- Email: [email protected]
Loss of NeuN immunoreactivity after cerebral ischemia does not indicate neuronal cell loss: a cautionary note
Brain Research 1015 (2004) 169 – 174 www.elsevier.com/locate/brainres
Research report
Loss of NeuN immunoreactivity after cerebral ischemia does not indicate neuronal cell loss: a cautionary note ¨ zdemir a, Gunfer Gurer a, Turgay Dalkara a,* ¨ nal-Cßevik a, Mu¨nire Kılıncßb, Yasemin Gu¨rsoy-O Isın U a
Institute of Neurological Sciences and Psychiatry, and Faculty of Medicine, Department of Neurology, Hacettepe University, Sıhhiye Ankara 06100, Turkey b Faculty of Medicine, Department of Neurology, Baskent University, Ankara, Turkey Accepted 9 April 2004 Available online
Abstract NeuN immunoreactivity is used as a specific marker for neurons. The number of NeuN-positive cells decreases under pathological conditions. This finding is usually considered as an evidence of neuronal loss. However, decrease in NeuN labeling may also be caused by depletion of the protein or loss of its antigenicity. Hence, we have investigated the morphological features of neurons that lost NeuN immunoreactivity and the NeuN protein levels in mouse brain after cerebral ischemia. The number of NeuN-labeled cells was decreased 6 h after a mild ischemic insult (30 min middle cerebral artery occlusion) in penumbral and core regions. Hematoxylin and eosin (H&E) staining of adjacent sections showed that neurons in the penumbra were not disintegrated but displayed early ischemic changes. The nuclear NeuN staining was dramatically reduced or lost in some neurons. However, Hoechst 33258 staining of the same sections revealed that these nuclei were preserved with an intact membrane. Labeling of neurons that had lost NeuN-positivity with antibodies against caspase-3-p20, which is constitutively not present but emerges in neurons after ischemia, disclosed that these neurons still preserved their integrity. Moreover, Western blots showed that NeuN protein levels were not decreased, suggesting that reduced NeuN antigenicity accounted for loss of immunoreactivity in this mild brain injury model. Supporting this idea, NeuN labeling was partially restored after antigenic retrieval. In conclusion, since NeuN immunoreactivity readily decreases after metabolic perturbations, reduced NeuN labeling should not be taken as an indicator of neuronal loss and, quantitative analysis based on NeuN-positivity should be used cautiously after central nervous system (CNS) injury. D 2004 Elsevier B.V. All rights reserved. Theme: Cellular and molecular biology Topic: Staining, tracing and imaging techniques Keywords: Cerebral ischemia; NeuN immunoreactivity; Caspase-3-p20
1. Introduction In order to differentiate neurons from glial cells in brain sections, specific neuronal markers, such as neuron-specific enolase, neurofilament proteins and calbindin, can be used in immunohistochemical procedures [17]. However, most of these proteins are present in distinct subpopulations of neurons, limiting their application for histological identification of these cells. In order to overcome this obstacle, Mullen et al. [15] developed a monoclonal antibody that recognizes a neuron specific nuclear protein. NeuN, which * Corresponding author. Tel.: +90-312-3051809; fax: +90-3123093451. E-mail address: [email protected] (T. Dalkara). 0006-8993/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2004.04.032
stands for neuron-specific nuclear protein, is expressed in nucleus and cell body of most neuronal cell types in rodents, chicks and humans [15]. It is not expressed in glial cells; neither oligodendrocytes nor astrocytes nor do microglial cells express NeuN. Since its development, NeuN has been successfully used as a neuronal marker in a wide variety of studies in cell cultures as well as in diagnostic histopathology. Thus, NeuN has been established as a universal marker for neurons although a few neuronal cell types such as cerebellar Purkinje cells, retinal photoreceptor cells, mitral cells of the olfactory bulbs, and sympathetic chain ganglion cells are devoid of staining [15]. NeuN primarily stains the nucleus but the cytoplasm is also immunoreactive, though to a lesser extent. For the most part, the cytoplasmic staining is concentrated in the soma
170
¨ nal-Cßevik et al. / Brain Research 1015 (2004) 169–174 I. U
although it extends a short distance into some of the processes (mostly dendritic). NeuN is likely to be a transcription factor that is expressed in the nucleus and cytoplasm of mature neurons [15]. NeuN does not stain the nuclei of immature nerve cells until they complete their development [18]. Although healthy neurons normally express an intense NeuN signal, NeuN immunoreactivity has been reported to decrease under several pathological conditions that adversely affect neuronal viability such as cerebral ischemia, hypoxia and trauma [2,9,19]. Interestingly, this reduction in immunostainability has been considered as an evidence of neuronal loss in several studies [2,9,19 –21]. However, loss of NeuN immunoreactivity may not necessarily indicate disintegration and disappearance of neurons, but may also be due to the depletion of NeuN protein or loss of its antigenicity. In this study, we investigated whether or not the neurons that lost their NeuN immunoreactivity were indeed dissipated and lost their morphological integrity after mild cerebral ischemia. We preferred a mild ischemia model because it induces less tissue destruction and less neuronal cell loss.
