NK cells (CD56dim) appears to be unchanged. These data suggest that subsets of NK cells in MS patients display a functionally different phenotype and may be implicated in the disease process. Further lessons on the relation between immune cells and CNS tissue can be learned from degenerative diseases and their respective animal models. Here, studies in mouse models of Alzheimer's disease (AD) as well as Huntington's disease (HD) recently gained much interest. Initially, AD has been characterized by formation of amyloid plaques, neurofibrillary tangles and subsequent neurodegeneration. More recently, the role of immune cells, most notably microglia, in this process has been characterized. While the exact sequence of events eventually leading to neuronal death still remain to be determined, several mediators involved in regulation of microglia have been identified. Here, especially chemokines and their receptors were found to play an important role, e.g. for cell migration (work presented by Marius Krauthausen from the group of Markus Müller and Michael Heneka, Dept. of Neurology, University of Bonn). The modulation of such factors may critically regulate microglial function and finally also influence the process of neurodegeneration. Similar observations were reported in models of HD which are characterized by formation of intraneuronal huntingtin aggregates and neuronal dysfunction. Here, innovative neurobiological treatment approaches such as anti-sense technologies or stem-cell based repair strategies can reduce huntingtin aggregate load and improve functional deficits in mouse models of HD as presented by Christian Saß from the Dept. of Neurology, University Clinic of Aachen. In addition, more established treatment strategies such as therapy with immunomodulators appear to be effective as well (Christiane Reick, Dept. of Neurology, St. Josef-Hospital, Ruhr-University Bochum). These data point to a role of the immune system in HD, but also to a putative neuroprotective function of these drugs, thus opening up an exciting new avenue of translational research linking the fields of neuroimmunology and neurobiology.

ristiane Reick, Dept. of Neurology, St. Josef-Hospital, Ruhr-University Bochum). These data point to a role of the immune system in HD, but also to a putative neuroprotective function of these drugs, thus opening up an exciting new avenue of translational research linking the fields of neuroimmunology and neurobiology. Contributions on stroke and vascular pathology Ischemic stroke is a devastating disease that represents the second leading cause of death worldwide. Each year, 575.000 people in Europe fall victim to ischemic stroke, which is estimated to cost 71.8 billion Euros [European Stroke Initiative]. It is estimated that the lifetime risk for stroke is between 8% and 10%. Early restoration of blood flow remains the treatment of choice for limiting brain injury following stroke. While reperfusion of the ischemic brain is desirable in principle, it may also foster tissue damage under certain conditions. Reperfusion appears to augment the inflammatory response and causes additional injury to adjacent brain tissue. Hence, a rapidly evolving area of stroke research involves defining the molecular and cellular basis for this secondary tissue injury and inflammation associated with transient cerebral ischemia. For this research, primarily the middle cerebral artery occlusion reperfusion model in mice is used.

adjacent brain tissue. Hence, a rapidly evolving area of stroke research involves defining the molecular and cellular basis for this secondary tissue injury and inflammation associated with transient cerebral ischemia. For this research, primarily the middle cerebral artery occlusion reperfusion model in mice is used. The inflammatory response seen in this model is initiated by an accumulation of microglia and the secretion of pro-inflammatory cytokines such as IL-1β, IL-6 or MCP-1 (researched by Mathias Gelderblom from the groupd of Tim Magnus, Dept. of Neurology, University Clinic Hamburg-Eppendorf, Hamburg). On the cellular level, infiltration of the ischemic hemisphere by macrophages, lymphocytes, and dendritic cells (DCs) within the first day precedes neutrophilic influx. Up-regulation of MHC-II and the co-stimulatory molecule CD80 demonstrate activation of these cells arguing for a pro-inflammatory environment. However, also regulatory immune cells (NKT cells, CD4-/CD8-T lymphocytes, Foxp3+ T cells) accumulate in the ischemic brain [24].

thin the first day precedes neutrophilic influx. Up-regulation of MHC-II and the co-stimulatory molecule CD80 demonstrate activation of these cells arguing for a pro-inflammatory environment. However, also regulatory immune cells (NKT cells, CD4-/CD8-T lymphocytes, Foxp3+ T cells) accumulate in the ischemic brain [24]. The functional relevance of inflammatory cells can be proven in knockout animals such as Rag deficient mice that lack B and T cells and are largely protected from inflammatory damage secondary to ischemic stroke [25]. Conversely, depletion of regulatory T cells (Treg) induces a significant increase in infarct size pointing to a relevant part of these cells in regulating post-stroke inflammation [26]. Furthermore, inhibition of cell migration into the lesioned brain might become an interesting approach to modulate stroke-induced immune pathways. Arthur Liesz from the group of Roland Veltkamp, Dept. of Neurology, University of Heidelberg, showed that blockage of immune cell entry results in smaller infarcts and an improved neurological outcome. This blockage was achieved by an antibody preventing the binding of α 4 integrins, a common therapeutic approach also in MS. So far, it remains unclear which cell type (if any) is the key player in ischemic stroke. However, it seems likely that T cells and their subtypes play an important role while the function of neutrophils, which also express α 4 integrins, is not yet clear.

ing of α 4 integrins, a common therapeutic approach also in MS. So far, it remains unclear which cell type (if any) is the key player in ischemic stroke. However, it seems likely that T cells and their subtypes play an important role while the function of neutrophils, which also express α 4 integrins, is not yet clear. In contrast to the local pro-inflammatory response within the CNS, changes in the systemic immune compartment indicate a more general stroke-associated immune suppression. The latter can, as Odilio Engel from Andreas Meisel's group at the Dept. of Experimental Neurology, Charité Universitätsmedizin Berlin, points out, be observed in patients as well as in the animal model, where an increased bacterial load is found in the lungs of stroked rodents. The immunosuppressive effect may be elicited by an increase in vagal activation and subsequent secretion of acetylcholine in lymph nodes and spleen.

Universitätsmedizin Berlin, points out, be observed in patients as well as in the animal model, where an increased bacterial load is found in the lungs of stroked rodents. The immunosuppressive effect may be elicited by an increase in vagal activation and subsequent secretion of acetylcholine in lymph nodes and spleen. Another facet in stroke research is related to the occurrence of oxidative stress. Cells and especially neurons have to deal within minutes with reactive oxygen and nitrogen species (ROS/RNS). One attractive candidate source for oxidative stress in acute ischemic stroke are NADPH oxidases, the only known enzyme family that has ROS as their sole enzymatic product. These are the molecules of a specific research interest for Tobias Schwarz form Christoph Kleinschnitz's group at the Dept. of Neurology, University Clinic of Würzburg. In rodents 4 NOX genes exist, and in the rodent brain NOX are mainly expressed in neurons and the vasculature with NOX4 being the most abundant isoform. In an interesting study using NOX1, NOX2 and NOX4 deficient mice as well as the specific NOX Inhibitor VAS2870, the pathophysiological role of the different NOX isoforms in ischemic stroke has now been assessed in terms of infarct development and blood-brain-barrier damage.

re with NOX4 being the most abundant isoform. In an interesting study using NOX1, NOX2 and NOX4 deficient mice as well as the specific NOX Inhibitor VAS2870, the pathophysiological role of the different NOX isoforms in ischemic stroke has now been assessed in terms of infarct development and blood-brain-barrier damage. Our current performance in the acute treatment of stroke patients is moderate at best and, therefore, additional efforts to enhance tissue repair are badly needed. As outlined above, inflammatory cascades are active during cerebral ischemia. However, their effects are not necessarily detrimental, as Karen Gertz from the group of Matthias Endres, Dept. of Neurology, Charité Universitätsmedizin Berlin, reminds us. IL-6, for example, helps to increase vascular repair and possibly neogenesis and can improve long term outcome in experimental stroke. Other strategies involve the use of stem cells (presented by Jens Minnerup from the group of Wolf-Rüdiger Schäbitz, Dept. of Neurology, University of Münster). However, it appears that significant tissue repair derived from endogenous stem cells is not realistic in ischemic stroke, at least in the near future. The systemic application has turned out difficult since it is not easy to derive the perfect cell that is undifferentiated enough to integrate and survive but develops into a functional neuron. Although some functional improvement can be seen in stroked rodents after the application of neurospheres depending on their differentiation protocol, the therapeutic effects are still relatively small. Finally, Jan Klohs from Ulrich Dirnagl's group at the Dept. of Experimental Neurology, Charité Universitätsmedizin Berlin, introduced a new imaging technique in rodent stroke models [27]. Near-infrared fluorescence (NIRF) imaging is suitable to visualize distinct molecules involved in the pathophysiology of ischemic stroke in rodents in vivo by utilizing specific NIRF probes, e.g. against matrix metalloproteinases (MMPs). Although its temporal resolution is still relatively low, this non-invasive method could be useful in monitoring treatment responses in individual animals over time.

es involved in the pathophysiology of ischemic stroke in rodents in vivo by utilizing specific NIRF probes, e.g. against matrix metalloproteinases (MMPs). Although its temporal resolution is still relatively low, this non-invasive method could be useful in monitoring treatment responses in individual animals over time. Concluding remarks Lesion formation and repair are always essential events in pathological conditions in the CNS. Although distinct diseases, vascular, inflammatory and neurodegenerative disorders of the CNS may share pathological sequences on the cellular and molecular level. Decades of experience and gathering of knowledge in the formation, trafficking and effector functions of immune cells in neuroinflammatory conditions may help to better understand the involvement of immune cells in vascular diseases and neurodegenerative disorders. Thus, comparing cellular reactions in the peripheral immune compartment and the CNS across different disease models may offer an unconventional but very efficient means to generate new ideas and promote research. We hope that our initiative will help young researches to create new concepts and opportunities to improve the understanding of many neurological diseases and to find new treatment options. We encourage all members of the neurological community to support our idea and provide constructive input for the upcoming meeting in October 2010. Competing interests The authors declare that they have no competing interests. Authors' contributions TM, RL, SGM, CK and TK wrote the paper. All authors read and approved the final manuscript.

Concluding remarks Lesion formation and repair are always essential events in pathological conditions in the CNS. Although distinct diseases, vascular, inflammatory and neurodegenerative disorders of the CNS may share pathological sequences on the cellular and molecular level. Decades of experience and gathering of knowledge in the formation, trafficking and effector functions of immune cells in neuroinflammatory conditions may help to better understand the involvement of immune cells in vascular diseases and neurodegenerative disorders. Thus, comparing cellular reactions in the peripheral immune compartment and the CNS across different disease models may offer an unconventional but very efficient means to generate new ideas and promote research. We hope that our initiative will help young researches to create new concepts and opportunities to improve the understanding of many neurological diseases and to find new treatment options. We encourage all members of the neurological community to support our idea and provide constructive input for the upcoming meeting in October 2010. Competing interests The authors declare that they have no competing interests. Authors' contributions TM, RL, SGM, CK and TK wrote the paper. All authors read and approved the final manuscript. Acknowledgements The first scientific meeting of NEUROWIND e.V. was kindly supported by Merck Serono GmbH, Darmstadt, Germany (unrestricted grant to NEUROWIND e.V.). The authors thank all speakers at the 1'st NEUROWIND e.V. scientific meeting. We thank Ms. Anke Bauer, Wuerzburg, for editing the manuscript.

Authors' contributions TM, RL, SGM, CK and TK wrote the paper. All authors read and approved the final manuscript. Acknowledgements The first scientific meeting of NEUROWIND e.V. was kindly supported by Merck Serono GmbH, Darmstadt, Germany (unrestricted grant to NEUROWIND e.V.). The authors thank all speakers at the 1'st NEUROWIND e.V. scientific meeting. We thank Ms. Anke Bauer, Wuerzburg, for editing the manuscript. Co-investigators at the 1st scientific meeting of NEUROWIND e.V. (in alphabetical order)

Acknowledgements The first scientific meeting of NEUROWIND e.V. was kindly supported by Merck Serono GmbH, Darmstadt, Germany (unrestricted grant to NEUROWIND e.V.). The authors thank all speakers at the 1'st NEUROWIND e.V. scientific meeting. We thank Ms. Anke Bauer, Wuerzburg, for editing the manuscript. Co-investigators at the 1st scientific meeting of NEUROWIND e.V. (in alphabetical order) Verena Brucklacher-Waldert, Center for Molecular Neurobiology, Hamburg, Germany; Malte Christian Claussen, Dept. of Neurology, Technical University of Munich, Germany; Tobias Dallenga, Dept. of Neuropathology, University of Göttingen, Germany; Petra Ehling, Dept. of Physiology, University of Münster, Germany; Odilio Engel, Dept. of Experimental Neurology, Charité Universitätsmedizin Berlin, Germany; Mathias Gelderblom, Dept. of Neurology, University Clinic Hamburg-Eppendorf, Hamburg, Germany; Karen Gertz, Dept. of Neurology, Charité Universitätsmedizin Berlin, Germany; Jan Klohs, Dept. of Experimental Neurology, Charité Universitätsmedizin Berlin, Germany; Marius Krauthausen, Dept. of Neurology, University of Bonn, Germany; Mario Kreutzfeldt, Dept. of Neuropathology, University of Göttingen, Germany; Florian Kurschus, Institute of Molecular Medicine, University Medical Center of the Johannes Gutenberg-University Mainz, Germany; Harald Langer, NIH, Bethesda, USA; Arthur Liesz, Dept. of Neurology, University Clinic of Heidelberg, Germany; Jan Lünemann, Institute for Experimental Immunology, University of Zuerich, Switzerland; Nico Melzer, Dept. of Neurology, University Clinic of Würzburg, Germany; Brady Messmer, Institute for Experimental Immunology, University of Zurich, Switzerland; Gerd Meyer zu Hörste, Dept. of Neurology, University of Düsseldorf, Germany; Jens Minnerup, Dept. of Neurology, University of Münster, Germany; Christiane Reick, Dept. of Neurology, St. Josef-Hospital, Ruhr-University Bochum, Germany; Christian Saß, Dept. of Neurology, University Clinic of Aachen, Germany; Tobias Schwarz, Dept. of Neurology, University Clinic of Würzburg, Germany; Volker Siffrin, Dept. of Neurology, University Clinic of Mainz, Germany; Karin Steinbach, Center for Molecular Neurobiology, Hamburg, Germany; Heinz Wiendl (keynote lecture), Dept. of Neurology, University Clinic of Würzburg, Germany

Editorial The recently published double-blind, placebo-controlled, randomized phase II/III German Multicenter EPO Stroke Trial was conducted to evaluate the efficacy and safety of Erythropoietin (EPO) in stroke patients [1]. Of the 522 patients enrolled in this trial 460 were treated as planned (per-protocol population) with either EPO or placebo within 6 hours of symptom onset. The primary endpoint, change in Barthel Index on day 90, and all secondary outcomes failed to show any benefit of EPO. Moreover, an increased rate of intracerebral haemorrhages was observed after EPO treatment, resulting in an increased mortality in the EPO group. This effect was pronounced in patients who received EPO in addition to rt-PA. In this Editorial we discuss potential reasons for the negative results of the German Multicenter EPO Stroke Trial, which contrasted the findings of a clinical pilot trial and several preclinical studies that showed beneficial effects of EPO [2,3]. Altogether we want to reflect on four major issues: 1. The overestimated efficacy of EPO in preclinical studies due to neglected quality characteristics in animal experiments. 2. An underpowering caused by the study design of the German Multicenter EPO Stroke Trial. 3. Unexpected side effects of EPO. 4. The future for a further development of EPO as a stroke drug.

sues: 1. The overestimated efficacy of EPO in preclinical studies due to neglected quality characteristics in animal experiments. 2. An underpowering caused by the study design of the German Multicenter EPO Stroke Trial. 3. Unexpected side effects of EPO. 4. The future for a further development of EPO as a stroke drug. So far EPO and EPO analogues were widely tested in animal stroke models [3]. In a meta-analysis of preclinical studies we analyzed the overall efficacy in focal cerebral ischemia. EPO and EPO analogues reduced infarct volumes by 32% and improved neurobehavioral deficits by 37% to 38%. However, Philip et al. recently showed that the quality of preclinical EPO studies as measured by a STAIR derived quality score was relatively low [4]. This is a crucial point because disregarding basic quality standards may cause an overestimation of a drug's efficacy in animal studies [5,6]. Indeed, this might be the case in experimental EPO studies. When animals were randomized to EPO treatment or placebo the efficacy was lower compared to studies in which randomization was not reported [7]. The way in which the outcome was assessed was identified as a further potential source of bias. When comparing studies that blindly assessed neurobehavioral deficits to studies with an unblinded assessment of outcome the latter reported a significantly higher efficacy of EPO [7].

ich randomization was not reported [7]. The way in which the outcome was assessed was identified as a further potential source of bias. When comparing studies that blindly assessed neurobehavioral deficits to studies with an unblinded assessment of outcome the latter reported a significantly higher efficacy of EPO [7]. The study design of the German Multicenter Stroke Trial is another potential reason for the failure to replicate the positive findings of prior preclinical and clinical studies. A particularly critical point is the allowed combination of rt-PA and EPO. This combination of treatments was neither investigated in animal models nor in the clinical pilot trial. Therefore adverse interactions of these two drugs as suggested by the increased rate of intracerebral haemorrhages in the German Multicenter Stroke Trial were unpredictable. A preceding investigation of EPO-rt-PA interactions could have had prevented a combination therapy in the clinical trial. In fact, a present mouse stroke study by Zechariah et al. showed that a combination of EPO and rt-PA induces blood-brain barrier permeability and extracellular matrix disaggregation [8]. However, it is uncertain whether the results of a single animal study based on surrogate markers would have influenced further clinical development of EPO, particularly in combination with rt-PA. When considering that an increasing number of stroke patients is treated with rt-PA within a therapeutical time window also adequate for neuroprotective therapies, the importance of thoroughly testing combination therapies in animal studies becomes evident [9]. This particularly includes investigations of whether the drugs interfere regarding their efficacy and safety. In addition, the allowed combination of EPO and rt-PA is critical for the issue of study power. Altogether, it is rather difficult to measure a beneficial effect on top of a highly effective therapy such as thrombolysis. The inclusion criteria also might have reduced the power: Patients with pre-existing disability were included in the study making it difficult to measure treatment-related differences on the primary outcome Barthel Index or on secondary outcomes such as the modified Rankin Scale. However, the pre-stroke Barthel Index and the pre-stroke modified Rankin Scale were not reported in the manuscript.

existing disability were included in the study making it difficult to measure treatment-related differences on the primary outcome Barthel Index or on secondary outcomes such as the modified Rankin Scale. However, the pre-stroke Barthel Index and the pre-stroke modified Rankin Scale were not reported in the manuscript. The negative findings of the German Multicenter EPO Stroke Trial could be a result of previously unknown side effects of EPO. So far, it was known that EPO increases the risk for myocardial infarction and composite end-points of death and cardiovascular events in patients with anaemia due to chronic kidney disease [10]. In addition, some years ago EPO was shown to enhance tumor progression and shorten survival in patients with some types of cancer [11]. Results of the more recent TREAT study suggest an intrinsic stroke-inducing capacity of EPO. In this study patients with diabetes, chronic kidney disease, and anaemia were randomly assigned to receive darbepoetin alfa or placebo [12]. Surprisingly, a significant higher number of strokes occurred in the darbepoetin alfa treated group compared to the placebo group. Unfortunately, it was not reported whether those strokes were ischemic or hemorrhagic [13].

c kidney disease, and anaemia were randomly assigned to receive darbepoetin alfa or placebo [12]. Surprisingly, a significant higher number of strokes occurred in the darbepoetin alfa treated group compared to the placebo group. Unfortunately, it was not reported whether those strokes were ischemic or hemorrhagic [13]. The question arises what the disappointing results of the recent EPO trial mean for a future clinical development of the drug. One might consider that a further clinical stroke trial which excludes patients treated with rt-PA might show beneficial effects of EPO. Results of the German Multicenter EPO Stroke Trial, however, do not strongly support this assumption. In a subgroup analysis of non-rt-PA group none of the primary endpoints differed significantly between EPO and placebo treated patients. Only one secondary outcome measure, the delta NIHSS (NIHSS Day 1 minus Day 90) [1], revealed a better outcome after EPO treatment. In non-rt-PA treated patients there was even a tendency toward a higher death rate in the EPO group. The authors point out that this might be explained by the higher stroke severity of the dead patients on inclusion. Overall, the potential side effects of EPO will presumably prevent the conduction of further clinical stroke trials. However, non-haematopoietic EPO analogues remain as a therapeutic option for stroke, since the adverse effects of EPO were assumed to be mainly caused by its erythropoiesis stimulating effects. In a meta-analysis of preclinical studies we showed that non-hematopoietic EPO analogues are at least as effective as hematopoietic EPO-analogues [3]. The reason therefore might be EPO's mode of action in ischemic stroke, which is assumed to be based on a direct effect on neurons rather than on an increased hematopoiesis (for review see [14]). It was shown, that EPO receptors are expressed in the brain and that the neuronal EPO receptors which are distinct from those expressed by erythropoid precursors are stimulated by non-hematopoietic EPO analogues. Evidence regarding the safety of non-hematopoietic EPO analogues in stroke patients is expected in the near future since one clinical pilot trial of Carbamylated EPO in stroke patients was recently completed and another pilot trial has already started (http://www.clinicaltrials.gov/; NCT00756249 and NCT00870844). The future of non-hematopoietic EPO analogues for a further clinical development for stroke therapy will depend on the safety results of these trials.

of Carbamylated EPO in stroke patients was recently completed and another pilot trial has already started (http://www.clinicaltrials.gov/; NCT00756249 and NCT00870844). The future of non-hematopoietic EPO analogues for a further clinical development for stroke therapy will depend on the safety results of these trials. Competing interests The authors declare that they have no competing interests. Authors' contributions JM wrote the manuscript HW wrote the manuscript WRS revised the manuscript Acknowledgements None

Background In various animal models of central nervous system (CNS) injury, bone marrow stromal cells (BMSC) have been reported to be effective in limiting tissue damage[1-4]. The therapeutic effects of stem cells have been attributed to their ability to release of a mixture of neurotrophins, growth factors, and other substances that induce restorative processes in post-ischemic brain tissue[1,5-8]. Although some studies have focused on the role of one or more potential neuroprotective factors that are released from BMSC, less effort has been made to evaluate the responses of BMSC once they infiltrate brain tissue following an ischemic insult[7,9].

induce restorative processes in post-ischemic brain tissue[1,5-8]. Although some studies have focused on the role of one or more potential neuroprotective factors that are released from BMSC, less effort has been made to evaluate the responses of BMSC once they infiltrate brain tissue following an ischemic insult[7,9]. Immortalized cell lines have long been used as a source of stem cells for in vivo studies of the therapeutic efficacy of stem cells in models of ischemic tissue injury[10-12]. While these cells have been shown to survive and differentiate in brain tissue and to afford some protection against ischemic stroke, they are difficult to recover from brain tissue to permit assessment of the phenotypic and genetic changes that underlie their protective actions. Immortalized BMSC isolated from the H-2Kb-tsA58 transgenic mouse express a gene for temperature sensitive conditional immortality that makes them a more suitable model for stem cell recovery in ischemic tissue[13]. Stem cells from this background exhibit stem cell marker characteristics for over a year[13] and can give rise to cells of the mesenchymal lineage [14]. A particularly advantageous characteristic of BMSC from the H-2 Kb-tsA58 transgenic strain, is that the cells are undifferentiated at 33°C, but will acquire a differentiated phenotype at 37°C to yield a large number of cells with stem cell properties.

year[13] and can give rise to cells of the mesenchymal lineage [14]. A particularly advantageous characteristic of BMSC from the H-2 Kb-tsA58 transgenic strain, is that the cells are undifferentiated at 33°C, but will acquire a differentiated phenotype at 37°C to yield a large number of cells with stem cell properties. Although it is widely accepted that stem cells administered in animal models of stroke selectively home to and infiltrate the site of brain injury[14,15], successful isolation and genetic evaluation of these cells after they have infiltrated the post-ischemic brain has not been reported to date. In this study, we employed immortalized BMSC from H-2 Kb-tsA58 mice to selectively isolate stem cells that infiltrate brain tissue and produce therapeutic benefit following focal ischemia and reperfusion. The infiltrating BMSC were probed using whole genome array and RT-PCR in order to identify genes that are persistently up- or down-regulated in the stem cells after their appearance in infarcted brain tissue. Methods Animals All in vivo experiments were performed on male C57Bl/6J mice (WT; 6 to 8 weeks old) (Jackson Laboratories, Bar Harbor, Me). BMSC were isolated from either H-2Kb-tsA58 mice expressing temperature-sensitive SV40 large T antigen (Large T; CBA/ca × C57Bl/10 hybrid, Charles River Laboratories) or from WT mice. The experimental procedures employed in this study were approved by the Louisiana State University Health Sciences Center Institutional Animal Care and Use Committee and are in compliance with the guidelines of National Institutes of Health.

arge T; CBA/ca × C57Bl/10 hybrid, Charles River Laboratories) or from WT mice. The experimental procedures employed in this study were approved by the Louisiana State University Health Sciences Center Institutional Animal Care and Use Committee and are in compliance with the guidelines of National Institutes of Health. BMSC isolation Primary cultures of BMSC were obtained from WT or Large T mice as previously described[16]. Briefly, fresh complete bone marrow was harvested aseptically from the tibias and femurs and then cultured in Iscove's Modified Dulbecco's medium (IMDM) supplemented with 10% fetal bovine serum (FBS). BMSC isolated from Large T mice were cultured at 33°C for selective isolation of the immortalized cells. After 3 days of incubation, non-adherent cells were removed and cells tightly adhered to plastic were isolated and resuspended to fresh Iscove's Modified Dulbecco's medium in new flasks for further growth. By passage 3, less than 1% of cells were positive for CD11b and CD45 as assessed by flow cytometry, and the BMSC were stem cell antigen-1 (sca-1) positive (70%). Prior to use in the in vivo model, BMSC were harvested using a non-enzymatic dissociation solution (Sigma Chemicals, St Louis), centrifuged at 1000 × g, filtered through a 70 um cell strainer (BD, Falcon), and resuspended in PBS (pH 7.4). 2 × 106 viable WT or Large T BMSC(in 150 μl of PBS) or PBS (150 μl) were administered intravenously at 24 hours after the induction of cerebral ischemia.

atic dissociation solution (Sigma Chemicals, St Louis), centrifuged at 1000 × g, filtered through a 70 um cell strainer (BD, Falcon), and resuspended in PBS (pH 7.4). 2 × 106 viable WT or Large T BMSC(in 150 μl of PBS) or PBS (150 μl) were administered intravenously at 24 hours after the induction of cerebral ischemia. Middle cerebral artery (MCA) occlusion and reperfusion (MCAo/R) The mice were anesthetized by intraperitoneal injection of ketamine (50 mg/kg) and xylazine (2.5 mg/kg). Transient (60 minutes) focal cerebral ischemia was induced by occlusion of the left middle cerebral artery (MCAo) using a modification of intraluminal filament method. Briefly, the blunted tip of a 6-0 nylon monofilament was advanced to the level of the carotid bifurcation via the internal carotid artery until light resistance was felt. The distance from the nylon thread tip to the internal carotid artery-pterygopalatine artery bifurcation was slightly greater than 6 mm, and the distance to the bifurcation of the internal and external carotid arteries was slightly less than 9 mm. The monofilament was removed after 60 minutes of occlusion. In the sham group, these arteries were visualized but not disturbed. Ischemia and reperfusion (I/R) were verified using a Laser Doppler flowmeter probe (MSP300XP, AD Instruments Inc.) attached to the left parietal cranium. At the end of experiments, mice were killed with a lethal dose of pentobarbital (150 mg/kg, i.p.). The brains were immediately removed, and then stained with 2% 2, 3, 5-triphenyltetrazolium chloride to confirm the production of an infarct.

ter probe (MSP300XP, AD Instruments Inc.) attached to the left parietal cranium. At the end of experiments, mice were killed with a lethal dose of pentobarbital (150 mg/kg, i.p.). The brains were immediately removed, and then stained with 2% 2, 3, 5-triphenyltetrazolium chloride to confirm the production of an infarct. Neurological score In another set of experiments, the therapeutic effects of SC were assessed in mice receiving either 2 × 106 viable WT SC or Large T SC (in 150 μl of PBS) or PBS (150 μl) intravenously at 24 hours after the induction of cerebral ischemia. The neurological outcome was assessed at 1, 7 and 14 days after administration of SC using a 5-point scale neurological deficit score (0 = no deficit, 1 = failure to extend right paw, 2 = circling to the right, 3 = falling to the right, 4 = unable to walk spontaneously)[17].

rs after the induction of cerebral ischemia. The neurological outcome was assessed at 1, 7 and 14 days after administration of SC using a 5-point scale neurological deficit score (0 = no deficit, 1 = failure to extend right paw, 2 = circling to the right, 3 = falling to the right, 4 = unable to walk spontaneously)[17]. Recovery of Large T BMSC from ischemic brains Mice were sacrificed 14 days after BMSC administration with a lethal dose of pentobarbital (150 mg/kg, i.p.). Ischemic hemispheres were removed from recipient mice that received either WT or Large T BMSC. The infarcted cerebral hemisphere was cut into small pieces and incubated with 2% collagenase at 37°C for 2 hours. The collagenase treated hemispheres were centrifuged at 1500 RPM for 10 min. The supernatants were discarded and tissues were resuspended in IMDM with 10% FBS and 1% streptomycin/penicillin. The suspensions were filtered through a 70-μm cell strainer (BD Falcon) and the filtered fractions were plated into 75 cm2 flasks, cultured at 33°C in a mixture of 5% carbon dioxide and 95% oxygen, with the media replaced as needed. The cells isolated from ischemic brain tissue of mice receiving Large T BMSC reached confluency in 7 - 10 days. However, no cell growth was detected when infarcted tissue derived from mice receiving either WT BMSC or saline was cultured at 33°C under identical culture conditions.

gen, with the media replaced as needed. The cells isolated from ischemic brain tissue of mice receiving Large T BMSC reached confluency in 7 - 10 days. However, no cell growth was detected when infarcted tissue derived from mice receiving either WT BMSC or saline was cultured at 33°C under identical culture conditions. Immunostaining of Large T antigen (TAg) Immortalized stem cells (~104 cells) were adhered onto 1.2 cm diameter coverslips by centrifugation at 1,500 × g for 15 min in culture medium. Coverslips were then fixed in 1% paraformaldehyde in PBS (30 min), and extracted in 0.2% Triton-X100/PBS (5 min). Permeabilized coverslips were incubated for 1 h in 75 ul of mouse anti-SV40 large TAg (1:150), (pab416; ts A58, AbCAM, Cambridge, MA) in 0.1% milk powder in PBS (MPBS) at 25°C. Coverslips were washed 3-times in 0.1% MPBS, and then incubated with 75 ul of goat anti-mouse fluorescein conjugated antibody (1:50) in MPBS (1 h). After 3 washes in 0.1% MPBS, coverslips were mounted in 1:5 diluted Vectashield/DAPI (Vector Labs, Burlingame, CA) and sealed with nail polish. Using fluorescence microscopy (Olympus AX70 microscope), the BMSC were examined for fluorescein (SV40 large Tag) and nuclei (DAPI), and the images captured with a Nikon Coolpix camera.

washes in 0.1% MPBS, coverslips were mounted in 1:5 diluted Vectashield/DAPI (Vector Labs, Burlingame, CA) and sealed with nail polish. Using fluorescence microscopy (Olympus AX70 microscope), the BMSC were examined for fluorescein (SV40 large Tag) and nuclei (DAPI), and the images captured with a Nikon Coolpix camera. Gene microarray Pair-wise gene expression analysis was performed to compare the differences in gene expression patterns of naïve Large T (cell population injected into mice) with the Large T cell population isolated from infarcted tissue (iLarge T). RNA was extracted from cells using QIAshredder (Qiagen, Hilden, GmbH) and an RNeasy mini kit (Qiagen, Maryland), according to manufacturer's directions. RNA integrity was assessed by electrophoresis on the Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA). Double-stranded cDNA was synthesized from approximately 7 ug total RNA, using a Superscript cDNA Synthesis Kit (Invitrogen, Carlsbad, CA) in combination with a T7-(dT)24 primer (Proligo, Boulder, CO). Biotinylated cRNA was transcribed in vitro using the BioArray High Yield RNA Transcript Labeling Kit (ENZO Biochem, New York, NY) and purified using the GeneChip Sample Cleanup Module (Affymetrix, Santa Clara, CA). Twenty micrograms of purified cRNA was fragmented by incubation in fragmentation buffer (200 mM Tris-acetate, pH 8.1, 500 mM potassium acetate, 150 mM magnesium acetate) at 94°C for 35 minutes and chilled on ice. Ten micrograms of fragmented biotin-labeled cRNA was hybridized to the Mouse Genome 430 2.0 Array (Affymetrix), interrogating over 39,000 transcripts. Arrays were incubated for 16 hr at 45°C with constant rotation (60 rpm). The arrays were washed and then stained for 10 min at 25°C with 10 ug/mL streptavidin-R phycoerythrin (Vector Laboratories, Burlingame, CA) followed by 3 ug/mL biotinylated goat anti-streptavidin antibody (Vector Laboratories) for 10 minutes at 25°C. Arrays were then stained once again with streptavidin-R phycoerythrin for 10 min at 25°C. After washing and staining, the arrays were scanned using a GeneChip Scanner 3000. Pixel intensities were measured, expression signals were analyzed and features extracted using the commercial software package GeneChip Operating Software 1.2 (Affymetrix).

once again with streptavidin-R phycoerythrin for 10 min at 25°C. After washing and staining, the arrays were scanned using a GeneChip Scanner 3000. Pixel intensities were measured, expression signals were analyzed and features extracted using the commercial software package GeneChip Operating Software 1.2 (Affymetrix). Three independent sets of experiments were performed, each containing RNA samples pooled from BMSC populations isolated from 3 ischemic hemispheres or 3 sets of naïve Large T SC. Data mining and statistical analysis were performed with genesifter.net software. A two-fold or more change in gene expression with an unpaired t-test corrected with the Benjamini-Hochberg procedure[18], and a p < 0.05 was considered significant and used for further analysis. The Z-score was used to detect the most affected gene ontology families, with a high Z-score indicating a highly affected pathway[19]. Genes related to molecular function, localization and biological processes were analyzed by gene ontology detecting software http://www.genesifter.net.

nt and used for further analysis. The Z-score was used to detect the most affected gene ontology families, with a high Z-score indicating a highly affected pathway[19]. Genes related to molecular function, localization and biological processes were analyzed by gene ontology detecting software http://www.genesifter.net. Real time quantitative PCR Total RNA extracted from naïve and iLarge T SC was used as a template for cDNA synthesis. For each sample, 100 ng of total RNA was used as a template for cDNA synthesis. The reverse transcription reaction was performed in master mix containing RT buffer, MgCl2, dNTPs, random hexamers, RNase inhibitor and multiscribe reverse transcriptase (Applied Biosystems, Foster City, CA). Incubations were performed in a Mastercycler-personal (Eppendorf, Westbury, NY) for 10 min at room temperature followed by 60 min at 42 C and 5 min at 95 C. For quantitative real-time PCR analysis the primers listed in Table S1, Additional file 1 were used (Realtime Primers, Elkins Park, PA). Fast SYBR Green master mix was used for amplification and detection. Reactions were performed in triplicate using an ABI Prism 7900 Sequence Detection System. Raw data were analyzed using the ABI Prism Sequence Detection 1.9.1 software. The comparative Cr method for relative quantification of gene expression was used to determine expression levels for target genes. Beta-actin was used as a housekeeping gene.

ed in triplicate using an ABI Prism 7900 Sequence Detection System. Raw data were analyzed using the ABI Prism Sequence Detection 1.9.1 software. The comparative Cr method for relative quantification of gene expression was used to determine expression levels for target genes. Beta-actin was used as a housekeeping gene. Results Large-T BMSC improve neurological outcome after MCAO Intravenous administration of 2 × 10 6 Large T BMSC after 1 hour of cerebral ischemia and 24 hrs reperfusion blunted the neurological deficits (improved the neurological scores) normally noted in untreated mice two weeks thereafter (Figure 1). The protective effect of Large T BMSC was comparable to that observed in mice treated with BMSC isolated from wild type mice. Even though an improvement was observed with administration of WT BMSC or Large T BMSC at 7 days after ischemic stroke, a statistical difference was observed only at 14 days after BMSC administration. No difference in infarct volumes was observed among untreated stroke group versus the groups that received either WT-BMSC or Large-T BMSC (data not shown). These results indicate that Large T BMSC exhibit similar neuroprotective properties as BMSC isolated from WT mice. Figure 1 Neurological scores of sham, stroke, stroke +WT BMSC and stroke + Large T BMSC groups over a two week period after stroke or sham surgery. * represents significant difference from sham group (p < 0.05), # represents significant difference from stroke group (P < 0.05), one-way ANOVA with Tukey's post-hoc test.

Results Large-T BMSC improve neurological outcome after MCAO Intravenous administration of 2 × 10 6 Large T BMSC after 1 hour of cerebral ischemia and 24 hrs reperfusion blunted the neurological deficits (improved the neurological scores) normally noted in untreated mice two weeks thereafter (Figure 1). The protective effect of Large T BMSC was comparable to that observed in mice treated with BMSC isolated from wild type mice. Even though an improvement was observed with administration of WT BMSC or Large T BMSC at 7 days after ischemic stroke, a statistical difference was observed only at 14 days after BMSC administration. No difference in infarct volumes was observed among untreated stroke group versus the groups that received either WT-BMSC or Large-T BMSC (data not shown). These results indicate that Large T BMSC exhibit similar neuroprotective properties as BMSC isolated from WT mice. Figure 1 Neurological scores of sham, stroke, stroke +WT BMSC and stroke + Large T BMSC groups over a two week period after stroke or sham surgery. * represents significant difference from sham group (p < 0.05), # represents significant difference from stroke group (P < 0.05), one-way ANOVA with Tukey's post-hoc test. Viable Large-T BMSC can be recovered from post-ischemic brain tissue In order to analyze potential phenotypic changes in BMSC that are recruited into post-ischemic brain, we devised a novel approach for tissue isolation of BMSC that capitalized on the ability of Large T BMSC to grow under culture conditions at 33°C. Using this approach, we found that Large T BMSC harvested from post-ischemic brain tissue and then cultured at 33°C exhibit cell growth and colony formation, while WT BMSC did not exhibit these responses under identical experimental conditions. Immunocytochemical staining confirmed the presence of intracellular Large T antigen in all cells isolated using this procedure (Figure 2)

post-ischemic brain tissue and then cultured at 33°C exhibit cell growth and colony formation, while WT BMSC did not exhibit these responses under identical experimental conditions. Immunocytochemical staining confirmed the presence of intracellular Large T antigen in all cells isolated using this procedure (Figure 2) Figure 2 Large T BMSC isolated from ischemic brain. A. Nuclear DAPI Staining, B. Large T immunostaining of the same sample.

post-ischemic brain tissue and then cultured at 33°C exhibit cell growth and colony formation, while WT BMSC did not exhibit these responses under identical experimental conditions. Immunocytochemical staining confirmed the presence of intracellular Large T antigen in all cells isolated using this procedure (Figure 2) Figure 2 Large T BMSC isolated from ischemic brain. A. Nuclear DAPI Staining, B. Large T immunostaining of the same sample. BMSC isolated from post-ischemic brain tissue exhibit an altered gene expression pattern compared to their naive counterparts The GeneChip® Mouse Genome 430 2.0 array, covering over 39,000 transcripts, was used to compare gene expression between BMSC isolated from infarcted tissue (iLarge T BMSC) and naïve Large T SC. A pair-wise analysis of three independent experiments revealed a dramatically altered gene expression profile in iLarge T BMSC. Using filtering criteria of a two-fold or more change in gene expression with an unpaired t-test corrected with Benjamini-Hochberg procedure[18] (P < 0.05) a list of 1885 differentially expressed genes were detected (Figure 3). Of this total, 995 genes from iLarge T BMSC exhibited reduced expression (Table S2, Additional file 2), while 890 genes showed increased expression (Table S3, Additional file 3). Among the highly up-regulated genes (adjusted p < 0.01), endothelial specific cell molecule-1, bone morphogenic protein-2, nerve growth factor beta, olfactomedin-1 were detected. The gene microarray was validated with RT-PCR. A strong correlation (r2 = 0.93) between mRNA detected by RT-PCR and the gene microarray results was observed (Figure 4). The data discussed in this publication have been deposited in NCBI's Gene Expression Omnibus[20] and are accessible through GEO Series accession number GSE21393 http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE21393.

r2 = 0.93) between mRNA detected by RT-PCR and the gene microarray results was observed (Figure 4). The data discussed in this publication have been deposited in NCBI's Gene Expression Omnibus[20] and are accessible through GEO Series accession number GSE21393 http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE21393. Figure 3 Volcano graph of gene expression changes in BMSC isolated from ischemic brain (Group1) versus naïve BMSC (Group2). Figure 4 Fold-change in gene expression detected by RT-PCR versus microarray. Genes in Large-T BMSC that encode factors released into the extracellular space are highly induced by ischemia-reperfusion A gene ontology analysis was performed on Large T BMSC using Genesifter.net ontology tools. Z-scoring was used to identify the most affected pathways in isolated BMSC. A positive Z-score indicates that more genes than expected fulfilled the criterion for altered expression in a certain group or pathway; therefore, that group or pathway is likely to be affected by the imposed condition (e.g., ischemia)[21]. Z-scores were detected for up-regulated transcripts related to biological processes, cellular location and molecular function (Figure 5). A very high Z-score was detected for genes targeting the extracellular region. Figure 5 Z-scores of gene ontology groups that changed in BMSC isolated from ischemic brain.

Genes in Large-T BMSC that encode factors released into the extracellular space are highly induced by ischemia-reperfusion A gene ontology analysis was performed on Large T BMSC using Genesifter.net ontology tools. Z-scoring was used to identify the most affected pathways in isolated BMSC. A positive Z-score indicates that more genes than expected fulfilled the criterion for altered expression in a certain group or pathway; therefore, that group or pathway is likely to be affected by the imposed condition (e.g., ischemia)[21]. Z-scores were detected for up-regulated transcripts related to biological processes, cellular location and molecular function (Figure 5). A very high Z-score was detected for genes targeting the extracellular region. Figure 5 Z-scores of gene ontology groups that changed in BMSC isolated from ischemic brain. Since previous reports suggest that stem cells likely exert their beneficial effects through the secretion of trophic factors that enhance brain repair[22], [23], we focused our analysis on the extracellular factors that exhibit altered expression in BMSC isolated from post-ischemic brains. Eighty extracellular factors were increased in iLarge T BMSC, compared to their naïve counterparts (Table S4, Additional file 4). Among the extracellular factors affected, some have previously been reported as secreted by BMSC (red highlighting in Table S4, Additional file 4), while others (blue highlighting in Table S4, Additional file 4) have been associated with brain ischemia. A KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway analysis was performed by setting the absolute Z-score above 2 and number of genes in a set as ≥ 10 (Figure 6). This analysis revealed an up-regulation in genes related to the mitogen-activated protein kinase (MAPK) and axon guidance pathways, and a down-regulation of genes in cell division-related pathways. The KEGG diagram for axon guidance pathways in iLarge T BMSC were consistent with activation of a number of relevant genes, including Eph receptors (mediate neuronal branching), GTPase activators, and semaphorins (Figure 7).

) and axon guidance pathways, and a down-regulation of genes in cell division-related pathways. The KEGG diagram for axon guidance pathways in iLarge T BMSC were consistent with activation of a number of relevant genes, including Eph receptors (mediate neuronal branching), GTPase activators, and semaphorins (Figure 7). Figure 6 Changes in KEGG pathways in BMSC isolated from ischemic brain. Changes filtered for parameters Z-score > 2 and 10 or more changed genes in related group. Figure 7 Axon guidance KEGG pathway affected in BMSC isolated from ischemic brain. Genes affected represented in red.

) and axon guidance pathways, and a down-regulation of genes in cell division-related pathways. The KEGG diagram for axon guidance pathways in iLarge T BMSC were consistent with activation of a number of relevant genes, including Eph receptors (mediate neuronal branching), GTPase activators, and semaphorins (Figure 7). Figure 6 Changes in KEGG pathways in BMSC isolated from ischemic brain. Changes filtered for parameters Z-score > 2 and 10 or more changed genes in related group. Figure 7 Axon guidance KEGG pathway affected in BMSC isolated from ischemic brain. Genes affected represented in red. Discussion Stem cell therapy has received much attention as a potential post-injury intervention to repair stroke-damaged brain tissue. While there is limited clinical evidence that convincingly demonstrates stem cell mediated symptomatic relief in individuals who have suffered a stroke[24], animals studies suggest that stem cell therapy has the potential to reverse some of the behavioral deficits that result from ischemic stroke[1,3,4,25,26]. The success in animal models has led several laboratories to focus on potential mechanisms that may underlie the neuroprotective effects of stem cell therapy. A limitation of these mechanistic studies is an inability to isolate and characterize the population of stem cells that infiltrate post-ischemic brain tissue. Single cell isolation methods, such as laser capture, have provided some insights, but the low yield and poor survival of the recovered stem cells limits the utility of this approach. Here, we introduce a novel approach for selective isolation of BMSC from infarcted brain tissue. This approach yields viable infiltrating stem cells of sufficient number to allow for a detailed analysis of changes in gene expression.

eld and poor survival of the recovered stem cells limits the utility of this approach. Here, we introduce a novel approach for selective isolation of BMSC from infarcted brain tissue. This approach yields viable infiltrating stem cells of sufficient number to allow for a detailed analysis of changes in gene expression. Stem cells that infiltrate a brain infarct are believed to synthesize and secrete growth and guidance factors that are protective and orchestrate tissue recovery following stroke. However, the epigenetic changes that occur in these infiltrating stem cells due to environmental cues produced by the ischemic insult remain poorly understood. The identification of genes that exhibit significantly altered expression by protective tissue-infiltrating stem cells represents a powerful approach for discovery of factors produced by stem cells that enable them to mediate tissue restitution and repair following an ischemic stroke. Our analysis revealed that the expression of a large number of genes is dramatically and persistently increased in BMSC harvested from the ischemic infarct, when compared to their 'naïve' (non-migrated) precursors. Interestingly, many of the genes whose messages were most dramatically increased (> 100 fold) appear to play important roles in neural patterning, remodeling, adhesion and angiogenesis. Many other genes not previously associated with neuroprotection were also dramatically increased and these genes may play important supporting roles in tissue recovery from stroke.

s were most dramatically increased (> 100 fold) appear to play important roles in neural patterning, remodeling, adhesion and angiogenesis. Many other genes not previously associated with neuroprotection were also dramatically increased and these genes may play important supporting roles in tissue recovery from stroke. Different mechanisms have been proposed to explain the improved neurological outcome following stem cell treatment, including trans-differentiation into neural lineages, cell fusion, and neuroprotection through trophic support. A role for neuroprotection mediated by released trophic factors is supported by reports that describe the wide array of neuroprotective factors released from stem cells that could lead to improved neurological outcome after stroke by promoting physiological responses such as angiogenesis and/or neurogenesis, and inhibition of scar formation. More direct support is provided by Qu and coworkers[7], who identified, using gene microarray, the factors secreted by BMSC after exposure to ischemic brain extract. Their comparison of the gene expression profiles between naive BMSC and ischemic extract-exposed BMSC revealed an intense up-regulation of genes that encode extracellular factors, such as fibroblast growth factor-2, epidermal growth factor, nerve growth factor-beta, insulin-like growth factor 1, VEGF-A, transforming growth factor, beta 1, and brain derived neurotrophic factor[7]. Our analysis of BMSC that infiltrate post-ischemic brain tissue demonstrated gene activation for most of the same trophic agents (Table S4, Additional file 4). In our study, the expression of VEGF-A, VEGF-C and TGF-beta were all increased in BMSC isolated from post-ischemic brains, but these responses did not reach statistical significance in the microarray assay. However, consistent with previous reports[27], we found that angiopoietin-1 and 4 were highly expressed in BMSC isolated from post-ischemic brains. Hence, our results provide strong support for the contention that within the post-ischemic environment, BMSC release angiogenic and neurotrophic factors that may mediate the neuroprotection observed following stem cell therapy.

that angiopoietin-1 and 4 were highly expressed in BMSC isolated from post-ischemic brains. Hence, our results provide strong support for the contention that within the post-ischemic environment, BMSC release angiogenic and neurotrophic factors that may mediate the neuroprotection observed following stem cell therapy. Ddr2 (CD167b), a collagen adhesive receptor that participates in matrix integrin signaling[28], was increased 145-fold over naïve BMSC in our study, suggesting that Ddr2 matrix signals may contribute to post-stroke remodeling. ZIC-3 (Zinc finger protein-3) message was increased 26-fold over naïve BMSC. ZIC-3, a member of the C2H2-type zinc finger protein family, is a nuclear transcription factor that functions in left-right body axis alignment, and is a 'pluripotency' factor expressed during cell regeneration[29]. Zic3 also interacts with BMP and FGF signaling to direct neural cell programming[30]. We also found that infiltrating BMSC expressed 195-fold more cytokine receptor-like factor-1 (Crlf-1, cytokine-like factor 1, CLF-1, CRLM-3, cytokine receptor like molecule 3, NR6) transcript than naïve BMSC. CRLF1 forms a heterodimeric complex with cardiotrophin-like cytokine factor 1 (CLC-1), and the Crlf-1/CLC-1 heterodimer competes with ciliary neurotrophic factor for binding to the ciliary neurotrophic factor receptor (CNTFR) complex. CrLF-1 is a cytokine ligand related to IL-12 that supports differentiation and survival of a wide range of neural cell types during embryonic development and in adult neural tissues[31]. CrLF-1 mRNA is up-regulated by inflammatory cytokines e.g. TNF-α, IL-6, and IFN-γ which are elevated in post-ischemic brain tissue[32].

LF-1 is a cytokine ligand related to IL-12 that supports differentiation and survival of a wide range of neural cell types during embryonic development and in adult neural tissues[31]. CrLF-1 mRNA is up-regulated by inflammatory cytokines e.g. TNF-α, IL-6, and IFN-γ which are elevated in post-ischemic brain tissue[32]. Expression of FAM19A5 (also 'TAFA5') transcript was elevated 164-fold in brain penetrating BMSC, compared to naïve BMSC. FAM195 is a novel neuropeptide that is highly expressed in the CNS, particularly in hypothalamic paraventricular nuclear vasopressin and oxytocin cells[33]. FAM19a5 has been proposed to regulate brain fluid balance, and elevated levels of TAFA5 may thus help to control cerebral edema after stroke. Brain penetrating stem cells were found to express 101-fold more transcript for osteopontin (OPN) compared to naïve BMSC. OPN and thrombin generated OPN-peptides have all been shown to confer protection in stroke models[34,35]. Bayless has reported that OPN interacts with αvβ3 integrin and binds endothelial and smooth muscle cells in an RGD motif dependent manner[36]. OPN binding to α4β1 integrin has been implicated in the organization of endothelial cells in the developing vasculature, the extravasation of immune cells into tissues, and the emigration of neuroblasts[37]. Fibroblast growth factor-7 (FGF7 or KGF), which was increased 173-fold in our study, is known to mediate cell proliferation and motility, protect against cell death,[38] and has been shown to limit ischemia-induced neuronal death[39]. We also found a 44-fold increase in olfactomedin-1 (Olfm1), which has been implicated in neuronal differentiation, axon extension and cell survival[40].

old in our study, is known to mediate cell proliferation and motility, protect against cell death,[38] and has been shown to limit ischemia-induced neuronal death[39]. We also found a 44-fold increase in olfactomedin-1 (Olfm1), which has been implicated in neuronal differentiation, axon extension and cell survival[40]. Several genes that are linked to the Wnt signaling pathway were also profoundly upregulated in infiltrating BMSC. Glypican-1 (a GPI-anchored heparin sulfate proteoglycan) which interact with and suppresses hedgehog, stimulates the Wnt pathway, and binds BMP and FGF[41], exhibited a 155-fold increase. Dickkopf-2 (DKK2) expression was elevated 145-fold. Its homologue, Dickkopf-1 is a Wnt antagonist that contributes to neuronal apoptosis following brain ischemia[42]; DKK1 has been described as a target for treatment in neurodegenerative disorders e.g. beta-amyloid deposition, epilepsy, excitotoxicity. While DKK-1, (and Dkk4) block Wnt signaling, DKK2 and DKK3 do not,[43] and in some systems DKK2 actually synergizes with Wnt signals[44,45]. Therefore, in the setting of stroke recovery, elevated levels of secreted DKK2 might limit neuronal apoptosis to preserve neuron survival and improve tissue integrity.

totoxicity. While DKK-1, (and Dkk4) block Wnt signaling, DKK2 and DKK3 do not,[43] and in some systems DKK2 actually synergizes with Wnt signals[44,45]. Therefore, in the setting of stroke recovery, elevated levels of secreted DKK2 might limit neuronal apoptosis to preserve neuron survival and improve tissue integrity. Sushi-repeat containing protein (SRPX2), a ligand for urokinase-type plasminogen activator that interacts with cathepsins B and ADAMTS4 to control extracellular matrix remodeling[46], was elevated 33-fold in infiltrating BMSC compared naïve BMSC. SRPX2 participates in cell migration and adhesion through regulation of FAK phosphorylation[47]. It also contributes to the modulation of endothelial remodeling in angiogenesis[48], which may be a factor in the enhanced cerebral angiogenesis that is associated with stem cell therapy in ischemic brain disorders.

e BMSC. SRPX2 participates in cell migration and adhesion through regulation of FAK phosphorylation[47]. It also contributes to the modulation of endothelial remodeling in angiogenesis[48], which may be a factor in the enhanced cerebral angiogenesis that is associated with stem cell therapy in ischemic brain disorders. Several gene messages were found to be highly suppressed in BMSC isolated from the ischemic brain. Among these, basonuclin (Bnc1), a zinc-finger protein that is highly expressed in early keratinocytes[49], was down regulated 523-times. Down-regulation of this message may indicate a change in cell lineage fate within the ischemic environment. Glycoprotein m6a (Gpm6a) expression (reduced 157-fold) is associated with neuronal development and migration, and high levels of Gpm6a in stem cells has been linked to enhanced neuronal cell differentiation and migration[50]. We also noted a 34-fold suppression of breast cancer 1 gene (Brca1), a nuclear phosphoprotein that plays a role in maintaining genomic stability and in tumor suppression. Brca1 has been implicated in preventing apoptosis in early neuronal progenitors[51] and its expression in adult life is associated with Alzheimer's disease[52]. Smpd3 (reduced 24-fold) catalyzes the hydrolysis of sphingomyelin to form ceramide and phosphocholine. Ceramide mediates apoptosis and regulates the cell cycle by acting as a growth suppressor in confluent cells. Smpd3 also mediates cellular responses to IL-1ß and TNFα [53]. Tetraspanin 8 (Tspan8 or Tm4sf3) (reduced 23-fold), a member of the transmembrane 4 superfamily, is known to mediate signal transduction events that contribute to the regulation of cell development, activation, growth and motility[54]. Replication factor C (Rfc4) (9-fold reduction), which is required in the elongation of primed DNA templates by DNA polymerase δ and DNA polymerase ε[55], is believe to ensure error-free proliferation of stem cells at early phases of cell growth.

o the regulation of cell development, activation, growth and motility[54]. Replication factor C (Rfc4) (9-fold reduction), which is required in the elongation of primed DNA templates by DNA polymerase δ and DNA polymerase ε[55], is believe to ensure error-free proliferation of stem cells at early phases of cell growth. Astrocytes form glial scars along ischemic lesions and produce proteoglycans that inhibit axonal growth[56]. Suppression of inhibitory factors by cell-based therapies leads to axonal growth that correlate with improved functional outcome after stroke[57]. For example, it has been reported that bone-marrow mesenchymal cells reduce the expression of axonal-growth inhibitory proteins that are released by astrocytes, thereby allowing axon formation in the ischemic brain[58]. In our study, BMSC harvested from post-ischemic brain tissue exhibited a persistent and altered expression of several of genes that have been implicated in the regeneration and guidance of axons (Figure 7). Among these genes, Robo-1, EphA4, Pak1 and SLIT-Robo RhoGTPase activating protein exhibited the most significant up-regulation, while the expression of semaphorin 3D and ephrin-B1 were significantly reduced. These findings support the role of BMSC re-programming in axon formation and guidance following ischemic stroke.

Among these genes, Robo-1, EphA4, Pak1 and SLIT-Robo RhoGTPase activating protein exhibited the most significant up-regulation, while the expression of semaphorin 3D and ephrin-B1 were significantly reduced. These findings support the role of BMSC re-programming in axon formation and guidance following ischemic stroke. Several studies suggest that stem cells can also attenuate immune responses of the host[59-61]. Supporting this view are changes in gene expression related to immune processes in human BMSC transplanted into the mouse hippocampus 1 day after global ischemia[9]. The transplanted BMSC exhibited a change in gene ontology groups for carbohydrate binding, cell adhesion, basement membrane as well as antigen presentation and processing. The results of our analysis suggest that the changes in immune-related gene processes were small in the BMSC that infiltrate ischemic brain tissue, particularly in comparison with the responses noted for genes related to axon guidance, MAPK pathway and the cell cycle. While an explanation for the different gene expression responses between the two studies is not readily apparent, it may result from differences in ischemic model, route of administration of BMSC, and/or brain region studied.

with the responses noted for genes related to axon guidance, MAPK pathway and the cell cycle. While an explanation for the different gene expression responses between the two studies is not readily apparent, it may result from differences in ischemic model, route of administration of BMSC, and/or brain region studied. Conclusions In conclusion, using a novel approach for BMSC isolation from postischemic brain tissue, we found that BMSC assume a new and very different genetic profile that favors the secretion of numerous extracellular factors into ischemic brain tissue that have the potential to facilitate neuroprotective responses such as angiogenesis, and axonal guidance and regeneration. Our findings may help to explain the neuroprotective effects previously proposed for stem cells in ischemic stroke. Competing interests All authors declare no conflict of interest. This work was supported by a grant from the National Heart Lung and Blood Institute (R01 HL26441-29). Authors' contributions GY designed the study, performed animal and cell culture experiments, analyzed data, and wrote the paper. JSA generated H-2Kb-tsA58 stem cells, analyzed gene array data and wrote the paper. CEY performed immunohistochemistry and flow cytometry. DNG designed the study, provided lab facilities, helped to interpret data and wrote the manuscript. All authors read and approved the final manuscript. Supplementary Material Additional file 1 Table S1. RT PCR primers Click here for file Additional file 2 Table S2. Down-regulated transcripts in isolated BMSC from ischemic brain. Click here for file

Authors' contributions GY designed the study, performed animal and cell culture experiments, analyzed data, and wrote the paper. JSA generated H-2Kb-tsA58 stem cells, analyzed gene array data and wrote the paper. CEY performed immunohistochemistry and flow cytometry. DNG designed the study, provided lab facilities, helped to interpret data and wrote the manuscript. All authors read and approved the final manuscript. Supplementary Material Additional file 1 Table S1. RT PCR primers Click here for file Additional file 2 Table S2. Down-regulated transcripts in isolated BMSC from ischemic brain. Click here for file Additional file 3 Table S3. Up-regulated transcripts in isolated BMSC from ischemic brain. Click here for file Additional file 4 Table S4. Extracellular factors that are expressed in BMSC isolated from ischemic regions compared to naïve BMSC. Click here for file Acknowledgements We would like to thank Paula Polk and Jeff Houghton for providing technical assistance in microarray assays.

Background Stroke is the third most common cause of death, after heart attack and cancer, and it has profound negative social and economic effects. The current treatment for complete stroke is only partially successful at reversing neurodegeneration and restoring premorbid function. Since stroke is one of the principal etiologies for neurological sequelae and/or death, it is very important to understand both its pathologic mechanisms and any effective treatments. Understanding the cellular mechanisms of ischemia-reperfusion injury is of great importance for stroke therapy; however, the precise mechanisms still remain unclear. The identification and characterization of the ischemic area are valuable and useful for understanding the pathogenesis and for establishing new therapeutic strategies to treat or prevent brain ischemia.