2. Material and methods 2.1. MCA occlusion Twenty-four Swiss albino mice weighing 35 – 40 g were housed under diurnal lighting conditions and fasted overnight but allowed free access to water before the experiment. Animal housing, care and application of experimental procedures were all done in accordance with institutional guidelines and all efforts were made to minimize animal suffering and to use only the number of animals necessary to produce reliable scientific data. Mice were intraperitoneally anesthetized with chloral hydrate 300 mg/kg. Body temperature was monitored by rectal probe and maintained at 37.0 F 0.1 jC via a homoeothermic blanket. Proximal occlusion of the right middle cerebral artery (MCA) was performed by a nylon filament as previously described [5]. Briefly, the right common and external carotid arteries were ligated by a 5-0 silk suture following a midline incision. A nylon filament (80) was inserted into the common carotid through a small incision proximal to the bifurcation and advanced in the internal carotid artery up to the origin of MCA (10 mm from the bifurcation). The distal 3-mm of 8-0 filament was coated with silicon. A flexible probe (PF-318 of PeriFlux PF 2B, Perimed) was placed over the skull (2 mm posterior, 6 mm lateral to bregma), away from large pial vessels to monitor the regional cerebral blood flow (rCBF) by laser-Doppler flowmetry. After obtaining a stable 10-min epoch of preischemic rCBF (100%), the MCA was occluded and rCBF (around 10 – 20% of baseline) was monitored for the duration of ischemia and for the first 5– 10 min of reperfusion until
the rCBF was recovered. Reperfusion was accomplished by pulling the filament back. Mice were sacrificed 6 h (n = 6), 24 h (n = 6) and 72 h (n = 3) h following 30 min of ischemia, or 24 h after 60 min (n = 3) of ischemia. 2.2. Histologic methods After reperfusion, mice were sacrificed by transcardially perfusing them with heparin containing saline followed by 4% formaldehyde. Subsequently, brains were removed and kept in formaldehyde for 2 days at + 4 jC and then embedded in paraffin. Coronal sections (5 Am thick) were obtained and the conventional avidin – biotin – peroxidase technique was carried out with anti-NeuN antibody (Chemicon, 1:500 dilution). Diaminobenzidine was used as chromogen, and hematoxylin, as counter stain. Adjacent sections were stained with H&E. For fluorescent immunolabeling, mice were perfused with heparin containing saline followed by 4% formaldehyde, postfixed in the same fixative for 24 h, and the brains were later cryoprotected in 20% and then 30% sucrose in phosphate-buffered saline (PBS) for 24 h at + 4 jC. Coronal sections (20 Am) were cut in a cryostat and double labeled with anti-NeuN (Chemicon; 1:500 dilution) and anti-caspase-3p20 antibodies (Cell Signaling; 1:200 dilution) by using Tyramide amplification kit (Molecular Probes). Hoechst-33258 (Molecular Probes) was added to the mounting medium to counter stain the nuclei. NeuN positive cells were counted manually under 400 magnification in penumbral and core cortical regions and their contralateral homologues. Three vertically adjacent, non-overlapping circular microscopic grids were sampled from each region and the mean number of NeuN-positive cells per mm2 of each section was calculated by averaging the counts obtained from these areas. Means were compared with Kruskal – Wallis variance analysis followed by Mann – Whitney U-test. A value of P < 0.05 was considered to be significant. Means in the text are given with their standard errors except in figures, where standard deviations are illustrated. 2.3. Western blot analysis Tissue corresponding to the MCA territory was separated from the ischemic (n = 6) and control non-ischemic brains (n = 3). Fresh brain samples were homogenized in RIPA buffer (20 mM Tris –HCl, pH 7.5, 50 mM NaCl, 1 mM EDTA, 0.5% NP40, 0.5% SDS, 0.5% deoxycholic acid). The homogenate was centrifuged at 14,000 g for 20 min at 4 jC, and the supernatant was used for analysis. After the same volume of Tris –glycine sodium dodecyl sulfate sample buffer (Invitrogen) was added to the supernatant, equal amounts of the samples were loaded per lane. Proteins were transferred to PVDF membranes and incubated with the primary antibody (anti-NeuN mouse monoclonal, Chemicon 1:500). Western blots were performed with horseradish
¨ nal-Cßevik et al. / Brain Research 1015 (2004) 169–174 I. U
171
peroxidase-conjugated anti-mouse immunoglobulin G (Amersham Biosciences) and were developed using enhanced chemiluminescence kit (Amersham Biosciences). h-Actin was used as an internal standard. Densitometric measurements are expressed as ratios to h-actin values.