-reperfusion injury is of great importance for stroke therapy; however, the precise mechanisms still remain unclear. The identification and characterization of the ischemic area are valuable and useful for understanding the pathogenesis and for establishing new therapeutic strategies to treat or prevent brain ischemia. Proteome analysis, an exhaustive analytical technique for analyzing proteins, is primarily conducted using ESI-LCMS or MALDI-TOFMS. Many proteins and peptides are extremely minute in quantity, requiring the enhanced sensitivity provided by mass spectrometers. Shimadzu's new LCMS-IT-TOF is a novel hybrid mass spectrometer that is applicable to a wide range of bioanalytical research, including biomarker discovery, metabolite identification, and drug development, among others. Coupling atmospheric pressure ionization with Ion-Trap (IT) and Time-of-Flight (TOF) technologies, the LCMS-IT-TOF delivers high mass accuracy and high mass resolution (10,000 at 1000 m/z) independent of MS mode. The LCMS-IT-TOF allows more qualitative information about a sample to be collected in a single run. This enables researchers and scientists to: 1. elucidate structures of new molecules; 2. identify impurities and contaminants; and 3. analyze metabolites and biomarkers to evaluate biological pathways. The LCMS-IT-TOF has been designed to maximize sensitivity and selectivity by optimizing the ion transport to the TOF analyzer and by redefining the capability of the quadrupole ion trap. The ion trap is used to focus ions prior to ejection into the TOF as well as to support MSn analysis with effective precursor ion selection capabilities (resolution > 1,000 at 1,000 m/z).

ity and selectivity by optimizing the ion transport to the TOF analyzer and by redefining the capability of the quadrupole ion trap. The ion trap is used to focus ions prior to ejection into the TOF as well as to support MSn analysis with effective precursor ion selection capabilities (resolution > 1,000 at 1,000 m/z). The quest for biological markers of disease is a challenge for investigators in many fields, and this is certainly the case in the study of neurologic disease. One of the goals of proteomics is to characterize cellular proteins, secreted proteins and peptides, and proteolytic fragments for use as potential biomarkers. The objective of the present study was to identify ischemia-related proteins using the new LCMS-IT-TOF to analyze the chronological change of protein peaks following ischemia/reperfusion. Methods Animals The experimental designs and all procedures were in accordance with the U.S. National Institutes of Health Guide for the Care and Use of Laboratory Animals and Animal Care Guidelines issued by the Animal Experimental Committee of Gifu Pharmaceutical University. All experiments were performed using male ddY mice (4-5 weeks old, Japan SLC, Ltd., Shizuoka, Japan). Every effort was made to minimize the number of animals used and their suffering.

and Use of Laboratory Animals and Animal Care Guidelines issued by the Animal Experimental Committee of Gifu Pharmaceutical University. All experiments were performed using male ddY mice (4-5 weeks old, Japan SLC, Ltd., Shizuoka, Japan). Every effort was made to minimize the number of animals used and their suffering. Drugs Formic acid (HCOOH) and trifluoroacetic acid (TFA) were purchased form Wako pure chemical (Osaka, Japan). HPLC-grade ACN (CH3CN) and water were also obtained from Wako pure chemical. A Liquid Tissue MS Protein Kit was obtained from Expression Pathology Inc. (Gaithersburug, MD, USA). Pentobarbital sodium and isoflurane were purchased from Nissan Kagaku (Tokyo, Japan) and Merck Hoei Ltd. (Osaka, Japan), respectively.

Japan). HPLC-grade ACN (CH3CN) and water were also obtained from Wako pure chemical. A Liquid Tissue MS Protein Kit was obtained from Expression Pathology Inc. (Gaithersburug, MD, USA). Pentobarbital sodium and isoflurane were purchased from Nissan Kagaku (Tokyo, Japan) and Merck Hoei Ltd. (Osaka, Japan), respectively. Focal cerebral ischemia model in mice Anesthesia was induced using 2.0 to 3.0% isoflurane and maintained using 1.0 to 1.5% isoflurane (both in 70% N2O/30% O2) by means of an animal general anesthesia machine (Soft Lander; Sin-ei Industry Co. Ltd., Saitama, Japan). Body temperature was maintained at 37.0 to 37.5°C with the aid of a heating pad and heating lamp. After a midline skin incision, the left external carotid artery was exposed, and its branches were occluded [1,2]. An 8-0 nylon monofilament (Ethicon, Somerville, NJ, USA) coated with a mixture of silicone resin (Xantopren; Bayer Dental, Osaka, Japan) was introduced into the left internal carotid artery through the external carotid artery stump so as to occlude the origin of the middle cerebral artery. Then, the left common carotid artery was occluded. After 2 h of occlusion, the animal was reanesthetized briefly and reperfusion initiated via withdrawal of the monofilament. After surgery, the mice were kept in the preoperative condition (24 ± 2°C) until sampling. To confirm the induction of MCAO, a laser-Doppler flowmetry (Omegaflow flo-N1; Omegawave Inc., Tokyo, Japan) measured the regional artery blood flow (rCBF) in the MCA territory from the temporal bone surface. Mice that did not demonstrate a significant reduction just before reperfusion (to less than 40% that of the contralateral rCBF values) were excluded.

ser-Doppler flowmetry (Omegaflow flo-N1; Omegawave Inc., Tokyo, Japan) measured the regional artery blood flow (rCBF) in the MCA territory from the temporal bone surface. Mice that did not demonstrate a significant reduction just before reperfusion (to less than 40% that of the contralateral rCBF values) were excluded. Tissue micro dissection The left ventricle was perfused with 4% paraformaldehyde in 0.1 M phosphate buffer (PB; pH 7.4). Brains were removed after 15 min perfusion fixation at 4°C, then immersed in the same fixative solution overnight at 4°C. They were then immersed in 25% sucrose in 0.1 M PB for 24 h, and finally frozen in liquid nitrogen. Coronal sections (14 μm thick) were cut on a cryostat at -20°C, and the ischemic core area and a non-ischemic area (2 mm × 2 mm) were excised and placed in clear 1.5 ml tubes. Protein Extraction Micro dissected specimens were processed by reagents according to the manufacturer's recommendations (Expression Pathology Inc., Rockville, MD, USA). Briefly, material was suspended in 20 μl of reaction buffer, incubated at 95°C for 90 min, then cooled on ice for 2 min, at which time 1 μl of trypsin was added followed by incubation at 37°C for 18 h. Dithiothreitol was added to a final concentration of 10 mM, and the sample were heated for 5 min at 95°C to reduce cysteine residues. Finally, Liquid Tissue MS Protein Kit extracts were desalted with C-18 Zip Tip microcolumns (Millipore, Bedford, MA, USA), lyophilized, and resuspended in a minimal amount of 0.1% formic acid in 2% ACN.

as added to a final concentration of 10 mM, and the sample were heated for 5 min at 95°C to reduce cysteine residues. Finally, Liquid Tissue MS Protein Kit extracts were desalted with C-18 Zip Tip microcolumns (Millipore, Bedford, MA, USA), lyophilized, and resuspended in a minimal amount of 0.1% formic acid in 2% ACN. NanoRPLC-MS/MS Analysis NanoRPLC was performed using a DiNa-2A nanoLC system (KYA Technologies, Tokyo, Japan) coupled online to a LCMS-IT-TOF (Shimadzu, Kyoto, Japan). NanoRPLC separation was performed using a Pico Frit Beta Basic C18 column (New Objective, Woburn, MA, USA) at a constant flow rate of 300 nl/min. After injection of 1 μl of sample, peptides were eluted using gradients of 5-40% solvent B (0.1% formic acid in 80% ACN)/0-30 min, 40-100% solvent B/30-40 min, and 100-100% solvent B/40-60 min. LCMS-IT-TOF was operated in the data-dependent MS/MS mode. The heated capillary temperature and electrospray voltage were set at 200°C and 2.3KV, respectively. Data were collected at scan ranges of 400-1500 for MS and 50-1500 for MS/MS. MS/MS spectra analysis MS/MS spectra were searched against the NCBI database using Mascot (Matrix Science Ltd, London, UK) with a peptide mass tolerance of ± 0.05 Da and a fragment mass tolerance of ± 0.05 Da.

NanoRPLC-MS/MS Analysis NanoRPLC was performed using a DiNa-2A nanoLC system (KYA Technologies, Tokyo, Japan) coupled online to a LCMS-IT-TOF (Shimadzu, Kyoto, Japan). NanoRPLC separation was performed using a Pico Frit Beta Basic C18 column (New Objective, Woburn, MA, USA) at a constant flow rate of 300 nl/min. After injection of 1 μl of sample, peptides were eluted using gradients of 5-40% solvent B (0.1% formic acid in 80% ACN)/0-30 min, 40-100% solvent B/30-40 min, and 100-100% solvent B/40-60 min. LCMS-IT-TOF was operated in the data-dependent MS/MS mode. The heated capillary temperature and electrospray voltage were set at 200°C and 2.3KV, respectively. Data were collected at scan ranges of 400-1500 for MS and 50-1500 for MS/MS. MS/MS spectra analysis MS/MS spectra were searched against the NCBI database using Mascot (Matrix Science Ltd, London, UK) with a peptide mass tolerance of ± 0.05 Da and a fragment mass tolerance of ± 0.05 Da. Western blotting Mice were deeply anesthetized and decapitated at 8 h after reperfusion (Figure 1), brains were quickly removed, and the left hemispheres were cut into 3 mm coronal sections (between 4 and 7 mm from the frontal forebrain). Brain samples were divided into cortex and striatum and each sample was homogenized in 10 ml/g tissue ice-cold lysis buffer [50 mM Tris-HCl (pH8.0) containing 150 mM NaCl, 50 mM EDTA, 1% Triton X-100, and protease/phosphatase inhibitor mixture] and centrifuged at 14,000 × g for 40 min at 4°C. An aliquot of 5 μg of protein was subjected to 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and then separated proteins were transferred onto a polyvinylidene difluoride membrane. For immunoblotting, the following primary antibodies were used: mouse anti-Rab33B monoclonal antibody (1 μg/mL: Frontier Science, Hokkaido, Japan), monoclonal anti-β-actin (1:1000 dilution: Sigma-Aldrich Co., St. Louis, MO, USA). The secondary antibody was anti-mouse HRP-conjugated IgG (1:2000 dilution, Pierce Biotechnology, Rockford, IL, USA). The immunoreactive bands were visualized using SuperSignal West Femto maximum sensitivity substrate (Pierce Biotechnology). The band intensity was measured using a Lumino imaging analyzer (LAS-4000: Fuji Film, Tokyo, Japan).

y was anti-mouse HRP-conjugated IgG (1:2000 dilution, Pierce Biotechnology, Rockford, IL, USA). The immunoreactive bands were visualized using SuperSignal West Femto maximum sensitivity substrate (Pierce Biotechnology). The band intensity was measured using a Lumino imaging analyzer (LAS-4000: Fuji Film, Tokyo, Japan). Figure 1 Protocol of all experiments. The sampling time are mentioned. See in method section. Immunohistochemistry At 1, 3, 6, and 10 h after the 2 h ischemia-reperfusion treatment (Figure 1), mice were injected with sodium pentobarbital (Nembutal; 50 mg/kg, i.p.), then perfused through the left ventricle with 4% paraformaldehyde in 0.1 M phosphate buffer (PB; pH 7.4). Brains were removed after 15 min perfusion fixation at 4°C, then immersed in the same fixative solution overnight at 4°C. They were then immersed in 25% sucrose in 0.1 M PB for 24 h, and finally frozen in liquid nitrogen. Coronal sections (14 μm thick) were cut on a cryostat at -20°C and stored at -80°C until use.

.4). Brains were removed after 15 min perfusion fixation at 4°C, then immersed in the same fixative solution overnight at 4°C. They were then immersed in 25% sucrose in 0.1 M PB for 24 h, and finally frozen in liquid nitrogen. Coronal sections (14 μm thick) were cut on a cryostat at -20°C and stored at -80°C until use. Sections were rinsed three times in PBS, incubated in 3% H2O2 in PBS for 30 min, then placed in PBS and blocked with 1% mouse serum for 30 min. A mouse monoclonal antibody against Rab33b (3 μg/mL: Frontier Science) was applied to sections overnight at 4°C. Secondary antibody (M.O.M. biotinylated anti-mouse) was applied for 10 min. The avidin/biotinylated horseradish peroxidase complex (ABC Elite kit; Vector Laboratories, Burlingame, CA, USA) was applied for 30 min, and the sections were allowed to develop the chromogen in 3,3-diaminobenzidine nickel solution for 2 min. Rab33B-stained cells were counted in the ischemic area.

s applied for 10 min. The avidin/biotinylated horseradish peroxidase complex (ABC Elite kit; Vector Laboratories, Burlingame, CA, USA) was applied for 30 min, and the sections were allowed to develop the chromogen in 3,3-diaminobenzidine nickel solution for 2 min. Rab33B-stained cells were counted in the ischemic area. Double-immunostaining For immunohistochemistry, mice were anesthetized with sodium pentobarbital and their brains were perfused with 4% paraformaldehyde at 8 h after the 2 h ischemia-reperfusion treatment (Figure 1). The perfused brains were dissected out, fixed in 4% paraformaldehyde for overnight, and frozen. Coronal brain sections (14 μm) were cut on a cryostat. For immunofluorescent double staining, sections were incubated overnight at 4°C with the primary antibodies: anti-Rab33B antibody (3 μg/mL: Frontier science), or anti- Iba-1 antibody (1:500 dilution: Wako pure chemical). Then, they were incubated for 3 h with Alexa Fluor 488 F (ab')2 fragment of goat anti-rabbit IgG (H+L) antibody, Alexa Fluor 546 F (ab')2 fragment of goat anti-mouse IgG (H+L) antibody. Cell counting To quantify Rab33B-positive cells after ischemia-reperfusion, the number of Rab33B-positive cells in the ischemic striatum; two areas, the superior and inferior areas were counted in a high-power field (×200) randomly chosen in a section through the anterior commissure.

Double-immunostaining For immunohistochemistry, mice were anesthetized with sodium pentobarbital and their brains were perfused with 4% paraformaldehyde at 8 h after the 2 h ischemia-reperfusion treatment (Figure 1). The perfused brains were dissected out, fixed in 4% paraformaldehyde for overnight, and frozen. Coronal brain sections (14 μm) were cut on a cryostat. For immunofluorescent double staining, sections were incubated overnight at 4°C with the primary antibodies: anti-Rab33B antibody (3 μg/mL: Frontier science), or anti- Iba-1 antibody (1:500 dilution: Wako pure chemical). Then, they were incubated for 3 h with Alexa Fluor 488 F (ab')2 fragment of goat anti-rabbit IgG (H+L) antibody, Alexa Fluor 546 F (ab')2 fragment of goat anti-mouse IgG (H+L) antibody. Cell counting To quantify Rab33B-positive cells after ischemia-reperfusion, the number of Rab33B-positive cells in the ischemic striatum; two areas, the superior and inferior areas were counted in a high-power field (×200) randomly chosen in a section through the anterior commissure. Statistical analysis Data are presented as the means ± S.E.M. Statistical comparison were made using a one-way ANOVA followed by Student's t-test or Dunett's test using Statview version 5.0 (SAS Institute Inc., Cary, NC, USA), with p < 0.05 considered statistically significant.

Cell counting To quantify Rab33B-positive cells after ischemia-reperfusion, the number of Rab33B-positive cells in the ischemic striatum; two areas, the superior and inferior areas were counted in a high-power field (×200) randomly chosen in a section through the anterior commissure. Statistical analysis Data are presented as the means ± S.E.M. Statistical comparison were made using a one-way ANOVA followed by Student's t-test or Dunett's test using Statview version 5.0 (SAS Institute Inc., Cary, NC, USA), with p < 0.05 considered statistically significant. Results NanoRPLC-MS/MS Analysis In the control group, 27 proteins were identified from NanoRPLC-MS/MS analysis (Figure 2A and 2B). In an ischemia group at 8 h after 2 h ischemia-reperfusion (Figure 1), NanoRPLC-MS/MS analysis etected 18 proteins from the ischemic area (Figure 2B and 2C). From the NanoRPLC-MS/MS analysis, we identified nine unique proteins that were detected only in the ischemic area (Figure 2C). Figure 2 Proteins identified with NanoRPLC-MS/MS Analysis. (A) 18 proteins identified from control group only. (B) 9 proteins identified from both control and ischemia group. (C) 9 proteins identified from ischemia group only.

Results NanoRPLC-MS/MS Analysis In the control group, 27 proteins were identified from NanoRPLC-MS/MS analysis (Figure 2A and 2B). In an ischemia group at 8 h after 2 h ischemia-reperfusion (Figure 1), NanoRPLC-MS/MS analysis etected 18 proteins from the ischemic area (Figure 2B and 2C). From the NanoRPLC-MS/MS analysis, we identified nine unique proteins that were detected only in the ischemic area (Figure 2C). Figure 2 Proteins identified with NanoRPLC-MS/MS Analysis. (A) 18 proteins identified from control group only. (B) 9 proteins identified from both control and ischemia group. (C) 9 proteins identified from ischemia group only. Western blotting for Rab33B Of these nine unique proteins (Figure 2C), we focused on RAB33B, a member of the RAS oncogene family. First, we examined whether Rab33B was up-regulated in an ischemic group by western blotting. Rab33B was increased in the ischemic striatum, but not in the ischemic cortex (Figure 3A). In quantitative analysis, Rab33B was present at significantly greater levels in the ischemic striatum than in the control striatum (Figure 3B), but no difference was noted between the ischemic cortex and the control cortex (Figure 3C). Figure 3 Western blot analysis of Rab33B in ischemic area at 8 h after 2 h ischemia -reperfusion. (A) Representative band image showed Rab33B expression of ischemia (I) group and control (C) group. (B) and (C) Rab33B expressions were quantified by densitometry and corrected by reference to β-actin. (* P < 0.05, t-test) Data are shown as mean ± S.E.M. (n = 4 or 5).

chemic area at 8 h after 2 h ischemia -reperfusion. (A) Representative band image showed Rab33B expression of ischemia (I) group and control (C) group. (B) and (C) Rab33B expressions were quantified by densitometry and corrected by reference to β-actin. (* P < 0.05, t-test) Data are shown as mean ± S.E.M. (n = 4 or 5). Expression of Rab33B after ischemia Next, we investigated changes in Rab33B expressions at 1, 3, 6, and 10 h after reperfusion by immunohistochemical analysis (Figure 4B). Quantitative analyses of Rab33B-positive cells are shown in Figure 4C. In the ischemic area, the number of Rab33B-positive cells was increased in a time-dependent manner from 1 h to 10 h following the 2 h ischemia-reperfusion treatment. No Rab33B-positive cells were detected in non-ischemic areas. Figure 4 Change in levels of Rab33B expression in the ischemic area after 2 h ischemia-reperfusion. (A) Schematic drawing showing brain regions at 0.4-1.0 mm anterior to bregma (through the anterior commissure). (B) Immunostaining for Rab33B at 1 h, 3 h, 6 h, and 10 h after reperfusion in mice. Scale bar = 50 μm. (C) The number of Rab33B-positive cells in the ischemic striatum. Values are shown the number of Rab33B-positive cells/mm2. Data are shown as mean ± S.E.M. (n = 4). ** P < 0.01 vs. control (Dunnett's test).

mmissure). (B) Immunostaining for Rab33B at 1 h, 3 h, 6 h, and 10 h after reperfusion in mice. Scale bar = 50 μm. (C) The number of Rab33B-positive cells in the ischemic striatum. Values are shown the number of Rab33B-positive cells/mm2. Data are shown as mean ± S.E.M. (n = 4). ** P < 0.01 vs. control (Dunnett's test). Localization of Rab33B after ischemia To identify Rab33B-positive cells, double immunofluorescence was performed for Rab33B and Iba-1 which is the marker of activated microlgia at 8 h after the 2 h ischemia-reperfusion. Iba-1-Positive microglias were expressed with Rab33b together in ischemic striatum (Figure 5B-b), but not in non-ischemic area (Figure 5B-a). Figure 5 Localization of Rab33B in ischemic area. (A) Schematic drawing showing brain regions at 0.4-1.0 mm anterior to bregma (through the anterior commissure): Measurement areas (a and b) in figure 5A in the contralateral non-ischemic and ischemic core, respectively. (B) Double-immunostaining of Rab33B and Iba-1. Rab33B expressed in microglia at 8 h after the ischemia reperfusion. (a) contralateral side (non-ischemic area). (b) ipsilateral side (ischemic area). Scale bar = 20 μm.

urement areas (a and b) in figure 5A in the contralateral non-ischemic and ischemic core, respectively. (B) Double-immunostaining of Rab33B and Iba-1. Rab33B expressed in microglia at 8 h after the ischemia reperfusion. (a) contralateral side (non-ischemic area). (b) ipsilateral side (ischemic area). Scale bar = 20 μm. Discussion In the present study, we identified ischemia-related proteins using an LCMS-IT-TOF to analyze the chronological changes in protein peaks following ischemia/reperfusion treatments in mice. LCMS analysis is the useful instrument which is able to detect the candidates of the protein markers for cerebral ischemia. For analyzing the peptides by LCMS, peptides which were obtained by extracting proteins from tissue samples and trypsin digestion were used. There are differences between control and ischemia brain groups in the amount of protein and peptide. And the LCMS detects from the peptide fragments in decreasing order. In the present study, beta-actin of house-keeping gene was detected in only ischemia group due to their differences. The beta-actin was expressed in both ischemia group and control groups, and the expression level was no significant change (Figure 3A). Therefore, the beta-actin may play a role as a house-keeping gene in the present study.

, beta-actin of house-keeping gene was detected in only ischemia group due to their differences. The beta-actin was expressed in both ischemia group and control groups, and the expression level was no significant change (Figure 3A). Therefore, the beta-actin may play a role as a house-keeping gene in the present study. We found an up-regulation of nine unique proteins in the ischemic brain areas. Among these proteins, we identified three Rab proteins. In the Rab family, the past reports concerning Rab33B were few and its function is not well known. Therefore, we focused on Rab33B in the present study. The Rab family of proteins is a member of the Ras superfamily of monomeric G proteins [3]. At least 60 Rab genes are found in the human genome and are widely distributed over the human chromosomes [4]. Rab GTPases have emerged as central regulators of vesicle budding, motility, and fusion and a number of these proteins are conserved from yeast to humans. Like other regulatory GTPases, the Rab proteins switch between two distinct conformations, one GTP-bound and the other GDP-bound. The GTP-bound conformation is generally regarded as the 'active' form, as it is the one that interacts with downstream effector proteins [5,6].

e proteins are conserved from yeast to humans. Like other regulatory GTPases, the Rab proteins switch between two distinct conformations, one GTP-bound and the other GDP-bound. The GTP-bound conformation is generally regarded as the 'active' form, as it is the one that interacts with downstream effector proteins [5,6]. In the present study, we focused on Rab33B, which was identified in ischemic areas by proteome analysis. Rab33B is one of two members of the Rab33 subfamily. In contrast to its neural and immune cell homolog, Rab33B is ubiquitously expressed in mammalian tissues and is localized to the medial golgi [7]. Rab33B has been described as a Golgi-resident protein which is involved in Golgi-to-endoplasmic reticulum transport and in modulation of autophagy [8-10].

e Rab33 subfamily. In contrast to its neural and immune cell homolog, Rab33B is ubiquitously expressed in mammalian tissues and is localized to the medial golgi [7]. Rab33B has been described as a Golgi-resident protein which is involved in Golgi-to-endoplasmic reticulum transport and in modulation of autophagy [8-10]. Autophagy is an intracellular pathway that is activated in response to cell stress. It is a phenomenon in which the cytoplasmic organelles in the cell are engulfed by double membrane vesicles, the autophagosomes, and are delivered to the lysosomes. There, the organelle proteins are broken down by lysosomal proteases and the resulting amino acids are recycled back into the cell machinery to aid in cell survival [11,12]. A number of key autophagy (Atg) proteins are apparently involved in this process [10,13]. Rab33B directly interacts with one of these, Atg16L, in a GTP-dependent manner. Therefore, activation and inactivation of Rab33B can modulate autophagy [7,8]. Autophagy appears to be a vital event in the development of the central nervous system [14,15] and is also constitutively active in healthy neurons, where it aids in survival [16].

eracts with one of these, Atg16L, in a GTP-dependent manner. Therefore, activation and inactivation of Rab33B can modulate autophagy [7,8]. Autophagy appears to be a vital event in the development of the central nervous system [14,15] and is also constitutively active in healthy neurons, where it aids in survival [16]. In the present study, we examined whether Rab33B colocalized with LC3 of autophagy marker. However, we could not detect the expression of LC3 in ischemic area at 8 h after 2 h ischemia-reperfusion (data not shown). It is reported that the autophagy was detected in ischemic area with a peak at 1 day after the transient MCAO [17]. In the present study, we focused on the expressions of the ischemia related proteins at early phase after ischemia-reperfusion. Further studies are needed to clarify the involvement of Rab33B to autophagy after cerebral ischemia, especially later phase than the present study.

eak at 1 day after the transient MCAO [17]. In the present study, we focused on the expressions of the ischemia related proteins at early phase after ischemia-reperfusion. Further studies are needed to clarify the involvement of Rab33B to autophagy after cerebral ischemia, especially later phase than the present study. Microglia, which is one of the glial cells, is approximately 20% of the total glial cell population within the brain. Unlike astrocyte, individual microglias are distributed in large non-overlapping regions throughout the brain and spinal cord [18]. Activation of microglia commonly occurs in the early response of the CNS to a wide variety of pathological stimuli, such as neurological degeneration, inflammation, and ischemia [19-21]. It is unclear that microglias are necessarily damaged following brain ischemia, but recent evidences suggest that activated microglia may contribute to the injury. Some neuroprotective drugs, such as minocycline and edravone, improved the stroke outcomes by inhibiting the microglial activation [22-24]. Microglia exerts a cytotoxic function by releasing reactive oxygen species, nitric oxide, and inflammatory cytokines, which trigger the neuronal damage [25-27]. Several studies have suggested that microglia may be rapidly and time-dependently activated after ischemia, and the microglial activation may reflect the extent of severity of ischemic injury. In the present study, an up-regulation and a time-dependent increase in Rab33B was observed in ischemic striatum (primarily in the ischemic core area) and was colocalized with Iba-1 positive activated microglia. Taken together, Rab33B may contribute to the neuronal cell death in the ischemic core area.

ischemic injury. In the present study, an up-regulation and a time-dependent increase in Rab33B was observed in ischemic striatum (primarily in the ischemic core area) and was colocalized with Iba-1 positive activated microglia. Taken together, Rab33B may contribute to the neuronal cell death in the ischemic core area. Conclusions These findings suggest that Rab33B is involved in the mechanism of pathogenesis of transient cerebral ischemia in mice and that it may be contribute to the neuronal cell death. Thus, control of the Rab33B gene and/or protein after ischemia may prove to be a useful strategy for therapeutic treatment of ischemic stroke. Further studies are needed to clarify the function of Rab33B in cerebral ischemia. Competing interests The authors declare that they have no competing interests. Authors' contributions HH participated in the design of the study. KT and MS analyzed the data. KH and JH performed the study. KH, JH and HH wrote the paper. All authors read and approved the final manuscript.

Background Unless immediately recognized and treated, heat stroke is often lethal, and victims who do survive may sustain permanent neurological damage [1]. The clinical diagnosis of heat stroke is demonstrated when hyperthermia is accompanied with circulatory shock (arterial hypotension), intracranial hypertension, and cerebral ischemia and injury [2,3]. Meanwhile, the heat stroke-induced central nervous system dysfunction includes delirium, convulsion, or coma [4]. Hence, prolonging survival time in heat stroke victims may offer more sufficient time for urgent treatment, thereby ameliorating the heat stroke-induced damage.

sion, and cerebral ischemia and injury [2,3]. Meanwhile, the heat stroke-induced central nervous system dysfunction includes delirium, convulsion, or coma [4]. Hence, prolonging survival time in heat stroke victims may offer more sufficient time for urgent treatment, thereby ameliorating the heat stroke-induced damage. Several lines of evidence indicate that rodents share with humans almost the same heat stroke syndromes, such as arterial hypotension, activated inflammation, and multiorgan dysfunction (in particular, cerebral ischemia, injury, and dysfunction [5-7]. In the rodents heat stroke model, significant decrements in both mean arterial pressure (MAP) and cerebral blood flow (CBF), but increments of cerebral monoamines levels, free radical productions and systemic cytokine levels are obtained in urethane-anaesthetized rats after heat stroke [8,9]. These pathophysiological changes are known to aggravate the conditions of cerebral ischemia and neuronal damage during heat stroke in rats [10]. Activated inflammation is evidenced by overproduction of pro-inflammatory cytokines (e.g., interleukin-1β (IL-1β) and tumor necrosis factor-α (TNF-α)) in circulation of heat stroke rats [11,12]. High levels of cytokines and radicals in the peripheral blood stream, as well as excessive accumulation of glutamate, hydroxyl radicals, dopamine (DA) and serotonin (5-HT) in the central brain, correlate with the severity of circulatory shock, cerebral ischemia and neuronal damage during heat stroke in rats [6,9,13,14].

igh levels of cytokines and radicals in the peripheral blood stream, as well as excessive accumulation of glutamate, hydroxyl radicals, dopamine (DA) and serotonin (5-HT) in the central brain, correlate with the severity of circulatory shock, cerebral ischemia and neuronal damage during heat stroke in rats [6,9,13,14]. Various clinical and experimental investigations have shown that single doses of dexamethasone (DXM; exogenous glucocorticoids) or hydroxyethyl starch (HES) are extensively used in the treatment of cerebral ischemia and/or cerebral injury [15-17]. In the studies of heat stroke, systemic treatment with DXM attenuated serum IL-1β levels and cerebral ischemia damage, and improved survival in heat stroke [18]. Additionally, the prolongation of survival in rats with HES therapy was found to be associated with augmentation of both arterial blood pressure and CBF as well as reduction of cerebral ischemia, hypoxia, and neuronal damage during heat stroke [19]. Although, many therapeutic agents show potential promise in many animal models, the results of most single-agent clinical trials are sobering. Consequently, various authors advocate studies to estimate the efficacy of combined therapeutic approaches [20,21]. Furthermore, there is less attention to evaluate immediate effects of both DXM and HES (the combined agent) on heat stroke-induced pathophysiological changes, let alone their neuroprotective underlying influences, especially in the aspects of striatal monoamines and hydroxyl radical production release. Based on these concepts, we propose whether application of the combined agent immediate treatment has efficacy to elongate the survival time, and improve the heat stroke-induced circulatory shock, cerebral ischemia, and neuronal damage in rats. Furthermore, we also attempt to ascertain whether the neuroprotective effects of the combined agent treatment are associated with inhibition of cerebral release of glutamate, DA, 5-HT, hydroxyl radicals, and the serum IL-1β, TNF-α and malondialdehyde (MDA) levels during heat stroke.

chemia, and neuronal damage in rats. Furthermore, we also attempt to ascertain whether the neuroprotective effects of the combined agent treatment are associated with inhibition of cerebral release of glutamate, DA, 5-HT, hydroxyl radicals, and the serum IL-1β, TNF-α and malondialdehyde (MDA) levels during heat stroke. Methods Experimental animals Male Spraque-Dawley rats weighing between 300 and 350 g were obtained from the National Science Council of Republic of China (Taiwan). Between experiments the animals were housed individually at an ambient temperature of 24 ± 1°C with a 12-h light-dark cycle, with the lights being switched on at 0600 h. Animal chow and water were allowed ad libitum. All protocol were approved by the Animal Ethics Committee of the Chia-Nan University of Pharmacy and Science, Tainan, Taiwan (approbated no. CN-IACUC-94007) in accordance with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and the guidelines of the Animal Welfare Act. Adequate anesthesia was maintained to abolish the corneal reflex and pain reflexes by tail-pinching throughout all experiments (approximately 8 hr) by a single intraperitoneal dose of urethane (1.4 g.kg-1 b.w., i.p.). At the end of the experiments, control rats (and any rats that had survived heat stroke) were killed with an overdose of urethane. One hundred thirty-eight rats were used in this study. Fifty-three rats of 138 were used for examining in Table 1, Figure 1 and 2 (three premature deaths during heat stroke induction and two premature deaths during animal surgery). Forty-three rats of 138 were used for examining in Table 2 (three premature deaths during heat stroke induction). Forty-two rats of 138 were used for examining in Figure 3 and 4 (two premature deaths during heat stroke induction). No premature deaths during anesthesia.

n and two premature deaths during animal surgery). Forty-three rats of 138 were used for examining in Table 2 (three premature deaths during heat stroke induction). Forty-two rats of 138 were used for examining in Figure 3 and 4 (two premature deaths during heat stroke induction). No premature deaths during anesthesia. Table 1 Effects of heat exposure (HE; Ta = 42°C) on both latency for the onset of heat stroke and survival time in rats treated with normal saline (NS), in rats treated with hydroxyethyl starch (HES), in rats treated with dexamethasone (DXM), and in rats treated with HES +DXM. Treatment Latency (mins) Survival time (mins) 1. Rats treated with NS and kept at 24°C > 480†, ‡ > 480†, ‡, # 2. Rats treated with NS (1 ml/kg, i.v.) and kept at 42°C 82 ± 3* 23 ± 2*,†, ‡ 3. Rats treated with NS (11 ml/kg, i.v.) and kept at 42°C 79 ± 4* 34 ± 6*, ‡ 4. Rats treated with HES (10%, 11 ml/kg, i.v.) and kept at 42°C 81 ± 3* 177 ± 15*,†, # 5. Rats treated with DXM (4 mg/kg, i.v.) and kept at 42°C 80 ± 3* 28 ± 7*, ‡ 6. Rats treated with DXM (4 mg/kg, i.v.)+HES (10%, 11 ml/kg, i.v.) and kept at 42°C 79 ± 4* 262 ± 17*, †, ‡, # NS or drugs were administered immediately after the onset of heat stroke.

%, 11 ml/kg, i.v.) and kept at 42°C 81 ± 3* 177 ± 15*,†, # 5. Rats treated with DXM (4 mg/kg, i.v.) and kept at 42°C 80 ± 3* 28 ± 7*, ‡ 6. Rats treated with DXM (4 mg/kg, i.v.)+HES (10%, 11 ml/kg, i.v.) and kept at 42°C 79 ± 4* 262 ± 17*, †, ‡, # NS or drugs were administered immediately after the onset of heat stroke. Values are the means ± SEM of 8 rats per group. Groups 2-6 exposed to 42°C had heat exposure withdrawn at the onset of heat stroke.*P < 0.05, compared with the corresponding control values (rats kept at 24°C; treatment 1). (one-way ANOVA, followed by Duncan's test). †P < 0.05, compared with the corresponding control values (rats treated with NS (11 ml/kg) and kept at 42°C; treatment 3). (one-way ANOVA, followed by Duncan's test). ‡P < 0.05, compared with the corresponding control values (rats treated with HES (11 ml/kg) and kept at 42°C; treatment 4). (one-way ANOVA, followed by Duncan's test). #P < 0.05, compared with the corresponding control values (rats treated with DXM (4 mg/kg) and kept at 42°C; treatment 5). (one-way ANOVA, followed by Duncan's test).

pared with the corresponding control values (rats treated with HES (11 ml/kg) and kept at 42°C; treatment 4). (one-way ANOVA, followed by Duncan's test). #P < 0.05, compared with the corresponding control values (rats treated with DXM (4 mg/kg) and kept at 42°C; treatment 5). (one-way ANOVA, followed by Duncan's test). Figure 1 Physiological parameters, and cellular ischemia and injury markers. Effects of heat exposure (ambient temperature; Ta = 42°C for 80 min) on colonic temperature (Tco), mean arterial pressure (MAP), heart rate (HR), cerebral blood flow (CBF) and the extracellular concentrations of glutamate, glycerol, and lactate/pyruvate ratio of the corpus striatum in normothermic control rats (open circles), 0.9% NaCl solution (11 ml/kg)-treated (filled circles), DXM (4 mg/kg)-treated (open squares), HES (10%, 11 ml/kg)-treated (filled squares), or the combined agent (DXM+HES)-treated rats (open triangles). The dotted line indicates time of heat stroke onset and drug injection. *P <0.05, compared with normothermic control rats. †P <0.05, compared with saline-treated rats (Ta = 42°C for 80 min).#P <0.05, compared with HES-treated rats (Ta = 42°C for 80 min) (ANOVA followed by Duncan's test).

-treated rats (open triangles). The dotted line indicates time of heat stroke onset and drug injection. *P <0.05, compared with normothermic control rats. †P <0.05, compared with saline-treated rats (Ta = 42°C for 80 min).#P <0.05, compared with HES-treated rats (Ta = 42°C for 80 min) (ANOVA followed by Duncan's test). Figure 2 Extracellular levels of dopamine, serotonin, and total production of DHBA. Effects of heat exposure (ambient temperature; Ta = 42°C for 80 min) on extracellular levels of dopamine, serotonin, and total production of dihydroxybenzoic acid (DHBA) of the cerebral corpus striatum in normothermic control rats (open circles), 0.9% NaCl solution (11 ml/kg)-treated (filled circles), DXM (4 mg/kg)-treated (open squares), HES (10%, 11 ml/kg)-treated (filled squares), or the combined agent (DXM+HES)-treated rats (open triangles). The dotted line indicates time of heat stroke onset and drug injection. *P <0.05, compared with normothermic control rats. †P <0.05, compared with saline-treated rats (Ta = 42°C for 80 min). #P <0.05, compared with HES-treated rats (Ta = 42°C for 80 min) (ANOVA followed by Duncan's test). Table 2 Effects of heat exposure (42°C for 80 min) plus 25 min room temperature (24°C) exposure on the neuronal damage score values of corpus striatum from normal saline-treated, dexamethasone(DXM), HES or the combined agent (DXM+HES)-treated ratsa. Treatment Score of neuronal damage (0-3)

Figure 2 Extracellular levels of dopamine, serotonin, and total production of DHBA. Effects of heat exposure (ambient temperature; Ta = 42°C for 80 min) on extracellular levels of dopamine, serotonin, and total production of dihydroxybenzoic acid (DHBA) of the cerebral corpus striatum in normothermic control rats (open circles), 0.9% NaCl solution (11 ml/kg)-treated (filled circles), DXM (4 mg/kg)-treated (open squares), HES (10%, 11 ml/kg)-treated (filled squares), or the combined agent (DXM+HES)-treated rats (open triangles). The dotted line indicates time of heat stroke onset and drug injection. *P <0.05, compared with normothermic control rats. †P <0.05, compared with saline-treated rats (Ta = 42°C for 80 min). #P <0.05, compared with HES-treated rats (Ta = 42°C for 80 min) (ANOVA followed by Duncan's test). Table 2 Effects of heat exposure (42°C for 80 min) plus 25 min room temperature (24°C) exposure on the neuronal damage score values of corpus striatum from normal saline-treated, dexamethasone(DXM), HES or the combined agent (DXM+HES)-treated ratsa. Treatment Score of neuronal damage (0-3) striatum 1. Normal saline (1 ml/kg, i.v.)-treated rats at 24°C 0 (0, 0.75) 2. Normal saline (1 ml/kg, i.v.)-treated rats at 42°C 2 (2, 2.25)b 3. DXM (4 mg/kg, i.v.)-treated rats at 42°C 2 (2, 2)b 4. HES (10%, 11 ml/kg, i.v.)-treated rats at 42°C 1 (0.25, 1)c 5. DXM (4 mg/kg, i.v.)+HES (10%, 11 ml/kg, i.v.)-treated rats at 42°C 1 (0, 1)c aValues are means ± S.E. of 8 rats per group followed by median with Q1and Q3 in parenthesis. For determination of neuronal damage score animals were killed after 80 mins heat exposure plus 25 mins room temperature exposure. The data were evaluated by a Wilcoxon signed rank test followed by Mann-Whitney test.

1 (0, 1)c aValues are means ± S.E. of 8 rats per group followed by median with Q1and Q3 in parenthesis. For determination of neuronal damage score animals were killed after 80 mins heat exposure plus 25 mins room temperature exposure. The data were evaluated by a Wilcoxon signed rank test followed by Mann-Whitney test. bP < 0.05, significance of difference from corresponding values (treatment 1, control rats). cP < 0.05 significance of difference from corresponding values (treatment 2). Figure 3 Serum levels of MDA. Effects of heat exposure (42°C) on serum levels of malondialdehyde (MDA) in normothermic control rats (white bar), 0.9% NaCl solution (11 ml/kg)-treated (black bar), DXM (4 mg/kg)-treated (diagonal bar), HES (10%, 11 ml/kg)-treated (cross bar) or the combined agent (DXM+HES)-treated rats (gray bar). *P <0.05, compared with normothermic control rats. †P <0.05, compared with saline-treated rats (Ta = 42 for 80 min). #P <0.05, compared with HES-treated rats (Ta = 42°C for 80 min) (ANOVA followed by Duncan's test). The blood samples were acquired after 100 min the initiation heat exposure in heat stroke rats or the equivalent time in normothermic controls.

ntrol rats. †P <0.05, compared with saline-treated rats (Ta = 42 for 80 min). #P <0.05, compared with HES-treated rats (Ta = 42°C for 80 min) (ANOVA followed by Duncan's test). The blood samples were acquired after 100 min the initiation heat exposure in heat stroke rats or the equivalent time in normothermic controls. Figure 4 Serum levels of IL-1β and TNF-α. Effects of heat exposure (42°C) on serum levels of interleukin-1β (IL-1β) and tumor necrosis factor-α (TNF-α) in normothermic control rats (white bar), 0.9% NaCl solution (11 ml/kg)-treated (black bar), DXM (4 mg/kg)-treated (diagonal bar), HES (10%, 11 ml/kg)-treated (cross bar) or the combined agent (DXM+HES)-treated rats (gray bar). *P <0.05, compared with normothermic control rats. +P <0.05, compared with saline-treated rats (Ta = 42°C for 80 min). #P <0.05, compared with HES-treated rats (Ta = 42°C for 80 min). (ANOVA followed by Duncan's test). The blood samples were acquired after 100 min the initiation heat exposure in heat stroke rats or the equivalent time in normothermic controls. Animal surgery and physiological parameter monitoring Under urethane anesthesia, the right femoral artery of the rats was cannulated with polyethylene tubing (PE50) for physiological monitoring, the right femoral vein was also cannulated for blood sampling and drug administration. The animals were then positioned in a stereotaxic apparatus (Kopf model 1460) for measurement of CBF and microdialysis experiment. Physiological monitoring included colon temperature (TCO), MAP, heart rate (HR) and CBF values in the cerebral corpus striatum.

s also cannulated for blood sampling and drug administration. The animals were then positioned in a stereotaxic apparatus (Kopf model 1460) for measurement of CBF and microdialysis experiment. Physiological monitoring included colon temperature (TCO), MAP, heart rate (HR) and CBF values in the cerebral corpus striatum. Experimental groups Rats were randomly assigned to one of six major groups. One group of rats (n = 8) with heat stroke received normal saline (NS) treatment (1 or 11 ml/kg body wt, 0.9% NaCl solution, i.v.) at the onset of heat stroke. Heat stroke was induced by exposing the animals to an ambient temperature of 42°C (with a relative humidity of 60% in a temperature-controlled chamber). The moment in which MAP and local CBF began to sharply decrease from their peak levels was arbitrarily defined as the onset of heat stroke, as shown in Figure 1. The interval between the start of heat exposure and onset of heat stroke were taken as values of latency. The interval between the initiation of heat stroke onset and animal death were taken as values of survival time. Another group of rats (n = 8) with heat stroke respectively received DXM 4 mg/kg i.v., HES (10%, 11 ml/kg i.v., Fresenius AG, Bad Homburg, Germany), or the combined agent [DXM (4 mg/ml/kg) together with HES (10%, 11 ml/kg)] i.v. at the onset of heat stroke. The other group of rats (n = 8) were normothermic control rats which were exposed to an ambient temperature of 24°C for at least 90 min to reach thermal equilibrium. Their physiological parameters were continuously recorded for up to 480 min (at the end of experiment). Their colon temperatures were maintained at about 36°C using the electric thermal mat before the start of experiments. The rats of these groups were continually monitored from physiological parameters (such as TCO, MAP, HR, and CBF) and ST during heat stroke. According to the values of ST, rats treated with the combined agent displayed the best beneficial effect on prolongation of ST in Table 1. Consequently, investigation of the heat stroke-induced circulatory shock, cerebrovascular dysfunction, cerebral ischemia and neuronal damage would be emphasized by the influence of the combined agent administration.

s treated with the combined agent displayed the best beneficial effect on prolongation of ST in Table 1. Consequently, investigation of the heat stroke-induced circulatory shock, cerebrovascular dysfunction, cerebral ischemia and neuronal damage would be emphasized by the influence of the combined agent administration. Measurements of Cellular Ischemia and Injury Markers After cannulation of vessels, the animal's head was mounted on a stereotaxic apparatus (Davis Kopf Instruments) with the nose bar positioned 3.3 mm below the horizontal line. Following a midline incision, the skull was exposed and a burr hole was made in the skull for the insertion of a dialysis probe (4 mm in length, CMA/12, Carnegie Medicine, Stockholm, Sweden). The microdialysis probe was stereotaxically implanted into the corpus striatum according to the atlas and coordinates of Paxinos and the coordinates of Paxinos and Watson (1982) [22]. As the methods described previously [19,23], an equilibrium period of 2 hours without sampling was allowed after probe implantation. The dialysis probe was perfused with Ringer's solution (147 mM Na+, 2.2 mM Ca2+, 4 mM K+, pH 7.0) at 2 μl/min using a CMA/100 microinfusion pump. Dialysates were collected every 20 min in a CMA140 fraction collector. Aliquots of dialysates (5 μl) were injected onto a CMA600 Microdialysis Analyzer (Carnegie Medicine) for measurement of lactate, glycerol, pyruvate and glutamate. Four analytes can be analyzed per sample and the result is displayed graphically within minutes. The thermal experiments were started after showing stabilization in four consecutive samples.

e injected onto a CMA600 Microdialysis Analyzer (Carnegie Medicine) for measurement of lactate, glycerol, pyruvate and glutamate. Four analytes can be analyzed per sample and the result is displayed graphically within minutes. The thermal experiments were started after showing stabilization in four consecutive samples. The lactate/pyruvate ratio is a well-known marker of cell ischemia, that is, an inadequate supply of oxygen and glucose. Glycerol is a marker of how severely cells are affected by the ongoing pathology. Glutamate is released from neurons during ischemia and initiates a pathological influx of calcium leading to cell damage. It is an indirect marker of cell damage in the brain, as described previously [19,23].

supply of oxygen and glucose. Glycerol is a marker of how severely cells are affected by the ongoing pathology. Glutamate is released from neurons during ischemia and initiates a pathological influx of calcium leading to cell damage. It is an indirect marker of cell damage in the brain, as described previously [19,23]. Measurements of Extracellular DA and 5-HT release Dialysates samples were collected at 20 min intervals and assayed by an HPLC system. Extracellular monoamine concentrations were assayed by HPLC combined with an electrochemical detection system. The HPLC system comprised a Beckman 126 pump (BeckmanInstru-ments) and a CMA-200 microautosampler (CMA/Microdialysis, Stockholm, Sweden), and a microbore reversed-phase column was filled with Inertsil ODS-2 (GSK-C18, 5-mmOD, 150 × 1.0-mmID). The performance of each microdialysis probe was calibrated by dialysis of a known amount of the standard mixture, and recovery of all analyses was then determined. Brain concentrations of DA and 5-HT were calculated by determining each peak height ratio relative to the internal standard and were also corrected by each probe performance. The internal standard 3-methoxy-tyramine and standard mixtures were prepared fresh daily. The mobile phase was prepared by adding 60 ml of acetonitrile, 0.42 g of SDS (2.2 mM), 200 g of sodium citrate (30 mM), 10 mg of EDTA (0.027 mM), and 1 ml of diethylamine in double-distilled water.

ected by each probe performance. The internal standard 3-methoxy-tyramine and standard mixtures were prepared fresh daily. The mobile phase was prepared by adding 60 ml of acetonitrile, 0.42 g of SDS (2.2 mM), 200 g of sodium citrate (30 mM), 10 mg of EDTA (0.027 mM), and 1 ml of diethylamine in double-distilled water. Hydroxyl Radical Production Monitoring The concentrations of hydroxyl radicals were measured by a modified procedure based on the hydroxylation of sodium salicylate by hydroxyl radicals, leading to production of 2,3-dihydroxybenzoic acid (2,3-DHBA) and 2,5-DHBA. These two compounds were then measured in dialysates by HPLC with electrochemical detection. A Ringer's solution (0.860 g of NaCl, 0.030 g of KCl and 0.033 g of CaCl2 per 100 ml) containing 2 mmol/l sodium salicylate was perfused through the microdialysis probe at a constant flow rate (1.2 μl/min) A reverse-phase C18 column [phase II; particle size, 3 μm; 100 × 3.2 mm (length×internal) diameter]; BioAnalytic Systems, West Lafayette, IN, U.S.A.] was used, and the mobile phase consisted of a mixture of 75 mmol/l monochloroacetic acid, 0.7 mmol/l disodium EDTA, 1.5 mmol/l sodium 1-octanesulphonate and 45 ml/l acetonitrile (pH 3.0). The retention times of 2,3-DHBA and 2,5-DHBA were 9.07 and 5.44 min respectively.

; BioAnalytic Systems, West Lafayette, IN, U.S.A.] was used, and the mobile phase consisted of a mixture of 75 mmol/l monochloroacetic acid, 0.7 mmol/l disodium EDTA, 1.5 mmol/l sodium 1-octanesulphonate and 45 ml/l acetonitrile (pH 3.0). The retention times of 2,3-DHBA and 2,5-DHBA were 9.07 and 5.44 min respectively. Measurement of Serum MDA Levels 0.25 mL of serum was added to 25 μL of 0.2% BHT and 12.5 μL of 10 N NaOH (to adjust to pH~13) and incubated at 60°C for 30 min in a shaking water bath. To this was added 1.5 mL of 0.44 mol/L (or 7.2%) TCA containing 1% KI, and the mixture was placed in ice for 10 min and centrifuged (1,000g, 10 min). To 1 mL of the supernatant was added 0.5 mL of 0.6% TBA, and the mixture was heated at 95°C for 30 min. After cooling the mixture was extracted with 1.5 mL of n-butanol, and 20 μL of the butanol layer was injected to a C-18 (4.6 × 150 mm) column fitted with a guard and eluted at 1 mL/min by using 65% (v/v) 50 mM KH2PO4-KOH and 35% (v/v) methanol with spectrophotometric (532 nm) detector.

re was heated at 95°C for 30 min. After cooling the mixture was extracted with 1.5 mL of n-butanol, and 20 μL of the butanol layer was injected to a C-18 (4.6 × 150 mm) column fitted with a guard and eluted at 1 mL/min by using 65% (v/v) 50 mM KH2PO4-KOH and 35% (v/v) methanol with spectrophotometric (532 nm) detector. Measurement of Serum IL-1β and TNF-α Levels The blood samples were acquired 100 min after the initiation of heat exposure (or 20 min after the onset of heat stroke) in heat stroke rats or the equivalent time in normothermic controls. 5 ml of blood was withdrawn from the femoral vein of each rat for measurement of serum IL-1β or TNF-α. Blood samples were allowed to clot for 2 hours at room temperature or overnight at 2-8°C before centrifuging for 20 minutes at approximately 2000 ×g. Serum was quickly removed from these plasma samples and assayed for IL-1β or TNF-α immediately. The DuoSet Enzyme-linked Immunosorbent Assay (ELISA) Development System rat IL-1β or TNF-α kit (R&D Systems, Minneapolis, MN, USA) was used for measuring the levels of active rat IL-1β or TNF-α present in serum. This assay employs the quantitative colorimetric sandwich ELISA technique.

d for IL-1β or TNF-α immediately. The DuoSet Enzyme-linked Immunosorbent Assay (ELISA) Development System rat IL-1β or TNF-α kit (R&D Systems, Minneapolis, MN, USA) was used for measuring the levels of active rat IL-1β or TNF-α present in serum. This assay employs the quantitative colorimetric sandwich ELISA technique. Neuronal Damage Score At the end of each experiment, the brain was removed, fixed in 10% neutral buffered formalin and embedded in paraffin blocks. Serial (10 μm) sections through the striatum were stained with hemotoxylin and eosin for microscopic evaluation. The extent of striatal neuronal damage was scored on a scale of 0-3, modified from the grading system of Pulsinelli et al. (1982) [24], in which 0 is normal, 1 means that ~30% of the neurons are damaged, 2 means that ~60% of that neurons are damaged, and 3 means that 100% of that neurons are damaged. Each hemisphere was evaluated independently without the examiner knowing the experimental conditions. When examined for neuronal damage in gray matter, only areas other than those invaded by probes were assessed.

damaged, 2 means that ~60% of that neurons are damaged, and 3 means that 100% of that neurons are damaged. Each hemisphere was evaluated independently without the examiner knowing the experimental conditions. When examined for neuronal damage in gray matter, only areas other than those invaded by probes were assessed. Statistical analysis Data are presented as the mean ± SEM. Repeated-measures analysis of variance was used for factorial experiments, whereas Duncan's multiple-range test was used for post hoc multiple comparisons among means. For scoring neuronal damage, the Wilcoxon signed rank test was used when only two groups were compared. The Wilcoxon tests which convert the scores or values of a variable to ranks, require calculation of a sum of the ranks and provide critical values for the sum necessary to test the null hypothesis at a given significant level. The data were given by "median", and first and third quartile. A P value less than 0.05 was considered as statistical significance.

ores or values of a variable to ranks, require calculation of a sum of the ranks and provide critical values for the sum necessary to test the null hypothesis at a given significant level. The data were given by "median", and first and third quartile. A P value less than 0.05 was considered as statistical significance. Results The combined agent (DXM+HES) improves survival time in heat stroke Table 1 summarizes the effects of heat exposure (42°C Ta for 80 min) on survival time in rat heat stroke. In anesthetized rats treated with intravenous (i.v.) doses of normal saline (NS) or drugs immediately after the onset of heat stroke, although showing no change in latency, displayed increases in survival time in some groups. For example, rats treated with an i.v. dose of 1 ml/kg or 11 ml/kg of NS had survival time values of 23 ± 2 min or 34 ± 6 min, respectively. Rats treated with HES solution at an i.v. dose of 11 ml/kg immediately after the onset of heat stroke had a survival time value of 177 ± 15 min. This increase in survival would indicate that the HES, but not the NS groups, are adequately resuscitated. Additionally, although immediate treatment with DXM (4 mg/kg) alone had no apparent beneficial effect, the combined agent (combined administration of DXM (4 mg/kg) plus HES (11 mg/kg)) immediately after the onset of heat stroke did prolong the survival time as compared with the controls (as shown in Table 1).

scitated. Additionally, although immediate treatment with DXM (4 mg/kg) alone had no apparent beneficial effect, the combined agent (combined administration of DXM (4 mg/kg) plus HES (11 mg/kg)) immediately after the onset of heat stroke did prolong the survival time as compared with the controls (as shown in Table 1). The combined agent (DXM+HES) attenuates hypotension, cerebrovascular dysfunction and neuronal damage during heat stroke The effects of heat exposure (42°C for 80 min) on several physiological parameters in NS-, DXM-, HES- and the combined agent-treated rats are shown in Figure 1. In NS-treated (11 ml/kg) and DXM (4 mg/kg)-treated groups, the values of MAP and CBF were significantly decreased at 10 or 20 min after the onset of heat stroke (or 90 or 100 min after the start of heat stress) compared with those of normothermic controls. On the other hand, the values of extracellular concentrations of glutamate, glycerol and lactate/pyruvate ratio in the corpus striatum were significantly greater than those of the normothermic controls. In HES (11 ml/kg)-treated group, rats displayed greater MAP and CBF, but lower striatal levels of glutamate, glycerol, and lactate/pyruvate ratio after onset of heat stroke than those of the NS-or DXM- treated rats. However, heat stroke-induced arterial hypotension, cerebral ischemia and increased levels of glutamate, glycerol, and lactate/pyruvate ratio in the extracellular levels of striatum were all significantly diminished by treatment with the combined agent immediately at the onset of heat stroke or 80 min after start of heat stress.

heat stroke-induced arterial hypotension, cerebral ischemia and increased levels of glutamate, glycerol, and lactate/pyruvate ratio in the extracellular levels of striatum were all significantly diminished by treatment with the combined agent immediately at the onset of heat stroke or 80 min after start of heat stress. In separate experiments, 25 min after the onset of heat stroke, rats were sacrificed for determination of neuronal damage score in the corpus striatum. The data are summarized in Table 2. After the onset of heat stroke, rats treated with NS (11 ml/kg) displayed higher values of striatal neuronal damage [2 (2, 2.25)] compared with those of normothermic controls [0 (0, 0.75)] (as shown in Table 2 and Figure 5B). Values of striatal neuronal damage score in the DXM-treated rats and in the HES-treated rats were respectively [2 (2, 2)] and [1 (0.25, 1)]. However, with the combined agent [DXM (4 mg/kg)+HES (11 ml/kg)] acute treatment, neuroprotection ensured [1 (0, 1)]. Figure 5 shows that heat stroke-induced cell shrinkage, pyknosis of nucleus, and disappearance of nucleolus in the corpus striatum were attenuated by the combined agent (Figure 5C).