3. Results The number of NeuN-labeled cells was significantly ( p < 0.05) decreased by 27% in the penumbra and by 62% in the ischemic core 6 h after brief focal ischemia (Fig. 1; for absolute values of cell counts, see Fig. 2). Despite a reduction in the number of NeuN-positive neurons, H&E staining of adjacent sections did not disclose advanced degenerative changes such as ghost cells and cellular disintegration in the penumbra region, but instead showed that presumable neurons (by H&E criteria) had shrunken cytoplasms and, some, pyknotic nuclei (Fig. 1). At 24 h, the loss of immunoreactivity spread to the whole ischemic MCA territory. Compared to the counts at the 6-h time point, there was a further but statistically insignificant decrease in the number of labeled cells in the penumbral and core regions (for absolute values of cell counts, Fig. 2). When sections from two mice in the 24-h group were immunostained following antigenic retrieval by pressure cooking in citrate buffer for 5 min, the number of NeuN-
Fig. 2. Quantification of the number of NeuN-immunopositive cells showed significant decline 6 and 24 h after a 30-min ischemia in the penumbral and core regions of the MCA territory. *Statistically significant ( P < 0.05).
positive neurons increased by 31% in the penumbra (1707 F 372 vs. 1295 F 305 cells/mm2) and by 26% in the core (955 F 295 vs. 755 F 245 cells/mm2) compared to the equivalent sections stained without retrieval. Most of this increase was due to easier identification of hardly noticeable perikaryal NeuN signal in shrunken neurons, in which the weak nuclear signal was possibly masked by hematoxylin staining. The reduction in staining intensity was also clearly illustrated with immunofluorescent labeling (Fig. 3c,d).
Fig. 1. The number of NeuN-labeled cells was decreased 6 h after brief (30 min) MCA occlusion even at the border zone of the ischemic area, which is supported by collateral blood flow during ischemia (left upper and lower panels, 100 and 200 , respectively). Loss of immunoreactivity was more prominent in core regions (left upper panel). Some of the neurons in the penumbral area (right lower panel corresponding to the boxed area in the left, 1000 ) had shrunken cytoplasm and pyknotic nuclei, and displayed reduced or no NeuN immunoreactivity in contrast to normally labeled cells (densely stained nucleus and cytoplasm extending into the proximal parts of a few processes). H&E staining of the adjacent section (right upper panel, 1000 ) illustrated that presumable neurons in this area displayed scalloping, shrunken cytoplasm and pyknotic nuclei but were not disintegrated, indicating that the disappearance of NeuN immunoreactivity in some neurons does not necessarily mean neuronal cell loss.
172
¨ nal-Cßevik et al. / Brain Research 1015 (2004) 169–174 I. U
Fig. 3. Immunofluorescent staining with anti-NeuN antibodies showed that the nuclear staining was dramatically reduced or completely lost in some penumbral neurons 72 h after 30 min of ischemia in the frontal cortex (a, 1000 ). Hoechst 33258 staining (b) of the same sections revealed that the nuclei of these neurons were preserved and had an intact nuclear membrane. Compared with the homologue region in the non-ischemic frontal cortex (c, 400 ) the number of NeuNpositive cells were reduced 24 h after 60 min of ischemia (d, 400 ). Double staining with antibodies against caspase-3p20 (e, 400 ) disclosed that some of the caspase-3p20-positive neurons were NeuN-negative. However, their cellular contours looked the same as the NeuN-positive ischemic neurons, suggesting that they preserved their cellular integrity although they had lost NeuN immunoreactivity.
The nuclear staining was considerably reduced or lost in some fluorescent-labeled neurons (Fig. 3a). However, Hoechst 33258 staining of the same sections revealed that the nucleus was preserved with an intact nuclear membrane in these neurons (Fig. 3b). We also attempted to see whether or not the ischemic neurons that had completely lost their NeuN immunoreactivity (cytoplasmic as well as nuclear) retained their morphological integrity by visualizing their cellular boundaries with antibodies against the cleaved, active form (p20) of caspase-3. Caspase-3-p20 appears exclusively in neurons after ischemia and is found in the cytoplasm and nuclei, usually unveiling the image of the neuron that it is expressed in [2,16]. We found that 24 h following a 1-h ischemia, 47 F 12% of the caspase-3-p20 positive neurons were not labeled with NeuN, yet they preserved their morphological integrity although they were shrunken like the NeuN-positive ischemic neurons (Fig. 3e). The considerable reduction in colocalization rate was in contrast to complete overlap of NeuN and caspase-3-p20 immunoreactivities 1 h after recirculation in preliminary experiments. To determine whether the reduced NeuN immunoreactivity was caused by depletion of NeuN protein or by loss of its antigenicity, we detected NeuN protein levels in the ischemic brain. Western blot analysis of MCA territories prepared from the non-ischemic brains of naı¨ve control mice revealed four NeuN immunoreactive bands in agreement with McPhail et al.’s study [13] (Fig. 4). However, we found no change in the intensity of any of these bands in samples prepared from the ischemic MCA territory 6 or 24 h later.