)] and [1 (0.25, 1)]. However, with the combined agent [DXM (4 mg/kg)+HES (11 ml/kg)] acute treatment, neuroprotection ensured [1 (0, 1)]. Figure 5 shows that heat stroke-induced cell shrinkage, pyknosis of nucleus, and disappearance of nucleolus in the corpus striatum were attenuated by the combined agent (Figure 5C). Figure 5 Histological examination of neuronal damage. The photomicrographs of the cerebral corpus striatum in a normothermic control rat treated with 0.9% NaCl solution (11 ml/kg) (A), a heat stroke rat treated with 0.9% NaCl solution (11 ml/kg) (B), or a heat stroke rat treated with the combined agent (DXM+HES) (C) immediately after the initiation of heat stroke. The striatal photomicrograph in a heat stroke rat treated with DXM (4 mg/kg) is similar to (B), and in a heat stroke rat treated with HES (10%, 11 ml/kg) is similar to (C) (data not shown). Twenty-five minutes after 80-min heat exposure, the corpus striatum of the rat treated with 0.9% NaCl solution showed cell shrinkage, pyknosis of the nucleus, and disappearance of nucleolus. After acute treatment with the combined agent, neuronal damage was reduced, as shown in C. The rats were sacrificed at 25 min after the termination of heat exposure or the equivalent time for the normothermic controls. Scale bar, 50 μm.

ion showed cell shrinkage, pyknosis of the nucleus, and disappearance of nucleolus. After acute treatment with the combined agent, neuronal damage was reduced, as shown in C. The rats were sacrificed at 25 min after the termination of heat exposure or the equivalent time for the normothermic controls. Scale bar, 50 μm. The combined agent (DXM+HES) reduces cerebral striatal levels of dopamine, serotonin and DHBA during heat stroke As shown in Figure 2, twenty minutes after the termination of heat stress in the NS-treated group or DXM-treated group, all the dopamine, serotonin, and DHBA (dihydroxybenzoic acid, indirectly stood for production of hydroxyl radicals) values in cerebral striatum were significantly greater than those of the normothermic controls (P < 0.05). Immediate treatment with the combined agent at the onset of heat stroke (80 min after the start of heat stress) significantly attenuated the heat stress-induced increases levels of dopamine, serotonin, and production of hydroxyl radicals in the corpus striatum.

reater than those of the normothermic controls (P < 0.05). Immediate treatment with the combined agent at the onset of heat stroke (80 min after the start of heat stress) significantly attenuated the heat stress-induced increases levels of dopamine, serotonin, and production of hydroxyl radicals in the corpus striatum. The combined agent (DXM+HES) reduces the levels of MDA, IL-1β and TNF-α in peripheral blood stream during heat stroke The serum IL-1β and TNF-α, and MDA levels for normothermic controls, NS-, DXM-, HES-treated heat stroke rats, and the combined agent-treated heat stroke rats are summarized in Figure 3 and 4. It can be seen from the figures that the serum IL-1β and TNF-α, and MDA levels in NS-treated heat stroke rats were all significantly higher at 20 min after the onset of heat stroke than those in the normothermic controls. The immediate treatment with HES alone and the combined agent at the onset of heat stroke attenuated the heat stroke-induced increased serum lipid peroxidation, as well as attenuating it increased the serum levels of IL-1β and TNF-α. However, these serum levels were more significantly diminished by treatment with the combined agent immediately at the onset of heat stroke (as shown in Figure 3 and 4).

stroke attenuated the heat stroke-induced increased serum lipid peroxidation, as well as attenuating it increased the serum levels of IL-1β and TNF-α. However, these serum levels were more significantly diminished by treatment with the combined agent immediately at the onset of heat stroke (as shown in Figure 3 and 4). Discussion It has been reported that pretreatment with DXM (4 mg/kg) single dose, but not immediate treatment with DXM, before heat stress could increase the ST in rats by attenuating serum levels of interleukins [18]; however, there are fewer studies showing the immediate treatment with the combined agent at the onset of heat stroke. It will be more meaningful the combined treatment shows neuroprotection after heat stroke attacks. Although the previous results [19] have shown an insignificant therapeutic effect of DXM (4 mg/kg) administered immediately after the onset of heat stroke alone, the combination of DXM and HES provides a better therapeutic effect for rats with heat stroke in the present study. Additionally, the volume-expanding effect of the HES would be thought to improve survival during heat stroke in resuscitation. The HES would be 1.5-2 times the volume administered by 1-2 hour after it is given. In fact, in our previous study, intravenous infusion of 2-11 ml/kg of HES solution improved survival in a dose-dependent manner during heat stroke, and suggesting treatment with 11mg/kg of HES had better prolongation in survival. Hence, the HES that acts to expand the circulatory volume and a potent inflammatory agent (such as DXM) might be combined to develop an improved fluid therapy for attenuation or prevention of heat stroke-induced damage. Although our previous results indicated that the combination of HES and DXM did provide a better survival effect for rats with heat stroke, the hemodynamic, histological and biological changes by the combined treatment immediately after the onset of heat stroke were not observed in detail. In this study, administration of the combined agent indeed appears more effective to prolong the ST in rats with heat stroke, by comparison to treatment of DXM or HES alone (shown in table 1). Similarly, in agreement with the present results, treatment of the combined agent (DXM and HES) can also offer beneficial amelioration from ischemic condition and therapeutic influence in ischemic experiments [25,26].

ng the ST in rats with heat stroke, by comparison to treatment of DXM or HES alone (shown in table 1). Similarly, in agreement with the present results, treatment of the combined agent (DXM and HES) can also offer beneficial amelioration from ischemic condition and therapeutic influence in ischemic experiments [25,26]. There is evidence that cerebral ischemia (due to arterial hypotension and intracranial hypertension) may be one of the major causes to induce further damage after heat stroke onset [13,18,19]. After heat stroke induction, the CBF instantly drops from highest peak, and it is concomitant with significant increments of cerebral ischemia and injury indexes, as shown in figure 1. The lactate/pyruvate ratio is a well known marker of cellular ischemia, whereas glycerol is a marker of how severely cells are affected by ongoing pathology [27]. Excessive accumulation of glutamate has been shown in ischemic brain tissue [27]. Indeed, both present and previous results [19,23,27] have demonstrated that extracellular levels of glutamate, glycerol and lactate/pyruvate in ischemic brain are greater in heat stroke rats compared with those of normothermic controls. Meanwhile, evidences of histopathological morphology and neuronal damage scores also reveal severe neuronal damage (shown in figure 5, table 2) in heat stroke rats. However, as shown in the present results, all these heat stroke-induced cerebral ischemia and injury can be alleviated by acute treatment with the combined agent.

le, evidences of histopathological morphology and neuronal damage scores also reveal severe neuronal damage (shown in figure 5, table 2) in heat stroke rats. However, as shown in the present results, all these heat stroke-induced cerebral ischemia and injury can be alleviated by acute treatment with the combined agent. There were many evidences [8,14,28] that the increased DA, 5-HT and glutamate in the brain during the rat heat stroke were mediated in the development of neuronal damage. Cerebral DA, 5-HT or/and glutamate overload resulting from arterial hypotension and intracranial hypertension might be responsible for occurrence of central nervous system syndromes associated with heat stroke [14,28]. Systemic administration of dopaminergic or serotoninergic nerve depletors or receptor antagonists, or glutamate receptor antagonists cloud protect against ischemic neuronal injury in experimental heat stroke [14,28,29]. In addition, recent studies showed that the excessive accumulation of cytotoxic free radicals in the brain and oxidative stress occurred during heat stroke [9,10,30]. Evidence had accumulated to suggest that heat stroke-induced cerebral ischemia and neuronal damage might be associated with an increased production of free radicals, specifically hydroxyl radicals [9,10]. Pretreatment with hydroxyl radical scavengers, such as α-tocopherol, prevented production of hydroxyl radicals, reduced lipid peroxidation and ischemic neuronal damage in the brain of rats exposed to heat stroke and prolonged subsequent survival [31]. In brief, as demonstrated by Chang et al, after the onset of heat stroke, cessation or reduction of blood flow to the brain induced neuronal damage. This neurotoxic cascade involved overproduction of glutamate, DA, and 5-HT as well as oxidative stress in the brain [6]. Likewise, in the present study, heat stroke also produces similar increases in cerebral striatal DA, 5-HT, glutamate and hydroxyl radical production in heat stroke rats. Additionally, the heat stroke rats also displayed increased levels of lipid peroxidation in the peripheral blood stream. Indeed, according to our present findings, the heat stroke-induced high levels of DA, 5-HT, glutamate, and hydroxyl radicals in rats' corpus striatum, and the elevated plasma MDA levels can be prevented by acute treatment with the combined agent.

ed increased levels of lipid peroxidation in the peripheral blood stream. Indeed, according to our present findings, the heat stroke-induced high levels of DA, 5-HT, glutamate, and hydroxyl radicals in rats' corpus striatum, and the elevated plasma MDA levels can be prevented by acute treatment with the combined agent. This probably implies that the immediate administration of the combined agent during heat stroke may be mediated with the decrements of cerebral monoamines and oxidative stress to prolong the ST and improve the cerebral neuronal damage in rats.

ed increased levels of lipid peroxidation in the peripheral blood stream. Indeed, according to our present findings, the heat stroke-induced high levels of DA, 5-HT, glutamate, and hydroxyl radicals in rats' corpus striatum, and the elevated plasma MDA levels can be prevented by acute treatment with the combined agent. This probably implies that the immediate administration of the combined agent during heat stroke may be mediated with the decrements of cerebral monoamines and oxidative stress to prolong the ST and improve the cerebral neuronal damage in rats. The serum concentrations of inflammatory cytokines (such as IL-1β and TNF-α) are elevated in humans and animals with heat stroke [12,18,23,32]. The levels of both TNF-α and IL-1 receptors correlate well with severity of heat stroke [32,33]. The previous studies had also shown that heat stroke induced systemic and cerebral striatal productions of IL-1β and TNF-α in both rats and rabbits [9,31,34,35]. Indeed, as it is shown in the present results, an increase of serum IL-1β and TNF-α levels is observed in heat stroke rats. The increase in the levels of these inflammatory cytokines is associated with arterial hypotension, cerebral ischemia and neuronal damage. Administration of IL-1 receptor antagonists could prevent arterial hypotension and cerebral ischemic damage, and improve survival in heat stroke. Furthermore, the present results show that treatment with the combined agent significantly attenuates the heat stroke-induced overproduction of IL-1β and TNF-α in the serum. Meanwhile, both arterial hypotension and cerebral ischemic damage are attenuated and survival of heat stroke rats is ameliorated following acute treatment with the combined agent. The immediate administration of this combined agent might exert its protective effects by attenuating the increased plasma level of IL-1β and TNF-α during heat stroke.

potension and cerebral ischemic damage are attenuated and survival of heat stroke rats is ameliorated following acute treatment with the combined agent. The immediate administration of this combined agent might exert its protective effects by attenuating the increased plasma level of IL-1β and TNF-α during heat stroke. Our results indicated that following heat stroke, arterial hypotension, decreased cerebral blood flow, increased serum levels of IL-1β, TNF-α and MDA, and increased striatal dopamine, serotonin and hydroxyl radicals and increased of levels of glutamate, glycerol and lactate/pyruvate ratio developed. Although HES administration alone showed a pronounced effect, it was found that treatment with the combined agent conferred a moderate further beneficial effect to ameliorate these changes, and improve neuronal damage and survival time. Various clinical and experimental investigations of stroke and brain injury have shown that HES administration might reduce brain edema and intracranial hypertension [36-38]. It was also obtained that the values of MAP, cerebral perfusion pressure (CPP), and cerebral levels of local CBF were significantly lower during heat stroke [6,27]. The maintenance of appropriate levels of CBF might be brought about by higher CPP resulting from lower intracranial pressure (ICP; possibly due to reduction in brain edema and cerebovascualr congestion) and higher MAP during development of heat stroke [39]. This raises the possibility that HES might be a beneficial treatment for heat stroke subjects with intracranial hypertension as well as decreased cerebral perfusion. As a result of present study, we see from Figure 1, 2, 3 and 4, and Table 1 and 2 that acute treatment with HES (11 ml/kg) alone at the onset of heat stroke can alleviate the heat stroke-induced arterial hypotension, cerebral monoamines and hydroxyl radical production overload, systemic inflammation, and severe cerebral ischemia and damage. However, treatment with the combined agent (both HES and DXM) has more effective therapy than treatment with HES alone to maintain appropriate levels of MAP and CBF by attenuating the heat stroke-induced abnormal physiological and pathological changes, and results in prolongation in survival (as shown in Figure 1, 2, 3 and 4, and Table 1). Therefore, HES treatment showed partial effects on those parameters after heat stroke induction.

to maintain appropriate levels of MAP and CBF by attenuating the heat stroke-induced abnormal physiological and pathological changes, and results in prolongation in survival (as shown in Figure 1, 2, 3 and 4, and Table 1). Therefore, HES treatment showed partial effects on those parameters after heat stroke induction. According to our present results, it is reasonable to assume that acute treatment with both HES and DXM has a better effectiveness on reducing the heat stroke-induced damage, and augmenting ST. It is not known whether HES treatment exerts its benefit effect in heat stroke by acting through attenuation of brain edema and intracranial hypertension in present study. Of course, this needs further investigation. Conclusions In the present study, the heat stroke-induced increases in arterial hypotension, cerebral ischemia and neuronal damage are associated with elevated levels of DA, 5-HT, glutamate and hydroxyl radicals in rat brain, and increased circulating IL-1β, TNF-α and MDA in the peripheral blood stream. The immediate systemic treatment with the combined agent (both DXM and HES), in addition to attenuating the elevating levels of IL-1β, TNF-α and MDA in blood stream, diminishes monoamines, glutamate, and hydroxyl radical formation, and ischemia injury in the brain, and improves ST in rats with heat stroke. Our results suggest that the combination of a colloid substance with a volume-expanding effect and an anti-inflammatory agent may provide a better resuscitation solution for victims with heat stroke.

, glutamate, and hydroxyl radical formation, and ischemia injury in the brain, and improves ST in rats with heat stroke. Our results suggest that the combination of a colloid substance with a volume-expanding effect and an anti-inflammatory agent may provide a better resuscitation solution for victims with heat stroke. Abbreviations CBF: cerebral blood flow; DA: dopamine; DHBA: dihydroxybenzoic acid; DXM: dexamethasone; ELISA: enzyme-linked immunosorbent assay; HES: hydroxyl starch; HR: heart rate; 5-HT: serotonin; IL-1β : interleukin-1β; MAP: mean arterial pressure; NS: normal saline; ST: survival time; Tco: colon temperature; TNF-α: tumor necrosis factor-α; ICP: intracranial pressure; CPP: cerebral perfusion pressure. Competing interests The authors declare that they have no competing interests. Authors' contributions All authors have read and approved the final manuscript. YTH and WMY operated the animals, assessed the neuron damage score and interpreted the data. HWY and LKL collected blood samples and performed the ELISA. SMF and WYS provided DXM and HES, and finalized the manuscript. YTH and LCC conceived the experiments, funded the project and wrote the manuscript. Acknowledgements We are grateful to Dr M. T. Lin and C. P. Chang for their kind help in performing the assay of dopamine, serotonin, and DHBA levels in the corpus striatum of rats. This work was supported by grants from the National Science Council of the Republic of China (grant no. NSC 94-2320-B-041-001).

ecreased in aged and significantly reduced in OVX females (P = 0.0061) attenuating bleeding. Moreover, the infarct volumes were reduced (Figure 4B; P < 0.0001) and neurologic scores were also improved (Figure 5B; P < 0.0001) in all minocycline treated females as compared to their corresponding vehicle treated controls. The inhibition of MMP-9 by minocycline Because MMP-9 down-regulation has been previously shown as a potential mechanism of minocycline protection in young males [14,15], we investigated if minocycline down-regulates MMP-9 in females. All MMP-9 studies were performed by group blinded investigators. Consistent with previous reports [15] we did not observe the increase in MMP-9 level in vehicle treated animals of either group at 24 hours after stroke compared to sham controls (data not shown). In time dependent studies we detected the maximum MMP-9 level in the brain at 6 hours post stroke (Additional file 6, Figure S6). However, in plasma samples we detected low, or no, MMP-9 after ischemia compare to sham controls. This may be associated with using citrate to prevent blood clotting for plasma harvesting in our study.

Introduction Common symptoms of stroke include motor disturbances, such as postural imbalance and disturbed skilled movement [1-3]. The degree of spontaneous functional recovery after stroke is determined by inflammatory processes, which are modulated by stress and activity of the hypothalamic-pituitary-adrenal (HPA) axis [4]. Stress and high levels of glucocorticoids (GCs) are associated with poor stroke outcome and high morbidity [5-8]. The experience of psychological distress is associated with increased stroke risk and represents a predictor of fatal stroke [9-13]. Stress has been recognized as a critical variable in determining the success of stroke therapy and rehabilitative interventions [14,15]. Because adverse experience modulates neuronal plasticity [16,17] the intrinsic properties of the HPA axis, such as GR density, may explain the large variability in recovery rates among individual patients [18]. Because GC therapy also represents a potential treatment for stroke, investigation of glucocorticoid receptor (GR) regulation is critical to explore potential therapeutic avenues. This study describes the effects of psychological stress on GR expression and activation, and its impact on recovery and outcome after ischemic lesion in a rat model. The findings suggest that stress-induced regulation of GR expression after ischemic stroke may influence HPA axis feedback mechanisms.

Stress and high levels of glucocorticoids (GCs) are associated with poor stroke outcome and high morbidity [5-8]. The experience of psychological distress is associated with increased stroke risk and represents a predictor of fatal stroke [9-13]. Stress has been recognized as a critical variable in determining the success of stroke therapy and rehabilitative interventions [14,15]. Because adverse experience modulates neuronal plasticity [16,17] the intrinsic properties of the HPA axis, such as GR density, may explain the large variability in recovery rates among individual patients [18]. Because GC therapy also represents a potential treatment for stroke, investigation of glucocorticoid receptor (GR) regulation is critical to explore potential therapeutic avenues. This study describes the effects of psychological stress on GR expression and activation, and its impact on recovery and outcome after ischemic lesion in a rat model. The findings suggest that stress-induced regulation of GR expression after ischemic stroke may influence HPA axis feedback mechanisms. Materials and methods Animals Twenty-two adult male Long-Evans rats (400-500 g) raised at the University of Lethbridge vivarium were used. For participation in skilled reaching, animals were food restricted to maintain 90-95% of baseline body weight. Animals were matched to two groups based on baseline reaching success: Stress (n = 11), and handled Controls (n = 11). All procedures were performed in accordance with the guidelines of the Canadian Council for Animal Care and approved by the local Animal Welfare Committee.

d to maintain 90-95% of baseline body weight. Animals were matched to two groups based on baseline reaching success: Stress (n = 11), and handled Controls (n = 11). All procedures were performed in accordance with the guidelines of the Canadian Council for Animal Care and approved by the local Animal Welfare Committee. Motor cortex lesion Rats were anesthetized using isoflurane in an oxygen/nitrous oxide mixture (isoflurane 4% for initiation, 2% for maintenance at an oxygen flow rate of 2.0 l/min). Motor cortex devascularization was induced contralateral to the paw preferred in skilled reaching [19]. Briefly, the skin over the skull was incised and the skull was exposed. Using a fine dental burr, a craniotomy was performed at the following coordinates: -1.0 to 4.0 mm anterior-posterior, 1.5 to 4.5 mm lateral to Bregma. The dura and blood vessels were carefully wiped off using a sterile cotton tip. The skin was sutured and the rats were given analgesic (Temgesic, Schering-Plough, Brussels, Belgium). Rats were allowed to recover in individual cages on a heating pad until fully awake and were then returned to their home cages.

o Bregma. The dura and blood vessels were carefully wiped off using a sterile cotton tip. The skin was sutured and the rats were given analgesic (Temgesic, Schering-Plough, Brussels, Belgium). Rats were allowed to recover in individual cages on a heating pad until fully awake and were then returned to their home cages. Stress paradigm The stress regimen was performed 7 days prior, and 21 days after motor cortex lesion. Rats were placed individually in a transparent Plexiglas cylinder [20-22]. The cylinder (5 cm inner diameter) had perforated ends to allow for ventilation and maintained the animals in a standing position without compression of the body. Restraint stress was given starting at 9:00 AM. Animals were restrained for 20 minutes and were tested 10 minutes later in the skilled reaching task [23,21,25]. Furthermore, on the last day of baseline (pre-stress) and after 3 weeks of post-lesion behavioural tests, blood samples were collected after a 10 minute post-stress interval. Thus, both behavioural testing and blood sampling took place at a time when elevated corticosterone levels after restraint stress can be expected [21,23].

re, on the last day of baseline (pre-stress) and after 3 weeks of post-lesion behavioural tests, blood samples were collected after a 10 minute post-stress interval. Thus, both behavioural testing and blood sampling took place at a time when elevated corticosterone levels after restraint stress can be expected [21,23]. Skilled reaching task The rats were trained in the single pellet reaching task to assess skilled forelimb function [26,27] (Figure 1A). The reaching boxes were made of clear Plexiglas (40 cm × 45 cm × 13 cm). The front wall of the box had a 1.3 cm wide vertical slit, allowing the rats to reach for a food pellet located on a shelf attached to the outside of the box. The shelf was located 4 cm above the floor. On top of the shelf were two indentations (5 mm in diameter, and 1.5 mm deep), each aligned with one side of the slit. These indentations stabilized the pellet and were located 1.5 cm away from the front wall [27]. In each training session, rats were placed individually in the reaching box and a food pellet (45 mg each, BioServ, Frenchtown, NJ) was placed contralaterally to the rats' preferred reaching paw. To readjust their body position, rats were trained to walk to the rear end of the box before reaching for a new pellet. Each rat was given 20 pellets per training and test session.

in the reaching box and a food pellet (45 mg each, BioServ, Frenchtown, NJ) was placed contralaterally to the rats' preferred reaching paw. To readjust their body position, rats were trained to walk to the rear end of the box before reaching for a new pellet. Each rat was given 20 pellets per training and test session. Figure 1 Stress induces motor impairments. (A), photograph of a rat reaching for a food pellet. (B), success percent; (C), pellets obtained percent; (D), number of attempts to grasp 20 pellets. All data are presented as group means + SEM. * p < 0.05, ** p < 0.01, *** p < 0.001, compared to the Control group. A successful reach was defined as obtaining the pellet on the first attempt, withdrawing the paw through the slit and releasing the pellet to the mouth. Success was calculated using the following formula: Percentage of reaching success =number of successful reaches20×100 The percentage of total number of pellets obtained was measured by counting the number of pellets eaten, regardless of whether the pellet was grasped and eaten on the first attempt. If the rat dropped the pellet, it did not count as a pellet eaten. The percentage of total pellets obtained was measured using the following formula: Percentage of total pellets obtained=number of pellets obtained20×100 To assess reaching accuracy, the number of attempts to grasp a single pellet was averaged.

The percentage of total number of pellets obtained was measured by counting the number of pellets eaten, regardless of whether the pellet was grasped and eaten on the first attempt. If the rat dropped the pellet, it did not count as a pellet eaten. The percentage of total pellets obtained was measured using the following formula: Percentage of total pellets obtained=number of pellets obtained20×100 To assess reaching accuracy, the number of attempts to grasp a single pellet was averaged. Once reaching success rates during training sessions reached an asymptotic level, performance was recorded for 5 days of baseline testing. These values were averaged for further analysis (pre-stress). After the start of the stress regimen animals were tested in skilled reaching for 7 days prior and for 21 days after lesion. Blood samples Blood samples were taken from the tail vein on the last day of baseline (pre-stress), and three weeks post-lesion (the day before sacrifice). An average of 0.6 ml of blood was collected 10 minute post-stress interval under 4% isoflurane anesthesia. Blood samples were collected between the hours of 9:10-10:30 AM. No behavioural testing was performed on days on which blood samples were taken. Plasma corticosterone concentrations were determined by radioimmunoassay using commercial kits (Coat-A-Count, Diagnostic Products Corp., Los Angeles, CA) [28].

ane anesthesia. Blood samples were collected between the hours of 9:10-10:30 AM. No behavioural testing was performed on days on which blood samples were taken. Plasma corticosterone concentrations were determined by radioimmunoassay using commercial kits (Coat-A-Count, Diagnostic Products Corp., Los Angeles, CA) [28]. Histology and immunohistochemistry After completion of behavioural tests, rats were sacrificed with an overdose of pentobarbital (Euthansol; CDMV Inc. Québec, Canada) and transcardially perfused with phosphate buffered saline followed by 4% formaldehyde. Brains were cut at 10 μm thickness. Forty-five sections per brain were cut in each of the anterior lesion site, the middle section and the posterior lesion site. Three sections of each area were used for immuno-histochemical analysis. Infarct volume Cresyl Violet stained brain sections (10 μm) were used to analyze infarct volume. Cross-sectional volumes of both hemispheres were calculated using a Zeiss Axiovision 4.3 microscope (Zeiss, Jena, Germany). Infarct volume was measured according to the Cavalieri method [29]. The following formulas were used (VH: volume of a hemisphere; VTL: volume of tissue lost): VH= average (average area of a complete coronal hemisphere – ventricles – area of damage) × interval between sections × number of sections VTL= tissue remaining in normal hemisphere – tissue remaining in injured hemisphere

Infarct volume Cresyl Violet stained brain sections (10 μm) were used to analyze infarct volume. Cross-sectional volumes of both hemispheres were calculated using a Zeiss Axiovision 4.3 microscope (Zeiss, Jena, Germany). Infarct volume was measured according to the Cavalieri method [29]. The following formulas were used (VH: volume of a hemisphere; VTL: volume of tissue lost): VH= average (average area of a complete coronal hemisphere – ventricles – area of damage) × interval between sections × number of sections VTL= tissue remaining in normal hemisphere – tissue remaining in injured hemisphere Immunostaining Serial coronal sections were incubated overnight in primary antibody rabbit anti-GR (Santa Cruz Biotech; 1:50). Controls were performed omitting the primary antibody. Sections were incubated with a secondary anti-rabbit biotinylated antibody (Vectastain ABC Kit: peroxidase rabbit IgG, 1:200), which was detected using the Elite ABC-peroxidase kit (Vectastain ABC Kit, Vector) with DAB as chromogen (according to [30,31] but modified). Sections were examined under light microscopy. Quantitative analysis of GR expression and activation Photographs were captured digitally and panoramic pictures (1× lens) from coronal brain sections were taken (Figure 2A). Three motor cortex areas in the lesion and non-lesion hemispheres were photographed (40×; Figure 3A1-A4). GR expression was measured using ImageJ software (NIH, Bethesda, Maryland, USA). Color images were converted to monochromatic photo images to determine mean gray scale values.

brain sections were taken (Figure 2A). Three motor cortex areas in the lesion and non-lesion hemispheres were photographed (40×; Figure 3A1-A4). GR expression was measured using ImageJ software (NIH, Bethesda, Maryland, USA). Color images were converted to monochromatic photo images to determine mean gray scale values. Figure 2 Chronic stress activates the stress response in absence of GR density alteration. (A), Coronal brain section showing GR immune-histochemistry; (B), total GR quantification; (C), plasma corticosterone levels (CORT). All data are presented as group means ± SEM. * p < 0.05, compared to the Control group. Figure 3 Stress modulates GR activation after ischemic lesion. (A), GR nuclear density; (B), GR nuclear translocation. Lesion increased activated GR density in the lesion hemisphere compared to the non-lesion hemisphere in stress (A1 and A3) and control (A2 and A4) groups. Stress increased GR activation in the lesion hemisphere (A1) when compared to the non-lesion hemisphere (A3) (*p < 0.05). GR is constitutively expressed in the cytoplasm (arrow in B2), and inactive as a GR-HSP complex. GC enters the cell and binds to a GR dissociating the GR-HSP complex (Schema B3). Consequently, GR is activated when a GR-GC dimer is formed and translocated into the nucleus (arrow in B1; schema B3). Scale bars: 10 μm. GR activation was quantified by selecting the cellular stained nucleus (according to [32-34] but modified). Cellular location of GR was analyzed semi-quantitatively (100× lens; Figure 3B1, B2).

Figure 3 Stress modulates GR activation after ischemic lesion. (A), GR nuclear density; (B), GR nuclear translocation. Lesion increased activated GR density in the lesion hemisphere compared to the non-lesion hemisphere in stress (A1 and A3) and control (A2 and A4) groups. Stress increased GR activation in the lesion hemisphere (A1) when compared to the non-lesion hemisphere (A3) (*p < 0.05). GR is constitutively expressed in the cytoplasm (arrow in B2), and inactive as a GR-HSP complex. GC enters the cell and binds to a GR dissociating the GR-HSP complex (Schema B3). Consequently, GR is activated when a GR-GC dimer is formed and translocated into the nucleus (arrow in B1; schema B3). Scale bars: 10 μm. GR activation was quantified by selecting the cellular stained nucleus (according to [32-34] but modified). Cellular location of GR was analyzed semi-quantitatively (100× lens; Figure 3B1, B2). Statistical analysis Statview software version 5.0 (SAS Institute, 1998) was used to perform analyses of variance (ANOVA), unpaired student t-tests for between-group comparisons and paired t-tests for within-group comparisons in the single pellet reaching task, and for GR expression patterns across hemispheres. A p-value of less than 0.05 was chosen as significance level. All data are presented as mean ± standard error of the mean.

ANOVA), unpaired student t-tests for between-group comparisons and paired t-tests for within-group comparisons in the single pellet reaching task, and for GR expression patterns across hemispheres. A p-value of less than 0.05 was chosen as significance level. All data are presented as mean ± standard error of the mean. Results Skilled reaching Reaching success There was a significant difference in the success rate between groups (F(1,28) = 56.34, p < 0.001). Stress animals had lower reaching success than Control animals (pre-lesion stress: day 1, t(21) = -3.80, p < 0.001; day 2, t(21) = -5.58, p < 0.001; day 3, t(21) = -5.58, p < 0.001; day 4, t(21) = -4.84, p < 0.001; day 5, t(21) = -5.72, p < 0.001; day 6, t(21) = -4.45, p < 0.001; day 7, t(21) = -3.48, p < 0.01; post-lesion stress: day 15, t(21) = -2.91, p < 0.01; day 16, t(21) = -2.14, p < 0.05; day 17, t(21) = -2.94, p < 0.01; day 20, t(21) = -2.63, p < 0.05; day 21, t(21) = -2.28, p < 0.05; Figure 1B). There was a decline from baseline to post-lesion testing in Stress animals, and from pre-lesion to post-lesion in Control animals (Stress: t(10) = 10.22, p < 0.001, t(10) = 5.14, p < 0.001; Control: t(11) = 10.77, p < 0.001).

-2.94, p < 0.01; day 20, t(21) = -2.63, p < 0.05; day 21, t(21) = -2.28, p < 0.05; Figure 1B). There was a decline from baseline to post-lesion testing in Stress animals, and from pre-lesion to post-lesion in Control animals (Stress: t(10) = 10.22, p < 0.001, t(10) = 5.14, p < 0.001; Control: t(11) = 10.77, p < 0.001). Number of pellets obtained There was significant difference in the number of pellets obtained between groups (F(1,28) = 21.71, p < 0.001). The Stress group showed a decrease on days 2, 3, 4 and 5 pre-lesion when compared to Controls [day 2: t(21) = -4.36, p < 0.001; day 3: t(21) = -3.48, p < 0.01; day 4: t(21) = -2.52, p < 0.05; day 5: t(21) = -3.02, p < 0.01; (Figure 1C)]. Performance decreased from baseline to pre-lesion in Stress animals, and from pre-lesion to post-lesion in Controls (Stress: t(10) = 4.56, p < 0.001; Control: t(11) = 5.93, p < 0.001).

t(21) = -4.36, p < 0.001; day 3: t(21) = -3.48, p < 0.01; day 4: t(21) = -2.52, p < 0.05; day 5: t(21) = -3.02, p < 0.01; (Figure 1C)]. Performance decreased from baseline to pre-lesion in Stress animals, and from pre-lesion to post-lesion in Controls (Stress: t(10) = 4.56, p < 0.001; Control: t(11) = 5.93, p < 0.001). Number of Attempts Overall there was a significant group difference in the number of attempts to grasp a food pellet (F(1,28) = 17.14, p < 0.001). Stress animals made more attempts than Controls on all pre-lesion days and days 6 and 15 post-lesion (pre-lesion: day 1, t(21) = 2.42, p < 0.05; day 2, t(21) = 3.59, p < 0.01; day 3, t(21) = 3.57, p < 0.01; day 4, t(21) = 3.61, p < 0.001; day 5, t(21) = 5.56, p < 0.001; day 6, t(21) = 3.87, p < 0.001; day 7, t(21) = 5.83, p < 0.001; post-lesion: day 6, t(21) = 2.08, p < 0.05; day 15, t(21) = 2.57, p < 0.05; Figure 1D). Controls made more attempts after lesion than pre-lesion (t(11) = -5.27, p < 0.001). Stress animals made more attempts pre-lesion than at baseline (t(10) = -4.92, p < 0.001). Infarct size The lesion included the primary and secondary motor cortex as well as the forelimb and hind limb areas of somatosensory cortex (Figure 2A). There was no significant difference in infarct size between groups. The Stress group lost on average 18.02 mm3 of tissue and the Control group lost on average 12.14 mm3 of tissue. There was no correlation between infarct size and skilled reaching success (r = 0.37).

hind limb areas of somatosensory cortex (Figure 2A). There was no significant difference in infarct size between groups. The Stress group lost on average 18.02 mm3 of tissue and the Control group lost on average 12.14 mm3 of tissue. There was no correlation between infarct size and skilled reaching success (r = 0.37). Plasma corticosterone The Stress group had higher plasma corticosterone (CORT) levels compared to Controls post-lesion (t(21) = 2.81, p < 0.05; Figure 2C). Glucocorticoid receptor expression Stress animals showed slightly lower total GR expression than Controls (Figure 2B). There was higher GR activation in the lesion hemisphere in Stress animals when compared to the non-lesion hemisphere (t(7) = 3.43, p < 0.05; Figure 3A1-A3). There was slightly higher GR activation in the lesion hemisphere of Stress animals, and lower GR activation in the non-lesion hemisphere, when compared to Control animals (Figure 3A). Semi-quantitative analysis allowed the discrimination between nuclear (Figure 3B1) and cytoplasm (Figure 3B2) GR sub-location. GR nuclear density was increased near the lesion site.

Glucocorticoid receptor expression Stress animals showed slightly lower total GR expression than Controls (Figure 2B). There was higher GR activation in the lesion hemisphere in Stress animals when compared to the non-lesion hemisphere (t(7) = 3.43, p < 0.05; Figure 3A1-A3). There was slightly higher GR activation in the lesion hemisphere of Stress animals, and lower GR activation in the non-lesion hemisphere, when compared to Control animals (Figure 3A). Semi-quantitative analysis allowed the discrimination between nuclear (Figure 3B1) and cytoplasm (Figure 3B2) GR sub-location. GR nuclear density was increased near the lesion site. Discussion The present findings provide a mechanistic link between stress-induced motor disability and biochemical changes in a rat model of stroke. At the behavioural level, stress diminished skilled limb use in naïve rats and hindered motor recovery from an ischemic infarct thus confirming previous studies [35,21,24,25,37]. Furthermore, chronic stress increased circulating corticosterone concentrations. At the molecular level, chronic restraint stress modulated GR activation of central motor systems that cause permanent alterations in GC susceptibility. These findings support the notion that stress drives biochemical changes that are accompanied by lasting functional loss.

ed circulating corticosterone concentrations. At the molecular level, chronic restraint stress modulated GR activation of central motor systems that cause permanent alterations in GC susceptibility. These findings support the notion that stress drives biochemical changes that are accompanied by lasting functional loss. The main finding of the present study revealed an additive effect of stress and stroke to enhance GR activation in the lesion hemisphere. While enhanced GR activation occurred without concomitant changes in infarct size, diminished recovery in these animals suggests that elevated GC levels are detrimental to functional outcome. Stroke can induce an inflammatory response in the brain, what might affect the immune-endocrine communication, such as GC levels, causing the alteration of essential biological functions. Although corticosteroids may represent a potential treatment for ischemic cerebral edema [38], even physiologically elevated GC levels may modulate edema formation and recovery after ischemic infarct [24,36]. The latter studies suggest that chronic stress and associated immuno-endocrine interactions may promote disease aggravation. These observations are not surprising given the large body of evidence documenting central GC effects. Elevated GC levels may promote pro-inflammatory cell migration, cytokine production, and transcription factor activity in the brain [4] resulting in necrotic cell death [39].

eractions may promote disease aggravation. These observations are not surprising given the large body of evidence documenting central GC effects. Elevated GC levels may promote pro-inflammatory cell migration, cytokine production, and transcription factor activity in the brain [4] resulting in necrotic cell death [39]. High GC levels render neurons more susceptible to neurological insults via disruption of anti-apoptotic factor [40] and neurotrophic factor expression [15,41,42]. GC-mediated regulation of these factors might exaggerate infarct size [43], limit structural plasticity and the capacity to compensate for functional loss [24,37]. In human acute stroke the GC component of the stress response can be harmful, at least when cortisol reaches high blood concentrations within the first days after stroke to exaggerate ischemic injury and neuronal death [12]. Furthermore, corticosteroid treatment has been shown to be ineffective for some survivors from acute ischemic stroke [44]. The concert of anti- versus pro-inflammatory effects of GCs complicates the interpretation of clinical trials.

hin the first days after stroke to exaggerate ischemic injury and neuronal death [12]. Furthermore, corticosteroid treatment has been shown to be ineffective for some survivors from acute ischemic stroke [44]. The concert of anti- versus pro-inflammatory effects of GCs complicates the interpretation of clinical trials. The mechanism of action of GCs is through its binding with the type I mineralocorticoid receptor (MR) and the type II GR [45]. MRs are mostly occupied by GCs at basal levels [46]. GRs have low affinity for GCs and become occupied at the time of stress-induced GC elevation. GRs typically reside in the cytoplasm of the cell, bound to chaperone heat shock proteins until GCs enter the cell. Upon ligand binding the GR-GC complex is released from the chaperone complex to translocate to the cell nucleus (see Figure 3B3). Once in the nucleus, the GR-GC complex binds to the DNA, influencing transcription [47,48]. These changes may be directly linked to lasting alterations in brain function and stroke outcome. Enhanced GR activation in the lesion hemisphere suggests altered sensitivity of the HPA axis induced by stress. GRs are involved with reactive feedback to restore disturbed homeostasis [49]. Modulatory influences of GR, such as GC negative feedback, may dampen the HPA response [50]. Once GCs induce neuronal loss, low GR density could reduce the HPA axis negative feedback, thus promoting further GC production [51,52].

by stress. GRs are involved with reactive feedback to restore disturbed homeostasis [49]. Modulatory influences of GR, such as GC negative feedback, may dampen the HPA response [50]. Once GCs induce neuronal loss, low GR density could reduce the HPA axis negative feedback, thus promoting further GC production [51,52]. The balance of MR- and GR-mediated effects exerted by GCs is critical for homeostatic control [53]. Here we found increased GR activation in the lesion hemisphere, suggesting effective negative HPA feedback regulation that confines deleterious effects of stress to protect or promote neuronal survival. This notion is supported by considerable motor recovery in the absence of exaggerated infarct size after stress. Additional to central regulation of the stress response, GRs are critical for physiological sustainability. GRs modulate synaptic plasticity associated with the plasmatic membrane acting through second messengers to regulate signal transduction cascades [54]. Increased GR expression in the lesion hemisphere could therefore promote plasticity and maintain integrity of existing pathways thus facilitating compensation after stroke.

modulate synaptic plasticity associated with the plasmatic membrane acting through second messengers to regulate signal transduction cascades [54]. Increased GR expression in the lesion hemisphere could therefore promote plasticity and maintain integrity of existing pathways thus facilitating compensation after stroke. Since GC therapy represents a treatment option for clinical stroke to contain inflammatory processes, the investigation of the role of stress and lesion in GR regulation is critical to explore future therapeutic avenues. The current findings suggest that one possible mechanism to affect stroke outcome is brain GR modulation, and consequent alteration of stress responsiveness and GC therapy. The investigation of GR regulation therefore is a critical step towards designing effective therapies and rehabilitation strategies for stroke survivors. Competing interests The authors declare that they have no competing interests. Authors' contributions FCRZ: Experimental design, behavioural testing, immuno-histochemical procedures, data analysis and interpretation, statistical analysis, manuscript preparation. N-FM: Brain microtomy, immuno-histochemical procedures, assistance with data acquisition. NB: Brain microtomy, immuno-histochemical procedures, assistance with data acquisition. GAM: Experimental design, project supervision, data interpretation, manuscript preparation. All authors read and approved the final manuscript.

-FM: Brain microtomy, immuno-histochemical procedures, assistance with data acquisition. NB: Brain microtomy, immuno-histochemical procedures, assistance with data acquisition. GAM: Experimental design, project supervision, data interpretation, manuscript preparation. All authors read and approved the final manuscript. Acknowledgements We gratefully acknowledge Douglas F. Bray and Byron Lee for valuable advice about imuno-histochemical quantification analysis. This research was supported by the Alberta Heritage Foundation for Medical Research (FCRZ, NFM, GAM), the Hotchkiss Brain Institute (FCRZ), the Norlien Foundation (FCRZ), the Canadian Stroke Network (GAM), and the Canadian Institutes of Health Research (GAM).

Background Recently transcranial direct current stimulation (tDCS), a non-invasive brain stimulation technique has been applied to facilitate skill acquisition and motor learning [1-4]. TDCS modulates cortical excitability in a polarity dependent manner, that is, anodal tDCS increases but cathodal tDCS decreases cortical excitability at stimulated sites [5,6]. Furthermore, anodal tDCS applied to the contralateral motor cortex of the motor performing hand [1,3] or cathodal tDCS applied to the ipsilateral motor cortex [7,8] have been reported to improve motor performance in healthy subjects. This concept has also demonstrated in stroke patients. In these studies, anodal tDCS applied to the affected motor cortex [9,10] or cathodal tDCS applied to the unaffected motor cortex [10,11] to diminish inter-hemispheric trans-callosal inhibition [7,12] was shown to improve affected hand motor performance. Given the findings of the above-mentioned reports, it is possible that a combination of anodal tDCS to the contralateral motor cortex and cathodal tDCS to the ipsilateral motor cortex of the motor performing hand would improve motor performance more than the application of anodal tDCS to the contralateral motor cortex alone. Therefore, the purpose of this study was to test the above hypothesis using an implicit finger-sequence learning paradigm [13] in healthy subjects. Furthermore, functional recovery after stroke is a motor relearning process [14,15], and thus, it was hoped that the results of this study may be applicable to stroke patients.

efore, the purpose of this study was to test the above hypothesis using an implicit finger-sequence learning paradigm [13] in healthy subjects. Furthermore, functional recovery after stroke is a motor relearning process [14,15], and thus, it was hoped that the results of this study may be applicable to stroke patients. Methods Subjects Eleven healthy young adults (three males, age 26.3 years ± 3.6 S.D.) without any medical or neurological disease participated in this study. All were right handed, as determined by the Edinburgh Handedness Inventory [16]. The experimental protocol was approved by the Institutional Review Board at our hospital and written informed consent was obtained from all subjects. Experimental design After being familiarized with the experimental setting, each of the 11 subjects underwent a randomized crossover experiment of Uni-tDCS, Bi-tDCS, or sham stimulation separated by at least 48 hours. Orders of stimulation conditions were counterbalanced (Figure 1). Figure 1 Experimental design. Motor sequence performance improvement was measured by calculating the ratio of reaction times for the predetermined repeating sequences and a random sequence (S/R block) at shaded blocks. R' = familiarizing random sequence block; R = random sequence block; S = predetermined repeating sequence block.

perimental design. Motor sequence performance improvement was measured by calculating the ratio of reaction times for the predetermined repeating sequences and a random sequence (S/R block) at shaded blocks. R' = familiarizing random sequence block; R = random sequence block; S = predetermined repeating sequence block. TDCS was delivered through two saline-soaked, sponge electrodes (25 cm2) using a constant-current stimulator (Phoresor®ΙΙ PM850; IOMED® Inc., Salt Lake City, Utah) as previously described [9]. Although Phoresor®ΙΙ PM850; IOMED® Inc. is not designed for tDCS, it has been widely used for tDCS studies. This device can control current intensity, duration, and ramp up time [17]. First, we placed three electrodes over C3 (corresponding to the left M1), C4 (corresponding to the right M1) of the international 10-20 EEG system, and the right supra-orbital region. For Uni-tDCS (2 mA for 20 minutes) and sham stimulation (2 mA for 1 minute), we used an anode electrode over C3 and a cathode over the right supraorbital region, and for Bi-tDCS (2 mA for 20 minutes) we used an anode over C3 and a cathode over C4.

the international 10-20 EEG system, and the right supra-orbital region. For Uni-tDCS (2 mA for 20 minutes) and sham stimulation (2 mA for 1 minute), we used an anode electrode over C3 and a cathode over the right supraorbital region, and for Bi-tDCS (2 mA for 20 minutes) we used an anode over C3 and a cathode over C4. The current was slowly increased to 2 mA from the onset of stimulation in a ramp-up like fashion over 30 sec. For real stimulation, the switch was toggled up and down for an additional 30 sec to match the sham procedure, and the current was then maintained at 2 mA for the remainder of the 20 min, whereas during sham stimulation sessions the current was slowly tapered down to zero over 30 sec. This procedure has been demonstrated to prevent subjects differentiating between real and sham stimulation [9,18]. We selected C3 and C4 of the international 10-20 EEG system for stimulation because it has been reported that the primary motor cortex (M1) mediates implicit motor sequence learning [19], and because a neuroimaging study showed that C3 and C4 correspond to the left and right M1 [20]. However, in the present study, stimulation may have extended beyond M1 due to the large electrode size used. The tDCS procedures were administered by a separate investigator who did not participate in outcome measurements or data analysis. Therefore, the subjects and the investigator who determined outcome measures were unaware of the intervention type.

The current was slowly increased to 2 mA from the onset of stimulation in a ramp-up like fashion over 30 sec. For real stimulation, the switch was toggled up and down for an additional 30 sec to match the sham procedure, and the current was then maintained at 2 mA for the remainder of the 20 min, whereas during sham stimulation sessions the current was slowly tapered down to zero over 30 sec. This procedure has been demonstrated to prevent subjects differentiating between real and sham stimulation [9,18]. We selected C3 and C4 of the international 10-20 EEG system for stimulation because it has been reported that the primary motor cortex (M1) mediates implicit motor sequence learning [19], and because a neuroimaging study showed that C3 and C4 correspond to the left and right M1 [20]. However, in the present study, stimulation may have extended beyond M1 due to the large electrode size used. The tDCS procedures were administered by a separate investigator who did not participate in outcome measurements or data analysis. Therefore, the subjects and the investigator who determined outcome measures were unaware of the intervention type. Serial reaction time task We used a serial reaction time task (SRTT) as an outcome measure. The SRTT is a simple task that provides a measure of implicit motor skill learning [21]. Subjects performed a total of 20 blocks of key presses with their right hands, and each block was composed of 10 repetitions of a 12-digit length sequence (Figure 1). Subjects were seated in front of a computer screen and asked to press the key corresponding to the location of asterisks with 4 fingers (2nd - 5th) of the right hand as quickly and as accurately as possible. The task was designed using Superlab pro v.4.0 software (Cedrus Corporation, San Pedro, CA).

equence (Figure 1). Subjects were seated in front of a computer screen and asked to press the key corresponding to the location of asterisks with 4 fingers (2nd - 5th) of the right hand as quickly and as accurately as possible. The task was designed using Superlab pro v.4.0 software (Cedrus Corporation, San Pedro, CA). After familiarization using a random block (R' in Figure 1), subjects were presented with random (R in Figure 1) or predetermined repeating sequence blocks (S in Figure 1) separated by resting 30 sec periods. Next blocks were presented when all keys presses were correct. For the R' and R blocks, an asterisk appeared randomly in one of four locations on a computer screen, whereas an asterisk appeared in a predetermined repeating sequence in an S block. We used three predetermined repeating sequence S blocks (3-4-2-1-2-4-1-3-4-2-1-3/2-4-1-3-2-1-2-1-3-4-3-4/1-2-1-4-2-3-2-4-3-1-4-3), one for each of the three stimulation types (Uni-tDCS, Bi-tDCS, or sham stimulation) in a randomly selected manner. These R'-R-S blocks were presented at baseline (Pre), immediately (Post 1), and 24 hours after stimulation (Post 2). During stimulation, subjects practiced using the same predetermined repeating sequence S blocks (block 4-8 and 10-14) interrupted by one R block. The ratios of reaction times for the predetermined repeating sequence and the random sequence (shaded S block/R block in Figure 1) were used as an outcome measure of motor sequence performance improvements by practice.

the same predetermined repeating sequence S blocks (block 4-8 and 10-14) interrupted by one R block. The ratios of reaction times for the predetermined repeating sequence and the random sequence (shaded S block/R block in Figure 1) were used as an outcome measure of motor sequence performance improvements by practice. Prior to each session, subjects described their levels of attention, perceived general fatigue, hand fatigue, and task difficulty using a numeric rating scale (range 0 ~ 10; 0 = lowest, 10 = highest). Data analysis The mean response time per each trial was calculated to quantify motor sequence performance improvements (motor sequence learning) achieved by repeated practice. The ratios of reaction times of predetermined repeating sequence per random sequence (shaded S block/R block in Figure 1) at Pre, Post 1, and Post 2 relative to baseline were analyzed using the paired t test for each stimulation type (Uni-tDCS versus Bi-tDCS versus sham stimulation) to demonstrate the motor sequence learning effect. Results ANOVARM revealed no effect of INTERVENTIONUni-tDCS, Bi-tDCS, Sham, TIMEPre.Post1, Post 2 or INTERVENTIONUni-tDCS, Bi-tDCS, Sham×TIMEPre.Post1, Post 2 interaction on subjects' perceived attention and general fatigue (P > 0.05). But, there was a significant effect of TIMEPre.Post1, Post2 on hand fatigue and task difficulty, which suggested that subject perceived hand fatigue increased and task difficulty decreased immediate after practice blocks (Table 1).

st1, Post 2 interaction on subjects' perceived attention and general fatigue (P > 0.05). But, there was a significant effect of TIMEPre.Post1, Post2 on hand fatigue and task difficulty, which suggested that subject perceived hand fatigue increased and task difficulty decreased immediate after practice blocks (Table 1). Table 1 Subject perceived levels of attention, general fatigue, hand fatigue, and task difficulty (rated using numeric 0~10 rating scales; 0 = lowest, 10 = highest). Stimulation type ANOVARM P-value Uni-tDCS Bi-tDCS Sham Interv Time Interv X Time Pre Post 1 Post 2 Pre Post 1 Post 2 Pre Post 1 Post 2 Attention 5.2 ± 1.3 5.0 ± 2.3 5.8 ± 1.6 5.6 ± 1.3 5.2 ± 2.1 4.8 ± 2.3 5.8 ± 2.0 5.7 ± 2.0 5.2 ± 1.4 0.598 0.764 0.431 Fatigue 4.9 ± 1.1 4.3 ± 1.8 5.6 ± 2.0 4.6 ± 1.6 4.2 ± 1.5 4.0 ± 1.9 4.7 ± 2.2 5.1 ± 1.4 4.5 ± 2.4 0.168 0.464 0.250 Hand fatigue 5.9 ± 2.3 5.1 ± 2.2 5.7 ± 1.7 6.0 ± 2.0 5.3 ± 2.2 5.1 ± 2.1 5.7 ± 2.3 5.5 ± 2.5 5.7 ± 2.4 0.822 0.046 0.619 Task difficulty 5.7 ± 1.8 5.4 ± 2.3 6.2 ± 1.8 6.0 ± 2.0 4.7 ± 1.7 5.2 ± 2.0 6.0 ± 2.0 5.7 ± 2.5 5.9 ± 2.2 0.295 0.008 0.242 For each stimulation type, mean reaction time shortened during the predetermined repeating sequence blocks, but return to the baseline level during the random sequence blocks (Figure 2). Figure 2 Serial reaction times for each stimulation type. Note that mean reaction time was shortened during the predetermined repeating sequence blocks but returned to baseline level during the random sequence blocks regardless of stimulation type.

Uni-tDCS Bi-tDCS Sham Interv Time Interv X Time Pre Post 1 Post 2 Pre Post 1 Post 2 Pre Post 1 Post 2 Attention 5.2 ± 1.3 5.0 ± 2.3 5.8 ± 1.6 5.6 ± 1.3 5.2 ± 2.1 4.8 ± 2.3 5.8 ± 2.0 5.7 ± 2.0 5.2 ± 1.4 0.598 0.764 0.431 Fatigue 4.9 ± 1.1 4.3 ± 1.8 5.6 ± 2.0 4.6 ± 1.6 4.2 ± 1.5 4.0 ± 1.9 4.7 ± 2.2 5.1 ± 1.4 4.5 ± 2.4 0.168 0.464 0.250 Hand fatigue 5.9 ± 2.3 5.1 ± 2.2 5.7 ± 1.7 6.0 ± 2.0 5.3 ± 2.2 5.1 ± 2.1 5.7 ± 2.3 5.5 ± 2.5 5.7 ± 2.4 0.822 0.046 0.619 Task difficulty 5.7 ± 1.8 5.4 ± 2.3 6.2 ± 1.8 6.0 ± 2.0 4.7 ± 1.7 5.2 ± 2.0 6.0 ± 2.0 5.7 ± 2.5 5.9 ± 2.2 0.295 0.008 0.242 For each stimulation type, mean reaction time shortened during the predetermined repeating sequence blocks, but return to the baseline level during the random sequence blocks (Figure 2). Figure 2 Serial reaction times for each stimulation type. Note that mean reaction time was shortened during the predetermined repeating sequence blocks but returned to baseline level during the random sequence blocks regardless of stimulation type. The mean S/R ratio (ratio of reaction time for a predetermined repeating sequence versus a random sequence) at Pre did not differ significantly for the three stimulation types (P = 0.57 by one way ANOVA). When comparing sham and Uni-tDCS or sham and Bi-tDCS at Post 2 using the paired t test, no significant differences were found (Sham vs. Uni-tDCS, P = 0.49; Sham vs. Bi-tDCS, P = 0.19). Furthermore, no significant S/R ratio difference was observed between Uni-tDCS and Bi-tDCS at Post 2. ANOVA also revealed no significant differences between stimulation types at Post 2 (P = 0.65).

2 using the paired t test, no significant differences were found (Sham vs. Uni-tDCS, P = 0.49; Sham vs. Bi-tDCS, P = 0.19). Furthermore, no significant S/R ratio difference was observed between Uni-tDCS and Bi-tDCS at Post 2. ANOVA also revealed no significant differences between stimulation types at Post 2 (P = 0.65). We believe these negative findings were caused by small subject numbers. Therefore, we performed paired t-testing between Pre and Post1 or Pre and Post 2 for each stimulation type. We found that S/R ratio significantly decreased for all stimulation types at Post 1 (P < 0.01), but at Post 2, this reduction was significant after Uni-tDCS (P < 0.01) and Bi-tDCS (P < 0.01), and only marginally significant after Sham (P = 0.05), which suggested that motor sequence performance improvement was maintained by Uni-tDCS and Bi-tDCS, but only partially by sham stimulation. However, no significant S/R ratio differences were observed between Uni-tDCS and Bi-tDCS at Post 1 and Post 2 (Figure 3).

1), and only marginally significant after Sham (P = 0.05), which suggested that motor sequence performance improvement was maintained by Uni-tDCS and Bi-tDCS, but only partially by sham stimulation. However, no significant S/R ratio differences were observed between Uni-tDCS and Bi-tDCS at Post 1 and Post 2 (Figure 3). Figure 3 Motor sequence performance improvement. The Y axis represents the ratios of the reaction times of the predetermined repeating sequence versus a random sequence (shaded S versus R blocks in Figure 1). The asterisk (*) represents P < 0.05 between Pre and Post 1 sessions for all stimulation types by the paired t test, which suggests motor sequence learning occurred at immediately after stimulation regardless of stimulation type. Cross(+) represents P < 0.05 between Pre and Post 2 sessions for Uni-tDCS and Bi-tDCS, but not for Sham stimulation by the paired t test, which suggests that motor sequence performance improvements were maintained after Uni-tDCS and Bi-tDCS, but not after Sham stimulation. Discussion In this study, we evaluated the combining effect of anodal tDCS applied to the contralateral motor cortex and cathodal tDCS applied to the ipsilateral motor cortex (Bi-tDCS) on the implicit motor learning process, and compared this with the effect of anodal tDCS applied to the contralateral motor cortex alone (Uni-tDCS). We found that combined bilateral stimulation did not improve implicit motor learning more than unilateral stimulation, but that Bi-tDCS and Uni-tDCS did improve implicit motor learning more than sham stimulation.

nd compared this with the effect of anodal tDCS applied to the contralateral motor cortex alone (Uni-tDCS). We found that combined bilateral stimulation did not improve implicit motor learning more than unilateral stimulation, but that Bi-tDCS and Uni-tDCS did improve implicit motor learning more than sham stimulation. Initially we hypothesized that decreasing inter-hemispheric trans-callosal inhibition from non-dominant to dominant M1 by right hemisphere cathodal tDCS in combination with increasing the excitability of dominant M1 by left hemisphere anodal tDCS would improve implicit motor learning more than increasing the excitability of dominant M1 by left hemisphere anodal tDCS alone. However, our findings did not support this hypothesis, although it should be noted that Bi-tDCS showed a tendency to more improve implicit motor learning than Uni-tDCS, as is shown by the raw data presented in Figures 1 and 2.

ng more than increasing the excitability of dominant M1 by left hemisphere anodal tDCS alone. However, our findings did not support this hypothesis, although it should be noted that Bi-tDCS showed a tendency to more improve implicit motor learning than Uni-tDCS, as is shown by the raw data presented in Figures 1 and 2. Recently, Vines et al. [22] found that Bi-tDCS improved motor performance of the non-dominant hand more than Uni-tDCS in healthy subjects, whereas in the present study only a weak trend was found. We believe that this discrepancy may have been caused by the different task paradigm used or the use of the non-dominant hand, because in this previous study a 5 digit sequence and non-dominant left hands were used. It is possible that dominant hands might have already reached a ceiling prior to stimulation [1,7,8], or that interhemispheric inhibition from non-dominant to dominant hemisphere might be trivial as compared with inhibition from dominant to non-dominant hemisphere [22-24]. It is also probable that healthy young subjects are more likely to display the ceiling effect than older subjects or stroke patients in implicit motor learning process. It is also possible that our study was underpowered due to small number of subjects recruited.

bition from dominant to non-dominant hemisphere [22-24]. It is also probable that healthy young subjects are more likely to display the ceiling effect than older subjects or stroke patients in implicit motor learning process. It is also possible that our study was underpowered due to small number of subjects recruited. Although the clinical applications of tDCS have expanded, the effects of electrode montages have not been well established. One unique aspect of tDCS application is the use of an electrodes pair. The belief that tDCS increases excitability just at the stimulating site under the anode and decreases excitability under the cathode is changing. Now it is generally believed that tDCS has both a regional effect on the cortex underlying electrodes and a remote effect on brain regions between electrodes [25-27]. Moreover, recently Moliadze et al. addressed the role of the "return" electrode position on tDCS induced excitability changes under an the "active" electrode using a computer model, and showed that the position and size of the ''return" electrode affects the electric field distribution across the entire cortex, and the electric field distribution in cortex directly under the "active"electrode [28].

osition on tDCS induced excitability changes under an the "active" electrode using a computer model, and showed that the position and size of the ''return" electrode affects the electric field distribution across the entire cortex, and the electric field distribution in cortex directly under the "active"electrode [28]. According to this view, the anodal effects on C3 during Uni-tDCS and Bi-tDCS differed in the present study, because the return cathode was positioned over the contralateral supraorbital region for Uni-tDCS and over C4 for Bi-tDCS. We used this Bi-tDCS electrode montage hoping to simultaneously up-regulate excitability of the motor cortex over C3 (anodal stimulation), and to down-regulate excitability of the motor cortex over C4 (cathodal stimulation). Recently Lindenberg et al. [29] also used the same Bi-tDCS electrode montage used in the present study.

this Bi-tDCS electrode montage hoping to simultaneously up-regulate excitability of the motor cortex over C3 (anodal stimulation), and to down-regulate excitability of the motor cortex over C4 (cathodal stimulation). Recently Lindenberg et al. [29] also used the same Bi-tDCS electrode montage used in the present study. Extracephalic electrode montages offer another approach [30]. According to this method one electrode is placed over the cortex and the other over an extracephalic region, such as, a shoulder or mastoid process. It would be interesting to compare Uni-tDCS and Bi-tDCS using this extracephalic electrode montage in the future. However, in the present study, we could not exclude the possibility that during Uni-tDCS, the reference cathode on the right supraorbital region, which corresponds to the right prefrontal cortex, might have had some beneficial effect on implicit motor sequence learning, which would have diluted the additive effect of Bi-tDCS over Uni-tDCS. Additional experimental studies are required to investigate the effects of various electrode montages on the effects of tDCS.

hich corresponds to the right prefrontal cortex, might have had some beneficial effect on implicit motor sequence learning, which would have diluted the additive effect of Bi-tDCS over Uni-tDCS. Additional experimental studies are required to investigate the effects of various electrode montages on the effects of tDCS. Another possibility is that the electrode size over M1 was large enough to cover the pre-motor cortex, which also would have had a diluting effect on Bi-tDCS versus Uni-tDCS. In a neuroimaging study, it was found that finger sequence performance recruits the pre-motor and supplementary motor cortex as well as the primary motor cortex [31], although we only intended to stimulate M1 as performed in a previous study [3], in which it was shown that finger sequence performance task results can be influenced by modulating M1 activity. In the present study, the reaction times of predetermined repeating sequence in the SRTT decreased regardless of stimulation type, whereas the reaction times of random sequences did not, which implies that implicit motor learning had occurred during training. However decreases in reaction times immediately after sham stimulation tended to diminish at 24 hours, but were maintained after Uni-tDCS and Bi-tDCS, which suggest that tDCS might consolidate implicit motor learning more than Sham stimulation, which is in accordance with a previous report [4]. Our results also reveal that tDCS mainly affected motor performance speed rather than accuracy.

tended to diminish at 24 hours, but were maintained after Uni-tDCS and Bi-tDCS, which suggest that tDCS might consolidate implicit motor learning more than Sham stimulation, which is in accordance with a previous report [4]. Our results also reveal that tDCS mainly affected motor performance speed rather than accuracy. Subject attention levels could also have contributed to SRTT results, though these were similar across sessions as determined by our numerical rating scale, and thus, we believe that subject attention level differences were adequately taken into account. TDCS is easily administered, comfortable for patients, relatively inexpensive and can be administered in combination with rehabilitative training [18], and for these reasons was recently introduced as an adjuvant strategy for hand motor rehabilitation after stroke [9,10]. Our results might be relevant to stroke hand motor rehabilitation, although its relevance is limited by potential differences in the implicit motor learning process between healthy subjects and stroke patients. Conclusions In conclusion, no significant difference was observed between Uni-tDCS and Bi-tDCS in terms of inducing implicit motor sequence learning, although both Uni-tDCS and Bi-tDCS led to greater consolidation of the learned motor sequences than sham stimulation. The findings of the present study need to be tested in the context of stroke hand motor rehabilitation. Competing interests The authors declare that they have no competing interests.

Conclusions In conclusion, no significant difference was observed between Uni-tDCS and Bi-tDCS in terms of inducing implicit motor sequence learning, although both Uni-tDCS and Bi-tDCS led to greater consolidation of the learned motor sequences than sham stimulation. The findings of the present study need to be tested in the context of stroke hand motor rehabilitation. Competing interests The authors declare that they have no competing interests. Authors' contributions NJP designed the study and EKK carried out the study and analyzed the data. Both authors drafted the manuscript, and finally read and approved the last version manuscript. Acknowledgements and Funding This study was supported by a grant of the Korean Health Technology R&D Project, Ministry for Health, Welfare & Family Affairs, Republic of Korea (Grant No: A101901).