4. Discussion Velier et al. [20] examined the temporal kinetics of neuronal cell loss and of cellular DNA damage after permanent MCA occlusion by using the neuron-specific marker NeuN and TUNEL. They concluded that there was an early loss of neurons (46% of the entire population) within the first 24 h that did not involve DNA damage and was presumably a form of necrotic cell loss. This conclusion was based on loss of NeuN immunoreactivity. Apparently, the authors considered loss of NeuN staining as a marker of neuronal loss, possibly induced by necrosis. Similarly, Sugawara et al. [19] demonstrated that 3 days after 10 or 5 min of global ischemia in rats, NeuN immunostaining showed a marked reduction in the CA1 pyramidal cell layer. These authors also used the NeuN positivity as a measure of the number of surviving neurons after ischemic insult. Davoli and his colleagues [2] investigated caspase-3 activation and DNA fragmentation following transient focal ischemia in the rat brain. After double labeling with TUNEL and NeuN, they showed that NeuN immunoreactivity was dramatically reduced by 24 h after ischemia and this was correlated with an increase in TUNEL labeling. Accordingly, they suggested that NeuN expression was decreased due to loss of neuronal viability. The authors also concluded that caspase-3 cleavage 3 and 6 h after ischemia was confined to the neuronal cells, which were still viable and thus immunoreactive for NeuN. Similar approaches were used to monitor neuronal loss after traumatic brain injury and in vitro oxygen-glucose deprivation [9,21].
¨ nal-Cßevik et al. / Brain Research 1015 (2004) 169–174 I. U
Fig. 4. (A) Western blot analysis of MCA territories prepared from the nonischemic controls revealed four NeuN immunoreactive bands in the mouse brain. No change was detected in the intensity of any of these bands in brain samples prepared from the ischemic (30 min) MCA territory 6 or 24 h after reperfusion. (B) Densitometric measurements of NeuN bands relative to h-actin values. The order of bands (1 – 4) are from the lowest to highest molecular weight.
In all of the examples above, the reduction in NeuN immunostainability was considered as an evidence of neuronal loss by the authors. Although there is no established evidence for this contention, it is conceivable that irreparably damaged neurons that are destined to die due to severe metabolic perturbations will lose the NeuN immunoreactivity. Our data showing presence of caspase activity in neurons that have lost NeuN staining support this idea. Hence, the loss of NeuN immunoreactivity may be considered as a surrogate marker of irreversible neuronal injury and a predictor of neuronal death. However, loss of NeuN labeling does not necessarily indicate disintegration (loss) of neurons; instead, it may also be caused by depletion of the protein or by a change in its antigenicity. In other words, an injured but still viable neuron may lose the NeuN protein due to suppression of protein synthesis and/or to its enhanced consumption [7]. A similar depletion has also been reported for several other proteins including another neuronspecific protein, the microtubule-associated protein-2 (MAP-2) [1,11,12,14]. The cerebral blood flow threshold for inhibition of protein synthesis is well above the threshold for irreversible neuronal injury [8]. Hence protein synthesis
173
decreases even after mild perturbations before irreversible damage takes place although some proteins are preferentially expressed in injured neurons [6]. To test the above possibility, we used a mild ischemia model (30 min ischemia). Contrary to transient focal ischemia with longer duration or to permanent ischemia models used in the above studies, neuronal death progresses more slowly in this model [3,4,16,22]. We have demonstrated by several means that neurons can still maintain the integrity of their nuclear and cytoplasmic morphology while NeuN immunoreactivity decreases. For example, NeuN immunoreactivity was significantly depressed 6 h after 30 min of MCA occlusion even in the peri-infarct areas that were supported by collateral blood supply during occlusion. H&E staining showed that neurons in these areas were at an early stage of neurodegeneration. Similarly, Hoechst 33258 staining revealed that the nucleus was preserved and had an intact membrane in neurons with no nuclear NeuN immunoreactivity. We have also shown that even ischemic neurons that completely lost their NeuN immunoreactivity maintained their cellular integrity, as demonstrated by a surrogate neuronal marker, caspase-3-p20 under ischemic conditions. We used caspase-3-p20 instead of other established markers such as MAP-2 (which are also depleted during ischemia) [1,11,12,14] because this cleaved, active form of caspase-3 is not constitutively present but emerges exclusively in neurons after ischemia [2,16,22]. Demonstration of the presence of morphologically intact neurons that have completely lost NeuN immunoreactivity points to an important potential source of error in interpreting NeuN labeling of the ischemic tissue. For example, although TUNEL staining is known to be observed almost exclusively in neurons in the ischemic brain [2,3,10]. Namura et al. [16] reported that about 60% of the TUNEL positive nuclei was not colocalized with NeuN labeling. This discrepancy was possibly caused by the decreased NeuN immunoreactivity of neurons, but not because of them being non-neuronal cells [16,22]. Interestingly, the NeuN protein level did not show any appreciable decrease even 24 h after reperfusion, strongly suggesting that loss of NeuN antigenicity—rather than reduced protein levels—accounts for loss of immunoreactivity in this relatively mild brain injury model. Indeed, the number of NeuN-labeled neurons was partially restored after antigenic retrieval, at least in the penumbra. However, NeuN protein levels may significantly decrease after a more severe injury as has been reported following axotomy [13]. In conclusion, since NeuN antigenicity is readily reduced in neurons after brain injury induced by metabolic perturbations, loss of NeuN immunoreactivity should not be taken as a definite indicator of neuronal loss and, quantitative analysis of NeuN positive neurons should be used with caution after central nervous tissue injury. Although severely injured neurons that have lost their NeuN immunoreactivity are very likely to be destined to die, there is as yet no evidence implying the loss of NeuN immunoreactivity as a
174
¨ nal-Cßevik et al. / Brain Research 1015 (2004) 169–174 I. U
specific and early marker of irreversible injury. On the other hand, erroneous conclusions may be reached due to falsenegative identification of neurons after CNS damage and also during evaluation of cell death after mild and potentially reversible injuries.
[10]
[11]
Acknowledgements
[12]
Dr. T. Dalkara’s work is supported by the Turkish Academy of Sciences.
[13]
References [1] C.J. Chen, S.L. Liao, W.Y. Chen, J.S. Hong, J.S. Kuo, Cerebral ischemia/reperfusion injury in rat brain: effects of naloxane, NeuroReport 12 (6) (2001 May 8) 1245 – 1249. [2] M.A. Davoli, J. Fortunis, J. Tam, S. Xanthoudakis, D. Nicholdon, G.S. Ng, G.Y.K. Ng, D. Xu, Immunohistochemical and biochemical assessment of caspase-3 activation and DNA fragmentation following transient focal ischemia in the rat, Neuroscience 115 (2002) 125 – 136. [3] M. Endres, S. Namura, M. Shimizu-Sasamata, C. Waeber, L. Zhang, T. Gomez-Isla, B.T. Hyman, M.A. Moskowitz, Attenuation of delayed neuronal death after mild focal ischemia in mice by inhibition of the caspase family, J. Cereb. Blood Flow Metab. 18 (3) (1998) 238 – 247. [4] K. Fink, J. Zhu, S. Namura, M. Shimizu-Sasamata, M. Endres, J. Ma, T. Dalkara, J. Yuan, M.A. Moskowitz, Prolonged therapeutic window for ischemic brain damage caused by delayed caspase activation, J. Cereb. Blood Flow Metab. 18 (10) (1998) 1071 – 1076. ¨ zdemir, H. Bolay, O. Saribas, T. Dalkara, Role of endo[5] Y. Gu¨rsoy-O thelial nitric oxide generation and peroxynitrite formation in reperfusion injury after focal cerebral ischemia, Stroke 31 (8) (2000 Aug) 1974 – 1980 (discussion 1981). [6] D.M. Hermann, E. Kilic, R. Hata, K.A. Hossmann, G. Mies, Relationship between metabolic dysfunctions, gene responses and delayed cell death after mild focal cerebral ischemia in mice, Neuroscience 104 (2001) 947 – 955. [7] K.A. Hossmann, Disturbances of cerebral protein synthesis and ischemic cell death, Prog. Brain Res. 96 (1993) 161 – 177. [8] K.A. Hossmann, Viability thresholds and the penumbra of focal ischemia, Ann. Neurol. 36 (1994) 557 – 565. [9] T. Igarashi, T.T. Huang, L.J. Noble, Regional vulnerability after traumatic brain injury: gender differences in mice that over express hu-
[14]
[15] [16]
[17] [18]
[19]
[20]
[21]
[22]
man copper, zinc superoxide dismutase, Exp. Neurol. 172 (2001 Dec) 332 – 341. Y. Li, M. Chopp, N. Jiang, F. Yao, C. Zaloga, Temporal profile of in situ DNA fragmentation after transient middle cerebral artery occlusion in the rat, J. Cereb. Blood Flow Metab. 15 (3) (1995) 389 – 397. Y. Li, N. Jiang, C. Powers, M. Chopp, Neuronal damage and plasticity identified by microtubule-associated protein 2, growth-associated protein 43, and cyclin D1 immunoreactivity after focal cerebral ischemia in rats, Stroke 29 (9) (1998) 1972 – 1980. D.F. Matesic, R.C. Lin, MAP-2 protein as an early indicator of ischemia induced neurodegeneration in gerbil forebrain, J. Neurochem. 63 (3) (1994 Sep) 1012 – 1020. L.T. McPhail, C.B. McBride, J. McGraw, J.D. Steeves, W. Tetzlaff, Axotomy abolishes NeuN expression in facial but not rubrospinal neurons, Exp. Neurol. 185 (2004) 182 – 190. S.L. Minger, J.W. Geddes, M.L. Holtz, S.D. Craddock, S.W. Whiteheart, R.G. Siman, L.C. Pettigrew, Glutamate receptor antagonists inhibit calpain-mediated cytoskeletal proteolysis in focal cerebral ischemia, Brain Res. 810 (1998) 181 – 199. R.J. Mullen, C.R. Buck, A.M. Smith, NeuN, a neuronal specific nuclear protein in vertebrates, Development 116 (1992) 201 – 211. S. Namura, J. Zhu, K. Fink, M. Endres, A. Srinivasan, K.J. Tomaselli, J. Yuan, M.A. Moskowitz, Activation and cleavage of caspase-3 in apoptosis induced by experimental cerebral ischemia, J. Neurosci. 18 (10) (1998) 3659 – 3668. A. Relo, J. Feldon, Specific neuronal protein: a new tool for histological evaluation of excitotoxic lesions, Physiol. Behav. 76 (2002) 449 – 456. H.B. Sarnat, D. Nochlin, D.E. Born, Neuronal nuclear antigen (NeuN): a marker of neuronal maturation in the early human fetal nervous system, Brain Dev. 20 (1998) 88 – 94. T. Sugawara, A. Lewen, N. Noshita, Y. Gasche, H.P. Chan, Effects of global ischemia duration on neuronal, astroglial, oligodendroglial and microglial reactions in the vulnerable hippocampal CA1 subregion in rats, J. Neurotrauma 19 (1) (2002) 85 – 98. J.J. Velier, J.A. Ellison, K.K. Kikly, A.P. Spera, F.C. Barone, G.Z. Feuerstein, Caspase-8 and caspase-3 are expressed by different populations of cortical neurons undergoing delayed cell death after focal stroke in the rat, J. Neurosci. 19 (14) (1999) 5932 – 5941. G.P. Xu, K.R. Dave, R. Vivero, R. Schmidt-Kastner, J.T. Sick, M. Perez-Pinzon, Improvement in neuronal survival after ischemic preconditioning in hippocampal slice cultures, Brain Res. 952 (2002) 153 – 158. C. Zhu, X. Wang, H. Hagberg, K. Blomerg, Correlation between caspase-3 activation and three different markers of DNA damage in neonatal cerebral hypoxia-ischemia, J. Neurochem. 75 (2) (2000 Aug) 819 – 829.
Research report
Loss of NeuN immunoreactivity after cerebral ischemia does not indicate neuronal cell loss: a cautionary note ¨ zdemir a, Gunfer Gurer a, Turgay Dalkara a,* ¨ nal-Cßevik a, Mu¨nire Kılıncßb, Yasemin Gu¨rsoy-O Isın U a