Summary of the scientific contributions to the NEUROWIND meeting 2010: Contributions in the fields of neuroimmunology and neurodegeneration T cell driven autoimmune inflammation in the CNS has widely been investigated in the model of experimental autoimmune encephalomyelitis (EAE) [1]. During decades of EAE research, it has been established that autoreactive T cells are activated in the peripheral immune tissue, then enter the CNS compartment and - upon local re-activation - acquire the ability to invade the CNS parenchyma and exert effector functions. Only with the advent of modern imaging techniques has it become possible to actually visualize the individual steps of T cell activation in the lymph nodes, of crossing the blood brain barrier, and of interaction between autoreactive T cells and their molecular targets within the CNS. Alexander Flügel has adapted the model of adoptive transfer EAE for imaging purposes making inflammatory processes accessible to two-photon-microscopy in situ. By retroviral expression of fluorescent proteins in encephalitogenic T cells, these T cells were visualized in vivo by two photon microscopy [2]. Christian Schläger from Alexander Flügel's group showed two-photon scanning data providing evidence that in the CNS vasculature encephalitogenic T cells tended to crawl against the blood stream before they left the vessel lumen in order to enter the perivascular space. Here, the cells appeared to be scanning their environment and only upon productive contact with antigen presenting cells that presented the appropriate antigen, T cells were instructed to infiltrate into the CNS parenchyma. It is becoming increasingly clear that many features of leukocyte extravasation in the CNS vasculature are unique and distinct from leukocyte extravasation in other vascular territories [3]. The advanced imaging tools that are now available hold promise to address current questions of T lymphocyte biology at the blood brain barrier: Why do lymphocytes move against the blood stream in the CNS microvasculature? Do lymphocytes trespass the endothelial barrier in a paracellular or transcellular way?

ories [3]. The advanced imaging tools that are now available hold promise to address current questions of T lymphocyte biology at the blood brain barrier: Why do lymphocytes move against the blood stream in the CNS microvasculature? Do lymphocytes trespass the endothelial barrier in a paracellular or transcellular way? How and to what extent do T cells become activated in the perivascular space? The technology of two-photon-microscopy even allows monitoring immune cell-target interactions within the CNS parenchyma. In a recently published study, Volker Siffrin from the group of Frauke Zipp investigated the interaction of encephalitogenic CD4+ T cells with neuronal structures in the brain stem in vivo. Interestingly, myelin antigen reactive (2D2) T cells of the Th17 phenotype were able to interact with (and damage) axons. While IFN-γ producing Th1 cells failed to induce neuronal apoptosis, Th17 cells were very efficient in promoting axonal damage. The mechanism of lesion development has not yet been entirely unraveled. While the interaction between CD4+ T cells and axons was independent of the T cell receptor (which in this case was MOG35-55 specific) and thus not restricted by MHC class II expression on axons, ICAM-1 expression by axons and LFA-1 expression by T cells was critically required for Th17-axonal interaction. Axons responded to Th17 cell-mediated attack by Ca2+ influx, which was partially reversible by blockade of NMDA receptors[4]. Thus, Th17 cells exerted effector functions in the CNS that appeared to be unique to this effector T cell subset.

LFA-1 expression by T cells was critically required for Th17-axonal interaction. Axons responded to Th17 cell-mediated attack by Ca2+ influx, which was partially reversible by blockade of NMDA receptors[4]. Thus, Th17 cells exerted effector functions in the CNS that appeared to be unique to this effector T cell subset. In order to test the functional relevance of susceptibility genes identified in the genome-wide association studies in MS, it is a promising approach to investigate whether the expression level of the corresponding gene products on T cells correlates with an altered functional phenotype of these cells. Melanie Piedavent from the group of Manuel Friese analysed the expression of CD226 on human and mouse CD4+ and CD8+ T cells. CD226 interacts with its ligand CD155 on antigen presenting cells and has a role as a costimulatory molecule. The nonsynonymous single nucleotide polymorphism (SNP) rs763361/Gly307Ser in exon 7 of CD226 leads to the substitution of serine for glycine in the amino acid sequence of CD226 and has been associated with increased risk for type 1 diabetes, MS, rheumatoid arthritis and autoimmune thyroid disease [5]. Both in mouse and in human CD4+ T cells, low and high expression of CD226 segregated with markers of naive and antigen experienced/memory T cells, respectively. CD8+ T cells expressed high amounts of CD226 in a constitutive manner. The functional consequences of rs763361/Gly307Ser are not known. It is possible that the amino acid substitution at position 307 alters the phosphorylation sites of CD226 at positions 322 and 329. Alternatively, an altered expression pattern of CD226 could be induced. Using a mouse model, the functional consequences of Gly307Ser can now be tested in vivo.

61/Gly307Ser are not known. It is possible that the amino acid substitution at position 307 alters the phosphorylation sites of CD226 at positions 322 and 329. Alternatively, an altered expression pattern of CD226 could be induced. Using a mouse model, the functional consequences of Gly307Ser can now be tested in vivo. The role of γδ T cells in EAE has recently been investigated in more detail. γδ T cells have drawn attention since a subset of γδ T cells was identified to be highly responsive to IL-23, which is known to be a potent driver of autoimmunity and chronic inflammation. Thus, the role of γδ T cells in models of autoimmunity has been revisited. Franziska Petermann from the group of Thomas Korn could show that γδ T cells that respond to IL-23 were very efficient in inhibiting Treg responses. As a result, adaptive immune responses flared up in a milieu that was enriched in IL-23R+γδ T cells [6]. While the mechanism of this particular function of γδ T cells has to be further investigated, the role of IL-23R+γδ T cells in restraining Treg responses was clinically relevant. Tcrd KO mice that lack γδ T cells had exaggerated Treg responses. Conversely, deletion of Tregs in Tcrd KO mice restored full susceptibility to EAE [6].

echanism of this particular function of γδ T cells has to be further investigated, the role of IL-23R+γδ T cells in restraining Treg responses was clinically relevant. Tcrd KO mice that lack γδ T cells had exaggerated Treg responses. Conversely, deletion of Tregs in Tcrd KO mice restored full susceptibility to EAE [6]. In addition to T cells, B cells are increasingly recognized as important players in neuroimmunological diseases. This concept is also supported by the therapeutic efficacy of the B cell depleting anti CD20 antibody rituxan in neuroimmunological disorders. Miguel Maurer from the group of Jan Lünemann, Zurich analyzed the B cell repertoire after rituxan treatment of anti-myelin associated glycoprotein (MAG) antibody positive paraproteinemic neuropathy, an autoimmune disorder of the peripheral nervous system characterized by the presence of antibodies against myelin associated glycoprotein MAG. Rituxan did not influence the B cell receptor repertoire, but reduced clonal expansions of IgM positive memory B cells with reactivity against MAG protein.

teinemic neuropathy, an autoimmune disorder of the peripheral nervous system characterized by the presence of antibodies against myelin associated glycoprotein MAG. Rituxan did not influence the B cell receptor repertoire, but reduced clonal expansions of IgM positive memory B cells with reactivity against MAG protein. The EAE model is an excellent model to investigate T cell development and T cell regulation in vivo. However, the role of antibodies in autoimmune neuroinflammation is not well addressed in classical MOG35-55 induced EAE. Moreover, since in MS the most relevant autoantigen is still unknown, the function of EAE as a model for MS is limited in various aspects. On the other hand, a considerable body of knowledge has emerged in the recent years on the target antigen in neuromyelitis optica (NMO) which has been regarded as a variant of MS. However, NMO is probably a distinct disease because the target autoantigen is aquaporin-4 (AQP4) which is not a myelin antigen. AQP4 is a water channel protein which is expressed in astrocytic endfeet of the lamina gliae limitans and thus plays an important part in the function of the blood/brain- and CSF/brain-barriers [7]. Antibodies to AQP4 (NMO-IgG) have been identified in sera of NMO patients and were proven to be a useful biomarker for NMO since NMO-IgG are negative in MS patients [8,9]. NMO-IgG have now been included in the diagnostic criteria for NMO [10]. However, it has still been unclear whether antibodies to AQP4 that are usually not produced intrathecally are pathogenetically relevant. Several laboratories have designed experiments in order to test whether NMO-IgG was able to induce damage to astrocytes [11-13]. In one approach, which was presented by Claudia Wrzos from the group of Christine Stadelmann, a subclinical EAE was induced in experimental rats by active immunization with MBP72-85 in CFA followed by intravenous transfer of either control IgG or recombinant monoclonal anti-AQP4 IgG. Recombinant anti-AQP4 antibodies were engineered from heavy and light chain genes isolated from intrathecal plasma cells of NMO patients. The heavy and light chain genes were cloned into human framework cassettes and expressed in HEK293 cells. Recombinant immunoglobulins recognized both the M1 and M23 translational isoform of AQP4. Only when rats received anti-AQP4 antibodies, they show astrocytic damage, Ig deposition, and complement activation at the blood brain barrier in histopathological analyses.

human framework cassettes and expressed in HEK293 cells. Recombinant immunoglobulins recognized both the M1 and M23 translational isoform of AQP4. Only when rats received anti-AQP4 antibodies, they show astrocytic damage, Ig deposition, and complement activation at the blood brain barrier in histopathological analyses. In a second approach, recombinant anti-AQP4 antibodies were injected intrathecally together with human complement. Here, astrocyte loss was detected as early as 1 h after injection and oligodendrocyte apoptosis (NogoA+caspase-3+) as early as 3 h after injection. These experiments were among the first to suggest that NMO-IgG might have pathogenic relevance beyond their great value as biomarker. Thus, astrocytes at the blood brain barrier appear to be a prime target of the inflammatory process in NMO.

oligodendrocyte apoptosis (NogoA+caspase-3+) as early as 3 h after injection. These experiments were among the first to suggest that NMO-IgG might have pathogenic relevance beyond their great value as biomarker. Thus, astrocytes at the blood brain barrier appear to be a prime target of the inflammatory process in NMO. The blood brain barrier (BBB) can be the primary target of an autoimmune reaction - as in NMO. However, the blood brain barrier is also crucial in modulating the pathogenic process in a series of inflammatory, ischemic, and degenerative diseases. Therefore, it is essential to understand the function of the BBB in health and disease. In the laboratory of Sven Meuth, it was found that a member of the two-pore domain potassium channel family (K2P), namely TWIK-related K+-channel gene (TREK-1) is expressed on murine and human endothelial cells. Inhibition of channel activity by pharmacological strategies or during inflammation was associated with a decreased transendothelial resistance (TER) in an in vitro model of the BBB. Activation of channel activity resulted in increased TER and decreased transmigration of immune cells in the same model. Translated to an in vivo model Stefan Bittner demonstrated an enhanced EAE disease course in TREK-/- mice after MOG35-55 immunization while activation of the channel in vivo using riluzole and/or α-linolenic acid resulted in a significantly ameliorated EAE phenotype with reduced cellular infiltrates.

in the same model. Translated to an in vivo model Stefan Bittner demonstrated an enhanced EAE disease course in TREK-/- mice after MOG35-55 immunization while activation of the channel in vivo using riluzole and/or α-linolenic acid resulted in a significantly ameliorated EAE phenotype with reduced cellular infiltrates. Dirk Hermann presented data on the regulation of luminal and abluminal ATP binding cassette transporters in CNS endothelial cells. ABCB1 is expressed in the luminal membrane and ABCC1 in the abluminal membrane. Upon ischemia, ABCB1 was up-regulated, while ABCC1 was down-regulated suggesting that the efflux of xenobiotics out of ischemic brain regions was facilitated while the influx of molecules using the ABC transporter system would be severely inhibited [14]. Interestingly, ApoE mediated the regulation of ABC transporters in the luminal and abluminal membranes via ApoE receptor 2 and the deactivation of JNK1/2 by dephosphorylation. As a consequence, ApoE KO mice showed decreased expression of ABCB1 and increased expression of abluminal ABCC1 upon ischemic brain injury. Thus, modulation of the ABC system appears to be possible by targeting ApoE which has the role of a molecular switch. This system could be exploited to facilitate the delivery of neuroprotective drugs into ischemic brain regions.

xpression of ABCB1 and increased expression of abluminal ABCC1 upon ischemic brain injury. Thus, modulation of the ABC system appears to be possible by targeting ApoE which has the role of a molecular switch. This system could be exploited to facilitate the delivery of neuroprotective drugs into ischemic brain regions. In addition to the investigation of immune cells and the blood brain barrier, studies on the target cells of the nervous system have been in the focus of interest in neuroimmunological research. To investigate the role of cells of the oligodendrocyte lineage, Karin Hagemeier from Tanja Kuhlmann`s group in Muenster presented a new co-culture system with primary oligodendrocyte precursor cells and retinal ganglion cells allowing for the analysis of oligodendrocyte - neuron interaction on a single animal basis with high cell purity. In particular, a focus of interest has been on factors influencing cell death of oligodendrocytes. Here, p53 induced pro-apoptotic member of the Bcl2 family (PUMA) is an interesting candidate and the analysis of PUMA deficient cells in culture and PUMA deficient mice in the cuprizone model of de- and remyelination in vivo will reveal the role of this factor in the regulation of oligodendrocyte survival.

dendrocytes. Here, p53 induced pro-apoptotic member of the Bcl2 family (PUMA) is an interesting candidate and the analysis of PUMA deficient cells in culture and PUMA deficient mice in the cuprizone model of de- and remyelination in vivo will reveal the role of this factor in the regulation of oligodendrocyte survival. In order to understand the role of microglial cells in remyelination, the cuprizone model was also investigated in Martin Stangel's laboratory. During 6 weeks of cuprizone feeding, toxic demyelination is induced in the absence of blood brain barrier breakdown. Remyelination occurs when cuprizone feeding is stopped and takes 6 weeks to be completed. Demyelination is preceded by microglia activation and proliferation by 2 weeks, and remyelination is preceded by proliferation of oligodendrocyte precursor cells (OPC). Thomas Skripuletz from Martin Stangel's laboratory realized that administration of LPS modulated both de- and remyelination in the cuprizone model. The net effect of LPS was beneficial because demyelination was decelerated and remyelination was enhanced. In histological analyses, the proliferation of microglial cells seemed to be inhibited while the proliferation of OPCs was increased by LPS. Thus, both de- and remyelination are modulated by TLR ligands in an indirect manner. However, it remains to be determined whether the decreased proliferation of microglia by LPS, which is able to cross the intact BBB, is the direct cause of decreased demyelination in this model.

liferation of OPCs was increased by LPS. Thus, both de- and remyelination are modulated by TLR ligands in an indirect manner. However, it remains to be determined whether the decreased proliferation of microglia by LPS, which is able to cross the intact BBB, is the direct cause of decreased demyelination in this model. In view of the degenerative changes in autoimmune demyelination, neuroprotective approaches are of high interest in MS therapy. De-Hyung Lee from the group of Ralf Linker, Erlangen, presented data on mechanisms of action of fumaric acid esters (FAE), which are currently under investigation as new oral disease modifying drug in relapsing remitting MS. Application of FAE in the MOG-EAE model resulted in an ameliorated course of chronic EAE and a preservation of neurons, oligodendrocytes and myelin as well as reduced astrogliosis without direct influence on the immune reaction. These neuroprotective effects were associated with the activation of nuclear factor (erythroid-derived 2)-like 2 (Nrf2), a transcription factor involved in the cellular detoxification pathways and the natural antioxidative response [15].

n as well as reduced astrogliosis without direct influence on the immune reaction. These neuroprotective effects were associated with the activation of nuclear factor (erythroid-derived 2)-like 2 (Nrf2), a transcription factor involved in the cellular detoxification pathways and the natural antioxidative response [15]. Finally, degenerative processes in autoimmune demyelination share several features with primary neurodegenerative diseases like amyotrophic lateral sclerosis (ALS) or Alzheimer's disease (AD). Sarah Kaiser from the Department of Neurology, University Hospital Ulm, presented data on several cerebrospinal fluid (CSF) biomarkers in ALS. The group of Johannes Brettschneider could show that SMI 35 (representing heavy neurofilaments) was increased in rapidly progressing ALS. In contrast, CSF levels of soluble amyloid precursor protein (sAPP) did not show differences to controls, but negatively correlated with disease duration. In conclusion, the NFH/sAPP ratio may represent a new biomarker for ALS progression and may also be of interest in diseases like MS.

ased in rapidly progressing ALS. In contrast, CSF levels of soluble amyloid precursor protein (sAPP) did not show differences to controls, but negatively correlated with disease duration. In conclusion, the NFH/sAPP ratio may represent a new biomarker for ALS progression and may also be of interest in diseases like MS. The intimate mechanistic relationship between neurodegenerative and neuroinflammatory disease processes is further highlighted by the immune reaction in AD which particularly involves microglial activation. Here, Marius Krauthausen from the group of Marcus Mueller, Bonn, gave an update on their studies on chemokines in the transgenic APP/PS1 model of AD. They investigated the role of CXCR3 in a genetic approach by crossing APP/PS1 mice with CXCR3-deficient mice. As compared to APP/PS1 transgenic controls, these mice display a reduced plaque burden and reduced Aβ protein load associated with a reduced activation and accumulation of periplaque microglia. These data argue for a role of chemokines in plaque formation which may be critically modulated by the function of microglia.

As compared to APP/PS1 transgenic controls, these mice display a reduced plaque burden and reduced Aβ protein load associated with a reduced activation and accumulation of periplaque microglia. These data argue for a role of chemokines in plaque formation which may be critically modulated by the function of microglia. So far, therapeutic options in neurodegenerative diseases are limited. Recently, the application of small inhibitory RNA (siRNA) came into the focus of interest. Anderson de Andrade and Xu Hong, both from Guenter Hoeglinger's group in Marburg, employed siRNA as new therapeutic approach in tauopathies and synucleinopathies. In vitro models using adenoviral overexpression of alpha synuclein in dopaminergic neurons and the in vivo model of P301S tau transgenic mice will allow for refined protocols of siRNA application and thus testing the effect of synuclein or tau knock-down as new therapeutic approaches in neurodegenerative diseases.

ies. In vitro models using adenoviral overexpression of alpha synuclein in dopaminergic neurons and the in vivo model of P301S tau transgenic mice will allow for refined protocols of siRNA application and thus testing the effect of synuclein or tau knock-down as new therapeutic approaches in neurodegenerative diseases. Contributions on stroke and vascular pathology In the stroke field, the role of the immune system was discussed. First studies show that specific inhibition of sphingolipid signaling or inhibition of adhesion molecules can be beneficial. Waltraud Pfeilschifter from Frankfurt presented data that treatment with FTY720, a functional sphingosin 1 phosphate receptor 1 antagonist which blocks the egress of lymphocytes from the lymph node, reduced ischemic damage in the middle cerebral artery occlusion (MCAO) model. It also reduced the activation of the immune system and apoptotic cell death. Arthur Liesz from Heidelberg demonstrated the potential of interfering with the migration of leucocytes across the blood brain barrier by inhibition of VCAM through antibody or siRNA[16]. Especially, the gene silencing resulted in a better outcome in the MCAO model. Furthermore, Friederike Vollmar from the group of Christoph Kleinschnitz in Wuerzburg highlighted novel findings on the role of the proinflammatory kallikrein-kinin-system (KKS) in the pathophysiology of acute ischemic stroke. As previously shown, depletion or pharmacological blockade of the bradykinin receptor B1 (B1R), but not B2R, attenuated postischemic inflammation and blood-brain-barrier damage both after transient middle cerebral artery occlusion or traumatic brain injury in mice [17,18]. In a follow-up project, the group currently investigates whether additional molecules located upstream of the kinin receptors such as kininogen or plasma kallikrein are likewise involved in stroke-induced inflammation and neuronal damage. Preliminary data obtained in kininogen(kng)-deficient mice were indicative for a potential role of KNG for thrombus formation and edema formation in the ischemic brain.

ocated upstream of the kinin receptors such as kininogen or plasma kallikrein are likewise involved in stroke-induced inflammation and neuronal damage. Preliminary data obtained in kininogen(kng)-deficient mice were indicative for a potential role of KNG for thrombus formation and edema formation in the ischemic brain. However, the immune regulation in stroke is quite complex and involves different subclasses of immune cells. Mathias Gelderblom from the group of Tim Magnus showed that three days following stroke γδ T cells emerged in the ischemic hemisphere. The majority of γδ T cells were located in direct proximity to the infarct core. 40% of these atypical T cells produced IL 17A and seemed to have a role in recruiting neutrophils to the area of destruction. The complexity of the immunologic reperfusion response after stroke and possible pitfalls in immune cell depletion approaches as a potential therapeutic strategy were further underscored by the observation of Michael Gliem from Sebastian Jander's group showing that depletion of macrophages with clodronate resulted in an increased bleeding rate after MCAO.

erfusion response after stroke and possible pitfalls in immune cell depletion approaches as a potential therapeutic strategy were further underscored by the observation of Michael Gliem from Sebastian Jander's group showing that depletion of macrophages with clodronate resulted in an increased bleeding rate after MCAO. The peptide hormone Ghrelin is known as the ligand of the growth hormone secretagogue receptor. Ghrelin crosses the blood-brain barrier and binds to hippocampal neurons thereby promoting dendritic spine synapse formation and proliferation of progenitor cells. Kai Diederich together with Jens Minnerup from Münster demonstrated that Ghrelin treatment improves functional recovery after photothrombotic stroke in rats probably by enhancing the generation of newborn hippocampal neurons. Felix Schlachetzki from Regensburg reviewed basic mechanisms of blood-brain barrier (BBB) damage following brain ischemia/reperfusion injury which is associated with intracerebral hemorrhage and edema formation. He pointed out that in experimental stroke BBB permeability is bi-phasic for certain contrast agents (para-endothelial efflux) yet vasogenic edema is a monophasic event (trans-endothelial efflux) as shown by serial post-contrast MRI and T2-relaxometry. However, the bi-phasic BBB response may be linked to both deleterious and regenerative effects at the neurovasular unit [19].

y is bi-phasic for certain contrast agents (para-endothelial efflux) yet vasogenic edema is a monophasic event (trans-endothelial efflux) as shown by serial post-contrast MRI and T2-relaxometry. However, the bi-phasic BBB response may be linked to both deleterious and regenerative effects at the neurovasular unit [19]. In conclusion, by bringing together researchers in the fields of neuroimmunology, neurodegeneration, and neurovascular diseases, this meeting has again been a valuable platform to discuss pathogenic cascades common to these different disorders. Access of immune cells (innate or adaptive) to different body compartments and in particular to the CNS are clearly common themes in a variety of neurological diseases. It is our hope that the NEUROWIND meeting will teach us how it might be possible to advance the understanding of pathogenic processes in neurological disorders by exchanging concepts and tools between various CNS disease models. Competing interests The authors declare that they have no competing interests. Authors' contributions TM, RL, SGM, CK, and TK wrote the paper. All authors read and approved the final manuscript. Acknowledgements The NEUROWIND e.V. scientific meeting was kindly supported by Merck Serono GmbH, Darmstadt, Germany (unrestricted grant to NEUROWIND e.V.). We thank Ms. Anke Bauer, Würzburg, and Patrick Meuth, Münster, for editing the manuscript. This publication was funded by the German Research Foundation (DFG) in the funding programme Open Access Publishing.

ic meeting was kindly supported by Merck Serono GmbH, Darmstadt, Germany (unrestricted grant to NEUROWIND e.V.). We thank Ms. Anke Bauer, Würzburg, and Patrick Meuth, Münster, for editing the manuscript. This publication was funded by the German Research Foundation (DFG) in the funding programme Open Access Publishing. List of speakers at the second scientific meeting of NEUROWIND e.V. (in alphabetical order) Anderson de Andrade, Dept. of Neurology, University of Marburg, Germany Stefan Bittner, Dept. of Neurology, University of Münster, Germany Kai Diederich, Dept. of Neurology, University of Münster, Germany Ulrich Dirnagl, Dept. of Neurology and Experimental Neurology and Center for Stroke Research, Berlin Charité University Medicine, Germany Mathias Gelderblom, Center for Molecular Neurobiology, Hamburg, Germany Michael Gliem, Dept. of Neurology, University of Düsseldorf, Germany Karin Hagemeier, Dept. of Neurology, University of Münster, Germany Dirk Hermann, Dept. of Neurology, University of Essen, Germany Sarah Kaiser, Dept. of Neurology, University of Ulm, Germany Marius Krauthausen, Dept. of Neurology, University of Bonn, Germany De-Hyung Lee, Dept. of Neurology, University of Bochum, Germany Arthur Liesz, Dept. of Neurology, University of Heidelberg, Germany Miguel Maurer, Institute for Experimental Neurology, University of Zürich, Switzerland Philipp Mergenthaler, Charité University Medicine, Berlin, Germany Franziska Petermann, Dept. of Neurology, Technical University of Munich, Germany Waltraud Pfeilschifter, Dept. of Neurology, University of Frankfurt, Germany

Arthur Liesz, Dept. of Neurology, University of Heidelberg, Germany Miguel Maurer, Institute for Experimental Neurology, University of Zürich, Switzerland Philipp Mergenthaler, Charité University Medicine, Berlin, Germany Franziska Petermann, Dept. of Neurology, Technical University of Munich, Germany Waltraud Pfeilschifter, Dept. of Neurology, University of Frankfurt, Germany Melanie Piedavent, Center for Molecular Neurobiology, Hamburg, Germany Franziska Scheibe, Dept. of Experimental Neurology, Berlin Charité University Medicine, Germany Felix Schlachetzki, Dept. of Neurology, University Regensburg, Germany Christian Schläger, Dept. of Neuroimmunology, Institute for MS Research, Göttingen, Germany Thomas Skripuletz, Hannover Medical School, Germany Volker Siffrin, Dept. of Neurology, University of Mainz, Germany Friederike Vollmar, Dept. of Neurology, University of Würzburg, Germany Claudia Wrzos, Dept. of Neurology, University of Göttingen, Germany Hong Xu, Marburg, Dept. of Neurology, University of Marburg, Germany

Introduction Interest in sex differences during acute stroke is an area of growing interest. A consistent finding in rodent models of cerebral ischemia is that young females have smaller infarct sizes and better outcomes than young male rodents [1]. This female protection is lost after ovariectomy. However, the sex difference in stroke is only present when the brain is reperfused; in permanent occlusion the sex difference vanishes [2]. Moreover, in older rodents, the sex difference seen in younger animals is lost [3]. Reproductively senescent older female and male mice have similar infarct sizes after 2 hours of ischemia and 22 hours of reperfusion [4]. The effect of sex on stroke outcome may also be hormone independent [3]. Recent studies suggest the existence of sex-divergent cell death pathways operating during cerebral ischemia [5]. The neuronal nitric oxide (NO)/Poly ADP ribose (PARP) pathways appear to only mediate cell death during cerebral ischemia in male rodents [5]. These sexually divergent pathways may influence how females and males respond to acute stroke treatments. For example, PARP inhibitors, and inhibitors of neuronal NOS are reported to be only neuroprotective in male mice [5,6]. This concern over sex-related effects has resulted in recommendations from the Stroke Academic Industry Roundtable to include female animals and older animals in pre-clinical testing [7].

treatments. For example, PARP inhibitors, and inhibitors of neuronal NOS are reported to be only neuroprotective in male mice [5,6]. This concern over sex-related effects has resulted in recommendations from the Stroke Academic Industry Roundtable to include female animals and older animals in pre-clinical testing [7]. The choice of experimental stroke model is also important. While the suture occlusion model is often used for both reperfusion and permanent ischemic models, a suture is an unnatural occlusion mechanism and reperfusion in human stroke is seldom achieved as abruptly as removal of the suture in an animal. This abrupt reperfusion may modify the cellular consequences of the ischemic process [8,9]. An embolic clot model better models the human clinical condition [10,11]. Moreover, mechanical and thrombolytic reperfusions have different profiles and time courses of cerebral blood flow (CBF) and barrier damage [12]. It is also important to test adult and older rodents of both sexes in embolic clot occlusion models where mostly young male rodents have been used to date.

condition [10,11]. Moreover, mechanical and thrombolytic reperfusions have different profiles and time courses of cerebral blood flow (CBF) and barrier damage [12]. It is also important to test adult and older rodents of both sexes in embolic clot occlusion models where mostly young male rodents have been used to date. Minocycline, a tetracycline derivative, is a promising neuroprotective drug that has reduced infarct size and improved functional outcomes in multiple experimental models [13-19]. Recent evidence suggests that it is also promising in clinical trials [20,21]. Minocycline inhibits PARP-1 at nanomolar concentrations, but has multiple mechanisms of action including inhibition of MMP-9 [14,15,22,23]. However, in a suture occlusion model, minocycline reduced infarct size in male mice, but not in recently ovariectomized (OVX) female mice [5,24,25]. In this study we used a new thromboembolic model with spontaneous reperfusion with a humanized clot. Our aims were to determine if minocycline was neuroprotective in an embolic clot model in mice of both sexes and in aged mice. We also utilized more than 12 weeks interval between ovariectomy in adult females and embolic stroke to mimic estrogen homeostasis post-menopause in humans. A secondary aim was to determine if there was a sex-specific change in MMP-9 levels.

uroprotective in an embolic clot model in mice of both sexes and in aged mice. We also utilized more than 12 weeks interval between ovariectomy in adult females and embolic stroke to mimic estrogen homeostasis post-menopause in humans. A secondary aim was to determine if there was a sex-specific change in MMP-9 levels. Materials and methods Animals All the experimental procedures have been approved by the Institutional Animal Care and Use Committee (IACUC) of Georgia Health Sciences University (GHSU) in accordance with NIH Guide for the Care and Use of Laboratory Animals. Wild type C57BL/6J male and female mice were purchased from the Jackson Laboratory (Bar Harbor, Maine) and housed in the GHSU Animal Facility approved by the American Association for Accreditation of Laboratory Animal Care. Female mice were ovariectomized at 11-12 weeks of age in the Jackson Laboratory and were about 12 weeks post ovariectomy prior to stroke. The mean ages of stroked adult animals were: 24.0 ± 4.6 weeks for adult males, 22.9 ± 3.3 weeks for adult females, and 23.9 ± 2.7 weeks for OVX females. The aged C57BL/6 animals (18.1 ± 0.8 month males and 16.0 ± 1.1 month females) were from GHSU in-house breeding.

were about 12 weeks post ovariectomy prior to stroke. The mean ages of stroked adult animals were: 24.0 ± 4.6 weeks for adult males, 22.9 ± 3.3 weeks for adult females, and 23.9 ± 2.7 weeks for OVX females. The aged C57BL/6 animals (18.1 ± 0.8 month males and 16.0 ± 1.1 month females) were from GHSU in-house breeding. Estrus cycle analysis in mice Five weeks after ovariectomy female mice (n = 7) were tested for estrous cyclicity to confirm loss of estrogenic effect and simulation of postmenopausal stage. The vaginal smears were compared to cycling adult females (n = 7) and aged females (n = 5). All vaginal smears were obtained daily between 9:00 am and 1:00 pm for 8 weeks and stained with 2% Giemsa solution (Sigma) as described [26]. Estrogen levels Plasma samples (50 μl) from all post-stroke survived females and eight randomly selected adult male mice were used to measure estrogen by enzyme immunoassay (Cayman Chemical, Ann Arbor, MI).