Institute of Neurological Sciences and Psychiatry, and Faculty of Medicine, Department of Neurology, Hacettepe University, Sıhhiye Ankara 06100, Turkey b Faculty of Medicine, Department of Neurology, Baskent University, Ankara, Turkey Accepted 9 April 2004 Available online
Abstract NeuN immunoreactivity is used as a specific marker for neurons. The number of NeuN-positive cells decreases under pathological conditions. This finding is usually considered as an evidence of neuronal loss. However, decrease in NeuN labeling may also be caused by depletion of the protein or loss of its antigenicity. Hence, we have investigated the morphological features of neurons that lost NeuN immunoreactivity and the NeuN protein levels in mouse brain after cerebral ischemia. The number of NeuN-labeled cells was decreased 6 h after a mild ischemic insult (30 min middle cerebral artery occlusion) in penumbral and core regions. Hematoxylin and eosin (H&E) staining of adjacent sections showed that neurons in the penumbra were not disintegrated but displayed early ischemic changes. The nuclear NeuN staining was dramatically reduced or lost in some neurons. However, Hoechst 33258 staining of the same sections revealed that these nuclei were preserved with an intact membrane. Labeling of neurons that had lost NeuN-positivity with antibodies against caspase-3-p20, which is constitutively not present but emerges in neurons after ischemia, disclosed that these neurons still preserved their integrity. Moreover, Western blots showed that NeuN protein levels were not decreased, suggesting that reduced NeuN antigenicity accounted for loss of immunoreactivity in this mild brain injury model. Supporting this idea, NeuN labeling was partially restored after antigenic retrieval. In conclusion, since NeuN immunoreactivity readily decreases after metabolic perturbations, reduced NeuN labeling should not be taken as an indicator of neuronal loss and, quantitative analysis based on NeuN-positivity should be used cautiously after central nervous system (CNS) injury. D 2004 Elsevier B.V. All rights reserved. Theme: Cellular and molecular biology Topic: Staining, tracing and imaging techniques Keywords: Cerebral ischemia; NeuN immunoreactivity; Caspase-3-p20
1. Introduction In order to differentiate neurons from glial cells in brain sections, specific neuronal markers, such as neuron-specific enolase, neurofilament proteins and calbindin, can be used in immunohistochemical procedures [17]. However, most of these proteins are present in distinct subpopulations of neurons, limiting their application for histological identification of these cells. In order to overcome this obstacle, Mullen et al. [15] developed a monoclonal antibody that recognizes a neuron specific nuclear protein. NeuN, which * Corresponding author. Tel.: +90-312-3051809; fax: +90-3123093451. E-mail address: [email protected] (T. Dalkara). 0006-8993/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2004.04.032
stands for neuron-specific nuclear protein, is expressed in nucleus and cell body of most neuronal cell types in rodents, chicks and humans [15]. It is not expressed in glial cells; neither oligodendrocytes nor astrocytes nor do microglial cells express NeuN. Since its development, NeuN has been successfully used as a neuronal marker in a wide variety of studies in cell cultures as well as in diagnostic histopathology. Thus, NeuN has been established as a universal marker for neurons although a few neuronal cell types such as cerebellar Purkinje cells, retinal photoreceptor cells, mitral cells of the olfactory bulbs, and sympathetic chain ganglion cells are devoid of staining [15]. NeuN primarily stains the nucleus but the cytoplasm is also immunoreactive, though to a lesser extent. For the most part, the cytoplasmic staining is concentrated in the soma
170
¨ nal-Cßevik et al. / Brain Research 1015 (2004) 169–174 I. U
although it extends a short distance into some of the processes (mostly dendritic). NeuN is likely to be a transcription factor that is expressed in the nucleus and cytoplasm of mature neurons [15]. NeuN does not stain the nuclei of immature nerve cells until they complete their development [18]. Although healthy neurons normally express an intense NeuN signal, NeuN immunoreactivity has been reported to decrease under several pathological conditions that adversely affect neuronal viability such as cerebral ischemia, hypoxia and trauma [2,9,19]. Interestingly, this reduction in immunostainability has been considered as an evidence of neuronal loss in several studies [2,9,19 –21]. However, loss of NeuN immunoreactivity may not necessarily indicate disintegration and disappearance of neurons, but may also be due to the depletion of NeuN protein or loss of its antigenicity. In this study, we investigated whether or not the neurons that lost their NeuN immunoreactivity were indeed dissipated and lost their morphological integrity after mild cerebral ischemia. We preferred a mild ischemia model because it induces less tissue destruction and less neuronal cell loss.
2. Material and methods 2.1. MCA occlusion Twenty-four Swiss albino mice weighing 35 – 40 g were housed under diurnal lighting conditions and fasted overnight but allowed free access to water before the experiment. Animal housing, care and application of experimental procedures were all done in accordance with institutional guidelines and all efforts were made to minimize animal suffering and to use only the number of animals necessary to produce reliable scientific data. Mice were intraperitoneally anesthetized with chloral hydrate 300 mg/kg. Body temperature was monitored by rectal probe and maintained at 37.0 F 0.1 jC via a homoeothermic blanket. Proximal occlusion of the right middle cerebral artery (MCA) was performed by a nylon filament as previously described [5]. Briefly, the right common and external carotid arteries were ligated by a 5-0 silk suture following a midline incision. A nylon filament (80) was inserted into the common carotid through a small incision proximal to the bifurcation and advanced in the internal carotid artery up to the origin of MCA (10 mm from the bifurcation). The distal 3-mm of 8-0 filament was coated with silicon. A flexible probe (PF-318 of PeriFlux PF 2B, Perimed) was placed over the skull (2 mm posterior, 6 mm lateral to bregma), away from large pial vessels to monitor the regional cerebral blood flow (rCBF) by laser-Doppler flowmetry. After obtaining a stable 10-min epoch of preischemic rCBF (100%), the MCA was occluded and rCBF (around 10 – 20% of baseline) was monitored for the duration of ischemia and for the first 5– 10 min of reperfusion until
the rCBF was recovered. Reperfusion was accomplished by pulling the filament back. Mice were sacrificed 6 h (n = 6), 24 h (n = 6) and 72 h (n = 3) h following 30 min of ischemia, or 24 h after 60 min (n = 3) of ischemia. 2.2. Histologic methods After reperfusion, mice were sacrificed by transcardially perfusing them with heparin containing saline followed by 4% formaldehyde. Subsequently, brains were removed and kept in formaldehyde for 2 days at + 4 jC and then embedded in paraffin. Coronal sections (5 Am thick) were obtained and the conventional avidin – biotin – peroxidase technique was carried out with anti-NeuN antibody (Chemicon, 1:500 dilution). Diaminobenzidine was used as chromogen, and hematoxylin, as counter stain. Adjacent sections were stained with H&E. For fluorescent immunolabeling, mice were perfused with heparin containing saline followed by 4% formaldehyde, postfixed in the same fixative for 24 h, and the brains were later cryoprotected in 20% and then 30% sucrose in phosphate-buffered saline (PBS) for 24 h at + 4 jC. Coronal sections (20 Am) were cut in a cryostat and double labeled with anti-NeuN (Chemicon; 1:500 dilution) and anti-caspase-3p20 antibodies (Cell Signaling; 1:200 dilution) by using Tyramide amplification kit (Molecular Probes). Hoechst-33258 (Molecular Probes) was added to the mounting medium to counter stain the nuclei. NeuN positive cells were counted manually under 400 magnification in penumbral and core cortical regions and their contralateral homologues. Three vertically adjacent, non-overlapping circular microscopic grids were sampled from each region and the mean number of NeuN-positive cells per mm2 of each section was calculated by averaging the counts obtained from these areas. Means were compared with Kruskal – Wallis variance analysis followed by Mann – Whitney U-test. A value of P < 0.05 was considered to be significant. Means in the text are given with their standard errors except in figures, where standard deviations are illustrated. 2.3. Western blot analysis Tissue corresponding to the MCA territory was separated from the ischemic (n = 6) and control non-ischemic brains (n = 3). Fresh brain samples were homogenized in RIPA buffer (20 mM Tris –HCl, pH 7.5, 50 mM NaCl, 1 mM EDTA, 0.5% NP40, 0.5% SDS, 0.5% deoxycholic acid). The homogenate was centrifuged at 14,000 g for 20 min at 4 jC, and the supernatant was used for analysis. After the same volume of Tris –glycine sodium dodecyl sulfate sample buffer (Invitrogen) was added to the supernatant, equal amounts of the samples were loaded per lane. Proteins were transferred to PVDF membranes and incubated with the primary antibody (anti-NeuN mouse monoclonal, Chemicon 1:500). Western blots were performed with horseradish
¨ nal-Cßevik et al. / Brain Research 1015 (2004) 169–174 I. U
171
peroxidase-conjugated anti-mouse immunoglobulin G (Amersham Biosciences) and were developed using enhanced chemiluminescence kit (Amersham Biosciences). h-Actin was used as an internal standard. Densitometric measurements are expressed as ratios to h-actin values.