Estrus cycle analysis in mice Five weeks after ovariectomy female mice (n = 7) were tested for estrous cyclicity to confirm loss of estrogenic effect and simulation of postmenopausal stage. The vaginal smears were compared to cycling adult females (n = 7) and aged females (n = 5). All vaginal smears were obtained daily between 9:00 am and 1:00 pm for 8 weeks and stained with 2% Giemsa solution (Sigma) as described [26]. Estrogen levels Plasma samples (50 μl) from all post-stroke survived females and eight randomly selected adult male mice were used to measure estrogen by enzyme immunoassay (Cayman Chemical, Ann Arbor, MI). Preparation of emboli The method of clot preparation was adapted from earlier reports [10,27] and modified to increase the strength of uniformity of the fibrin rich core and stability of occlusion. Briefly, mouse arterial blood was supplemented with human fibrinogen (2 mg/mL), and immediately clotted in PE-50 tubing for 6 hours at room temperature followed by storage at 4°C. Before use, the clot (~5 cm) was transferred into a modified PE-10 tube filled with sterile saline and retracted. The clot was then transferred to a Petri dish containing phosphate-buffered saline and left for further retraction at room temperature for 4 hours. A single 9.0 ± 0.5 mm long clot was transferred to a modified PE-10 catheter for embolization.

nsferred into a modified PE-10 tube filled with sterile saline and retracted. The clot was then transferred to a Petri dish containing phosphate-buffered saline and left for further retraction at room temperature for 4 hours. A single 9.0 ± 0.5 mm long clot was transferred to a modified PE-10 catheter for embolization. Thromboembolic stroke model Mice were anesthetized with 3.5% isofluorane and maintained with 2.0% during the surgery. Body temperature was maintained at 37°C by a thermo-regulated surgery pad. By a midline incision on the ventral side of the neck, the right common carotid artery, the right external carotid artery (ECA), and the right internal carotid artery (ICA) were assessed [10]. A modified PE-10 catheter containing a clot was introduced into the ECA lumen through a small hole, advanced into the ICA, and the clot was gently injected with 100 μL of the sterile phosphate-saline buffer (PBS). After thromboembolization the catheter was removed immediately. To identify the location of an embolus after injection, the fibrin-rich clot was labeled by Evans blue before injection [10]. In the sham group, an equal volume of PBS without clot was delivered. Occlusion was confirmed by ≥70% drop in cerebral blood flow (CBF) compared to the pre-ischemic value. Animals that showed sustained occlusion were included. The success rate of thromboembolic MCA occlusion was 95% (208 successes from 219 total stroked animals) based on changes in CBF. Animals were randomized immediately after clot injection and treated with either phosphate saline (PBS, vehicle) or minocycline (Sigma; 6 mg/kg) via bolus IV injection to tail vein (0.1 mL/10 g body weight). Sham-operated mice served as controls.

% (208 successes from 219 total stroked animals) based on changes in CBF. Animals were randomized immediately after clot injection and treated with either phosphate saline (PBS, vehicle) or minocycline (Sigma; 6 mg/kg) via bolus IV injection to tail vein (0.1 mL/10 g body weight). Sham-operated mice served as controls. Regional cortical laser-doppler flowmetry Cortical laser-Doppler flowmetry ([LDF], Perimed. Inc.) was performed 30 min and 3 min before occlusion to record a consistent basal level of peripheral blood flow in the middle cerebral artery region, and also recorded during occlusion [28]. For this purpose, shallow indentation was made in the parietal skull (A - P 2 mm, and lateral 3 mm with respect to bregma) with a low-speed drill for placement of the LDF probe holder (PH07-6, Perimed. Inc.). The LDF signal was recorded semi-continuously and averaged over 10-minute intervals for each time point. Neurological assessment Neurologic deficits in the animals were assessed at 24 hr post stroke by a 5-point scale scoring: 0, no deficit; 1, forelimb flexion deficit on contralateral side; 2, decreased resistance to lateral push and torso turning to the ipsilateral side when held by tail; 3, very significant circling to affected side and reduced capability to bear weight on the affected side; 4, rarely moves spontaneously and prefer to stay in rest.

cit; 1, forelimb flexion deficit on contralateral side; 2, decreased resistance to lateral push and torso turning to the ipsilateral side when held by tail; 3, very significant circling to affected side and reduced capability to bear weight on the affected side; 4, rarely moves spontaneously and prefer to stay in rest. Infarct analysis Mice were euthanized at 24 hours after stroke for injury assessment. Brains were perfused with ice cold 0.01 M phosphate-buffered saline (PBS), cut into 1-mm coronal slices. Every other slice was stained with 2% 2,3,5-triphenyltetrazolium chloride (TTC, Sigma) for 30 minutes at 37°C, then fixed with 10% formalin in PBS. The images were digitalized and the infarct volume was analyzed by software SPOT Advanced (Sterling Heights, MI) as previously described [14,28,29]. The infarct volumes were quantified as both direct volume (in mm3) and indirect volume (percent volume of the total ischemic hemisphere). The measurement of infarct size was made by an investigator blinded to treatment group.

analyzed by software SPOT Advanced (Sterling Heights, MI) as previously described [14,28,29]. The infarct volumes were quantified as both direct volume (in mm3) and indirect volume (percent volume of the total ischemic hemisphere). The measurement of infarct size was made by an investigator blinded to treatment group. Immunoblotting Six hours after the stroke onset, mice were euthanized, and brains were perfused with cold phosphate-buffered saline and extracted. Ischemic hemisphere tissue was homogenized and lysed in complete Lysis-M EDTA-free buffer (Roche Diagnostics, Indianapolis, IN). The amount of total protein was quantified using the EZQ® Protein Quantitation Kit (Invitrogen). Samples (30 micrograms of total protein) were subjected to SDS-PAGE using 10% NuPAGE® Novex® Bis-Tris gels (Invitrogen) and transferred to 0.2 μm PVDF membranes (Millipore, Billerica, MA). The membrane was blocked by for non-specific binding (5% BSA solution), and incubated with polyclonal anti-MMP-9 antibody (G6571, Cell Signaling Technology, Danvers, MA) at 4°C overnight, followed by HRP-conjugated donkey anti-rabbit IgG antibody (Jackson ImmunoResearch, West Grove, PA). Membranes were re-probed with mouse monoclonal anti-β-actin antibody (Sigma-Aldrich Co.) as a loading control. Proteins were visualized with the ECL detection system (Pierce, Thermo Fisher Scientific) on autoradiography film (Denville Scientific, Metuchen, NJ). The image was scanned and processed for densitometric measurement in Image-J software. The analysis of MMP-9 antigen was made by an investigator blinded to treatment group.

teins were visualized with the ECL detection system (Pierce, Thermo Fisher Scientific) on autoradiography film (Denville Scientific, Metuchen, NJ). The image was scanned and processed for densitometric measurement in Image-J software. The analysis of MMP-9 antigen was made by an investigator blinded to treatment group. Gelatin zymography Basal plasma MMP activity, and at 6 hours after stroke, was detected using gelatin zymography [14,23,28,29] in duplicate for each sample. The ischemic hemisphere tissue was homogenized in 0.05 mol/L Tris buffer (pH 7.4) containing 0.1 mol/L NaCl and 0.17 ng/mL PMSF. Citrated blood samples were immediately centrifuged at 4000 rpm for 20 min at 4°C and total plasma protein was determined with the BCA assay (Pierce). The total lysed basal protein (100 micrograms) or plasma protein (30 micrograms) was loaded and separated by a 10% Tris-glycine gel with 0.1% porcine gelatin (Sigma). The gels were washed with renaturing buffer and incubated with developing buffer (BioRad Labs.) at 37°C for 20 hours. Finally, the gels were stained with Coomassie blue R-250 followed by appropriate destaining. The gelatinolytic activity of the samples was assessed by densitometric analysis (Gel-Pro v 3.1, Media Cybernetics, Carlsbad, CA) of the bands as a relative comparison to a standard band of recombinant enzyme. To minimize inter-gel variability, all gels had a control lane loaded with 0.5 ng recombinant enzyme, which was used as a standard optical density and enzyme amount (in ng). The density of the sample bands were expressed as maximal optical density relative to the standard band. The analysis of MMP-9 activity was made by an investigator blinded to treatment group.

ad a control lane loaded with 0.5 ng recombinant enzyme, which was used as a standard optical density and enzyme amount (in ng). The density of the sample bands were expressed as maximal optical density relative to the standard band. The analysis of MMP-9 activity was made by an investigator blinded to treatment group. Measurement of cerebral perfusion and MRI procedureare described in the Additional Methods. Statistical analysis All data are expressed as mean ± SD. Statistical analyses were performed using SAS® 9.2 (SAS Institute, Inc., Cary, NC). Infarction and neurologic score were analyzed for males using a 2 Age (adult vs. aged) × 2 TRT (saline vs. minocycline) ANOVA and for females using a 3 Group (adult, aged, OVX) × 2 TRT (saline vs. minocycline) ANOVA. Adult males and females were compared using a 2 Sex (male vs. female) × 2 TRT (saline vs. minocycline) ANOVA. MMP-9 antigen and activity levels were analyzed using a 2 Sex (male vs. OVX) × 3 TRT (sham, stroke+vehicle, stroke+minocycline) ANOVA. In all analyses, interactions were tested for possible differential effects of minocycline treatment on age, sex or group. Tukey's multiple comparison tests were used to compare means for significant main effects. The effect of minocycline on mortality at 24 hours was analyzed in adult and aged mice using Fisher's Exact test. Group numbers are shown in parentheses. Statistical significance was determined at P < 0.05.

on age, sex or group. Tukey's multiple comparison tests were used to compare means for significant main effects. The effect of minocycline on mortality at 24 hours was analyzed in adult and aged mice using Fisher's Exact test. Group numbers are shown in parentheses. Statistical significance was determined at P < 0.05. Results Thromboembolic stroke model The model was initially optimized using C57BL/6 wild type male mice (24 ± 4 weeks old, 25 - 30 g) by injecting a fibrin rich clot formed ex-vivo into the right MCA. Figure 1A represents the brain image with the clot occluding the ipsilateral MCA origin. The embolization led to consistent reduction of CBF to 18.1 ± 4.7% of baseline that persisted (25.4 ± 4.6%) for 6 hours (Figure 1B and Additional file 1, Figure S1). There was a slow spontaneous CBF restoration by 24 hours that reached 65 ± 5% from the baseline as determined by cortical laser-Doppler flowmetry (Figure 1B). Cerebral perfusion imaging with the PeriScan system yielded similar results (Additional file 2, Figure S2). Thromboembolic stroke resulted in reproducible infarct (115 ± 22 mm3, n = 14) which was confirmed by T2-weighted MRI (Figure 1C and Additional file 3, Figure S3) or coronal brain staining with 2,3,5-triphenyltetrazolium chloride (Figure 1D and Additional file 3, Figure S3). In this model adult males treated with saline had reliable injury and low mortality rate at 24 hours (3 out of 17 animals, Table 1). The model was applied to study the gender difference in neuroprotection with minocycline using various subgroups of male (Figure 2) and female (Figure 3) mice.

itional file 3, Figure S3). In this model adult males treated with saline had reliable injury and low mortality rate at 24 hours (3 out of 17 animals, Table 1). The model was applied to study the gender difference in neuroprotection with minocycline using various subgroups of male (Figure 2) and female (Figure 3) mice. Figure 1 Thromboembolic stroke model in C57BL/6 adult male mice. A) The clot is inserted at the MCA origin. Clots were visualized with Evan's Blue dye (arrowhead). B) Regional cerebral blood flow (CBF) over time as measured with Laser Doppler flowmetry (means ± SD, n = 7). The CBF declines to 18 ± 4% of baseline following embolization, which was associated with a reproducible sized infarct of the MCA territory (115 ± 22 mm3) at 24 post stroke hours. C, D) Representative infarct area at 24 hours as determined by T2 diffusion-weighted MRI (C) and with 2,3,5-triphenyltetrazolium chloride staining of the coronal brain sections (D). Additional representative images of thromboembolic stroke model are shown in the Additional file 1 (Figure S1), Additional file 2 (Figure S2), Additional file 3 (Figure S3), Additional file 9 (Additional Figure Legends) and Additional file 10 (Additional Methods). Table 1 Mortality rates at 24 hours after stroke Group Treatment Total in studies, n Dead, n (mortality rate) Adult males Vehicle 17 3 (17%) Minocycline 13 1 (8%) Aged males Vehicle 17 10 (59%) Minocycline 9 2 (22%) Adult females Vehicle 15 2 (13%) Minocycline 15 3 (20%) Aged females Vehicle 23 11 (48%) Minocycline 15 4 (27%)

Figure 1 Thromboembolic stroke model in C57BL/6 adult male mice. A) The clot is inserted at the MCA origin. Clots were visualized with Evan's Blue dye (arrowhead). B) Regional cerebral blood flow (CBF) over time as measured with Laser Doppler flowmetry (means ± SD, n = 7). The CBF declines to 18 ± 4% of baseline following embolization, which was associated with a reproducible sized infarct of the MCA territory (115 ± 22 mm3) at 24 post stroke hours. C, D) Representative infarct area at 24 hours as determined by T2 diffusion-weighted MRI (C) and with 2,3,5-triphenyltetrazolium chloride staining of the coronal brain sections (D). Additional representative images of thromboembolic stroke model are shown in the Additional file 1 (Figure S1), Additional file 2 (Figure S2), Additional file 3 (Figure S3), Additional file 9 (Additional Figure Legends) and Additional file 10 (Additional Methods). Table 1 Mortality rates at 24 hours after stroke Group Treatment Total in studies, n Dead, n (mortality rate) Adult males Vehicle 17 3 (17%) Minocycline 13 1 (8%) Aged males Vehicle 17 10 (59%) Minocycline 9 2 (22%) Adult females Vehicle 15 2 (13%) Minocycline 15 3 (20%) Aged females Vehicle 23 11 (48%) Minocycline 15 4 (27%) OVX females Vehicle 26 13 (50%) Minocycline 15 1 (7%) Figure 2 Representative effect of minocycline in adult (A) and aged (B) males. Coronal TTC sections, outlining the infarct area in representative subjects treated by the phosphate-saline buffer (vehicle) or minocycline at 24 hours, are showed in the left panels. The corresponding change of regional CBF during the first 6 hours after stroke is showed on the right panel. The value of CBF is presented as means ± SD.

ons, outlining the infarct area in representative subjects treated by the phosphate-saline buffer (vehicle) or minocycline at 24 hours, are showed in the left panels. The corresponding change of regional CBF during the first 6 hours after stroke is showed on the right panel. The value of CBF is presented as means ± SD. Figure 3 Representative effect of minocycline in adult (A), retired (B), and OVX (C) females. Coronal TTC section, outlining the infarct area in a representative subjects treated by vehicle or minocycline at 24 hours, are showed on the left panel. The corresponding change of regional CBF during the first 6 hours after stroke is showed on right panel. The value of CBF is presented as means ± SD.

OVX (C) females. Coronal TTC section, outlining the infarct area in a representative subjects treated by vehicle or minocycline at 24 hours, are showed on the left panel. The corresponding change of regional CBF during the first 6 hours after stroke is showed on right panel. The value of CBF is presented as means ± SD. Acute minocycline treatment is neuroprotective for male mice We tested the neuroprotective efficiency of minocycline in two different groups of male mice: adult males (24.0 ± 4.6 weeks old) and aged males (18.1 ± 0.8 months old). The embolization lead to significant and stable drop of CBF in the MCA territory (18-21% of pre-ischemic values) as shown in Figures 2A and 2B. However, in adult males the occlusion was more stable (24% ± 4% of pre-ischemic values) for 6 hours than in aged group. We observed a modest spontaneous reperfusion in the aged males (41% ± 9% of pre-ischemic values for 6 hours after occlusion) that may resulted from the increased endogenous tPA activity in elderly [30,31]. Animals were randomized immediately after clot injection and treated with ether PBS (vehicle) or minocycline (6 mg/kg). Treatment with minocycline did not significantly affect regional CBF in aged males (Figure 2A), but induced modest spontaneous reperfusion in the adults (Figure 2B) at 6 hours (48% ± 8% of pre-ischemic values). Aged males had increased mortality at 24 hours post stroke that was markedly attenuated after minocycline treatment (Table 1). Minocycline significantly reduced the infarct volumes (P < 0.0001) and also improved functional outcomes (p < 0.0001) in surviving males in both adult and aged groups. Infarct volumes for adult males were reduced to 18.8% ± 14.5% in minocycline treated mice versus 43.1% ± 10.6% for the saline group, and treatment of aged group with minocycline decreased the infarct volumes to 16.8% ± 8.9% versus 34.6% ± 9.9% (Figure 4A). Neurologic scores at 24 hours post stroke in adult males were 2.0 ± 0.8 for the minocycline treated group versus 3.2 ± 0.6 for the saline group, and 2.0 ± 0.8 versus 3.6 ± 0.5 in the corresponding aged males (Figure 5A). Representative coronal sections of 24 hours post stroke brain are shown in Figures 2A and 2B.

4A). Neurologic scores at 24 hours post stroke in adult males were 2.0 ± 0.8 for the minocycline treated group versus 3.2 ± 0.6 for the saline group, and 2.0 ± 0.8 versus 3.6 ± 0.5 in the corresponding aged males (Figure 5A). Representative coronal sections of 24 hours post stroke brain are shown in Figures 2A and 2B. Figure 4 Minocycline reduces brain tissue injury in males (A) and females (B). The infarct size of ischemic MCA territory was estimated as the percent volume of the total ischemic hemisphere. Regardless of treatment adult females had significantly smaller infarct volumes (both P < 0.0001, ***) than aged and OVX females who were not different than each other (P = 0.77). All data expressed as means ± SD. Figure 5 Minocycline improves neurologic scores in males (A) and females (A). Neurologic evaluation was performed for surviving animals at 24 hours. Regardless of treatment adult females had significantly improved neurologic scores (both P < 0.0001, ***) than aged and OVX females who were not different than each other (P = 0.42). All data expressed as means ± SD.

es in males (A) and females (A). Neurologic evaluation was performed for surviving animals at 24 hours. Regardless of treatment adult females had significantly improved neurologic scores (both P < 0.0001, ***) than aged and OVX females who were not different than each other (P = 0.42). All data expressed as means ± SD. Acute minocycline treatment is neuroprotective for female mice To analyze the effect of minocycline on the ischemic injury in female animals, we studied 3 groups of female mice: adult females (22.9 ± 3.3 weeks old), aged females (16.0 ± 1.1 months old), and ovariectomized (OVX) females (23.9 ± 2.7 weeks old). The age of adult and old females was selected based on reported time for maximal and stable cycle frequency (7-10 months) or onset of acyclicity (13-16 months) in C57BL/6 strain [26]. Consistent with the early report [26], we observed the cycle regularity in adult females (4-5 days) and significantly lengthened cycles in old females (7 - 11 days) with predomination of diestrus stage. Unlike other studies [5,24,25], ovariectomy was performed in early age at least 10 weeks before stroke onset. The estrous cyclicity test in OVX mice has confirmed their complete acyclicity and loss of estrogenic effect.

d significantly lengthened cycles in old females (7 - 11 days) with predomination of diestrus stage. Unlike other studies [5,24,25], ovariectomy was performed in early age at least 10 weeks before stroke onset. The estrous cyclicity test in OVX mice has confirmed their complete acyclicity and loss of estrogenic effect. Consistent with cyclicity data the estrogen levels, determined at experimental endpoint, were 31.5 ± 51.0 pg/ml for adult females, 13.5 ± 9.5 pg/ml for aged females, and 6.6 ± 5.3 pg/ml for OVX females (Additional file 4, Figure S4). Estrogen levels were equivalent in OVX female mice and adult male mice (6.2 ± 0.7 pg/ml). All female mice were used regardless of cycle stage that should reflect the clinical scenario including both pre- and post menopause females in different estrogen cycle stages.

pg/ml for OVX females (Additional file 4, Figure S4). Estrogen levels were equivalent in OVX female mice and adult male mice (6.2 ± 0.7 pg/ml). All female mice were used regardless of cycle stage that should reflect the clinical scenario including both pre- and post menopause females in different estrogen cycle stages. The degree of CBF reduction after embolization was similar among the groups (Figure 3). The degree of spontaneous CBF restoration was also similar among the female groups (39% - 51%), but was higher than in adult males due to possible enhanced endogenous fibrinolytic activity in females [32,33]. As expected, adult females were significantly more resistant to cerebral ischemic injury than males, but this advantage was abolished by aging or lack of estrogen in OVX females (P < 0.001). Mortality rates in the 24 hours studies were significantly higher in the aged and OVX groups compared to adults (P < 0.001, Table). Unlike aged animals, the mortality in OVX females was associated with intracranial post-stroke bleeding (Additional file 5, Figure S5). Minocycline treatment provided significant neuroprotection in all female groups. The mortality rates were markedly decreased in aged and significantly reduced in OVX females (P = 0.0061) attenuating bleeding. Moreover, the infarct volumes were reduced (Figure 4B; P < 0.0001) and neurologic scores were also improved (Figure 5B; P < 0.0001) in all minocycline treated females as compared to their corresponding vehicle treated controls.

endent studies we detected the maximum MMP-9 level in the brain at 6 hours post stroke (Additional file 6, Figure S6). However, in plasma samples we detected low, or no, MMP-9 after ischemia compare to sham controls. This may be associated with using citrate to prevent blood clotting for plasma harvesting in our study. To estimate the gender effect, we used adult males and OVX females subjected to 6 hours stroke and randomized to sham-control, vehicle and minocycline treated (ns = 6-8) groups (Table 2). Figure 6A shows that stroke up-regulated brain level of MMP-9 protein and minocycline treatment reduced its expression in both genders (P < 0.0001). The vehicle-treated mice had significantly higher levels of MMP-9 protein than sham operated animals (P = 0.0007) and minocycline treated animals (P < 0.0001). No statistical difference between sham and minocycline groups were found (P = 0.77). Males had significantly higher level of MMP-9 expression than the OVX females (P = 0.0095). In the vehicle groups brain MMP-9 activity, as determined by zymography was highly variable in both genders (Figure 6B and Additional file 7, Figure S7). As shown in Additional file 7 (Figure S7), in a minority of animals acute ischemia did not result in up-regulation of MMP-9, such that no significant differences were found for sex or treatment. Table 2 Mortality rates at 6 hours after stroke Group Treatment Total in studies, n Dead, n Adult males Vehicle 8 1 Minocycline 7 0 Sham 6 0

To estimate the gender effect, we used adult males and OVX females subjected to 6 hours stroke and randomized to sham-control, vehicle and minocycline treated (ns = 6-8) groups (Table 2). Figure 6A shows that stroke up-regulated brain level of MMP-9 protein and minocycline treatment reduced its expression in both genders (P < 0.0001). The vehicle-treated mice had significantly higher levels of MMP-9 protein than sham operated animals (P = 0.0007) and minocycline treated animals (P < 0.0001). No statistical difference between sham and minocycline groups were found (P = 0.77). Males had significantly higher level of MMP-9 expression than the OVX females (P = 0.0095). In the vehicle groups brain MMP-9 activity, as determined by zymography was highly variable in both genders (Figure 6B and Additional file 7, Figure S7). As shown in Additional file 7 (Figure S7), in a minority of animals acute ischemia did not result in up-regulation of MMP-9, such that no significant differences were found for sex or treatment. Table 2 Mortality rates at 6 hours after stroke Group Treatment Total in studies, n Dead, n Adult males Vehicle 8 1 Minocycline 7 0 Sham 6 0 OVX females Vehicle 6 0 Minocycline 6 0 Sham 6 0 Figure 6 Minocycline reduces level of MMP-9 in both genders. A) Densitometric analysis of immunoreactive band intensities and representative Western Blot showing expression of 92 kDa-MMP-9 in ipsilateral hemispheres of adult male and OVX female mice (n = 6-7 animals per group) at 6 hours after thromboembolization. Values are expressed as relative intensity normalized to 42 kDa-β-actin intensity. The saline-treated mice had higher levels of MMP-9 than minocycline treated animals (P < 0.0001, ***). Minocycline and sham groups were not significantly different from each other (P = 0.77). B) Densitometric analysis of brain MMP-9 activity and representative zymography. The brain injury shows increased variability of MMP-9 activity relative to sham, but there are no sign differences between treatment or sex groups. All data expressed as means ± SD.

e not significantly different from each other (P = 0.77). B) Densitometric analysis of brain MMP-9 activity and representative zymography. The brain injury shows increased variability of MMP-9 activity relative to sham, but there are no sign differences between treatment or sex groups. All data expressed as means ± SD. Mortality Relative to saline, minocycline treatment significantly reduced mortality at 24 hours post-ischemia for OVX females (P = 0.006) and for aged mice (54% versus 25%, P = 0.037) (Table 1). There was not a difference in mortality (P = 1.0) for adult male and female mice (14% versus 16%). Collapsing across sex and age mortality was overall reduced in stroked mice with minocycline treatment compared to the non-minocycline controls (16% vs. 40%). No mortality was observed in the 6 hour study (except one of the eight males in vehicle group, Table 2).

Mortality Relative to saline, minocycline treatment significantly reduced mortality at 24 hours post-ischemia for OVX females (P = 0.006) and for aged mice (54% versus 25%, P = 0.037) (Table 1). There was not a difference in mortality (P = 1.0) for adult male and female mice (14% versus 16%). Collapsing across sex and age mortality was overall reduced in stroked mice with minocycline treatment compared to the non-minocycline controls (16% vs. 40%). No mortality was observed in the 6 hour study (except one of the eight males in vehicle group, Table 2). Discussion This study has several novel and important findings. First, we used a thromboembolic clot model rather than an intraluminal suture model (MCAO). Although MCAO, as the occlusion-reperfusion model, is widely utilized, the evolution of infarct within the territory of blood supplied by MCA is not been well explored [3]. Some study suggests that the fast reperfusion (at suture removal) may accelerate infarct development and modify the cellular mechanisms of the ischemic process [3,8,9]. A thromboembolic rodent model better mimics human stroke and has been previously described. Reperfusion is gradually and partially restored spontaneously but not until 6 to 12 hours when the ischemic cascade is already well advanced. This mimics the clinical scenario where the clot persists with slow spontaneous reperfusion and thrombolytic or mechanical reperfusion fails or is not performed. The model prevents the rapid evolution of the penumbra (as one of the limitations in clinical neuroprotective studies) and allows administration of a neuroprotective agent in the ischemic time windows that are suitable to mimic in clinical trials. Moreover, this represents the vast majority of human middle cerebral artery territory strokes (95% of stroke patients) where t-PA and mechanical removal are not given, or are ineffective.

es) and allows administration of a neuroprotective agent in the ischemic time windows that are suitable to mimic in clinical trials. Moreover, this represents the vast majority of human middle cerebral artery territory strokes (95% of stroke patients) where t-PA and mechanical removal are not given, or are ineffective. Second, in this study we applied the novel approach to partially "humanize" the stroke mouse model. By supplementing the clot with human fibrinogen we adjusted the physiological level of fibrinogen in mouse blood clot (1.5 g/L) to the range of human normal value (1.5 - 4 g/L). This resulted in the increased strength and uniformity of the fibrin-rich clot and stabilized the occlusion. The stabilization of occlusion for 6 hours in adult males may be applied for further neuroprotective study beyond the window of thrombolytic therapy. The "humanization" of clot may also partially eliminate the cross species restriction barriers for binding of t-PA to the clot surface. Third, to the best of our knowledge, this is the first study that tested minocycline in an thromboembolic stroke model to investigate gender and age-dependent influences on stroke injury and outcomes. Our novel findings provide evidence that minocycline was effective at reducing infarct size and improving short-term neurological outcome in young male and female mice, OVX female mice and aged male and female mice.

boembolic stroke model to investigate gender and age-dependent influences on stroke injury and outcomes. Our novel findings provide evidence that minocycline was effective at reducing infarct size and improving short-term neurological outcome in young male and female mice, OVX female mice and aged male and female mice. Overall, analyzing all subgroups and reflecting to the clinical situation, minocycline reduced mortality (16% vs. 40%), decreased the infarct size (13.3% ± 1.4% vs. 32.8 ± 1.7%, P < 0.0001, Additional file 8, Figure S8A) and improved neurological outcomes (1.9% ± 0.1% vs. 3.2 ± 0.1%, P < 0.0001, Additional file 8, Figure S8B). Thus, this is the first comprehensive attempt to study minocycline's effect by both sex and age. Using the thromboembolic stroke model we found that minocycline protected not only young adult male but also female and aged (male & female) brains compared to control vehicle treatment.

Overall, analyzing all subgroups and reflecting to the clinical situation, minocycline reduced mortality (16% vs. 40%), decreased the infarct size (13.3% ± 1.4% vs. 32.8 ± 1.7%, P < 0.0001, Additional file 8, Figure S8A) and improved neurological outcomes (1.9% ± 0.1% vs. 3.2 ± 0.1%, P < 0.0001, Additional file 8, Figure S8B). Thus, this is the first comprehensive attempt to study minocycline's effect by both sex and age. Using the thromboembolic stroke model we found that minocycline protected not only young adult male but also female and aged (male & female) brains compared to control vehicle treatment. This study is also novel in the context of the revised preclinical STAIR criteria call for testing of neuroprotective agents in female mice and aged mice. Most experimental stroke studies have been done exclusively in young male animals, although stroke mainly affects the elderly. Only a few studies have used female or aged rats, particularly with an embolic clot model [11,34,35]. Moreover, we could find no comprehensive reports of the thromboembolic clot model in aged or female mice. One reason may be the high mortality in this model in aging animals. In our model we found a mortality of nearly 50% in both aged females and males and a similar high mortality in OVX females. Minocycline significantly reduced this mortality in OVX females and aged mice. Further studies are needed and planned to determine the effect of t-PA in the embolic clot model in younger and older female mice and the expansion of this time window for minocycline treatment. Including animals of both sexes and aged animals in an embolic clot model is warranted for design of future clinical trials.

Further studies are needed and planned to determine the effect of t-PA in the embolic clot model in younger and older female mice and the expansion of this time window for minocycline treatment. Including animals of both sexes and aged animals in an embolic clot model is warranted for design of future clinical trials. It is interesting that in MCAO preclinical model minocycline was neuroprotective in male mice but not in recently OVX females [24]. While minocycline is a potent PARP inhibitor at nanomolar concentrations [5,22], minocycline acts by multiple mechanisms of action [13,14,16,18,19], one of which is MMP-9 inhibition [14,23]. Activation of MMP-9 plays an important role in mediating tissue injury during human ischemic stroke and is associated with ICH after t-PA [36-40]. Suppression of MMP-9 may lead to safer therapeutic outcomes in acute stroke [41]. Although the different mechanisms may be responsible for neuroprotection with minocycline, the difference in stroke models and following ischemic sequences may be the key factors of attenuation or augmentation of the particular mechanism. Here, we demonstrated that brain MMP-9 was up-regulated in an embolic model after ischemia in both male and female mice. Moreover, minocycline reduces MMP-9 expression for both sexes. MMP-9 activity has been shown to be up-regulated in the blood and brain of ischemic rodents, but all studies to date have used male rodents. MMP-9 activity was highly variable following ischemia in this model and others [15,41] when t-PA is not given. While there appeared to be a trend towards less variability of MMP-9 brain activity with minocycline treatment we did not detect a significant overall reduction in activity.