3. Results The number of NeuN-labeled cells was significantly ( p < 0.05) decreased by 27% in the penumbra and by 62% in the ischemic core 6 h after brief focal ischemia (Fig. 1; for absolute values of cell counts, see Fig. 2). Despite a reduction in the number of NeuN-positive neurons, H&E staining of adjacent sections did not disclose advanced degenerative changes such as ghost cells and cellular disintegration in the penumbra region, but instead showed that presumable neurons (by H&E criteria) had shrunken cytoplasms and, some, pyknotic nuclei (Fig. 1). At 24 h, the loss of immunoreactivity spread to the whole ischemic MCA territory. Compared to the counts at the 6-h time point, there was a further but statistically insignificant decrease in the number of labeled cells in the penumbral and core regions (for absolute values of cell counts, Fig. 2). When sections from two mice in the 24-h group were immunostained following antigenic retrieval by pressure cooking in citrate buffer for 5 min, the number of NeuN-
Fig. 2. Quantification of the number of NeuN-immunopositive cells showed significant decline 6 and 24 h after a 30-min ischemia in the penumbral and core regions of the MCA territory. *Statistically significant ( P < 0.05).
positive neurons increased by 31% in the penumbra (1707 F 372 vs. 1295 F 305 cells/mm2) and by 26% in the core (955 F 295 vs. 755 F 245 cells/mm2) compared to the equivalent sections stained without retrieval. Most of this increase was due to easier identification of hardly noticeable perikaryal NeuN signal in shrunken neurons, in which the weak nuclear signal was possibly masked by hematoxylin staining. The reduction in staining intensity was also clearly illustrated with immunofluorescent labeling (Fig. 3c,d).
Fig. 1. The number of NeuN-labeled cells was decreased 6 h after brief (30 min) MCA occlusion even at the border zone of the ischemic area, which is supported by collateral blood flow during ischemia (left upper and lower panels, 100 and 200 , respectively). Loss of immunoreactivity was more prominent in core regions (left upper panel). Some of the neurons in the penumbral area (right lower panel corresponding to the boxed area in the left, 1000 ) had shrunken cytoplasm and pyknotic nuclei, and displayed reduced or no NeuN immunoreactivity in contrast to normally labeled cells (densely stained nucleus and cytoplasm extending into the proximal parts of a few processes). H&E staining of the adjacent section (right upper panel, 1000 ) illustrated that presumable neurons in this area displayed scalloping, shrunken cytoplasm and pyknotic nuclei but were not disintegrated, indicating that the disappearance of NeuN immunoreactivity in some neurons does not necessarily mean neuronal cell loss.
172
¨ nal-Cßevik et al. / Brain Research 1015 (2004) 169–174 I. U
Fig. 3. Immunofluorescent staining with anti-NeuN antibodies showed that the nuclear staining was dramatically reduced or completely lost in some penumbral neurons 72 h after 30 min of ischemia in the frontal cortex (a, 1000 ). Hoechst 33258 staining (b) of the same sections revealed that the nuclei of these neurons were preserved and had an intact nuclear membrane. Compared with the homologue region in the non-ischemic frontal cortex (c, 400 ) the number of NeuNpositive cells were reduced 24 h after 60 min of ischemia (d, 400 ). Double staining with antibodies against caspase-3p20 (e, 400 ) disclosed that some of the caspase-3p20-positive neurons were NeuN-negative. However, their cellular contours looked the same as the NeuN-positive ischemic neurons, suggesting that they preserved their cellular integrity although they had lost NeuN immunoreactivity.
The nuclear staining was considerably reduced or lost in some fluorescent-labeled neurons (Fig. 3a). However, Hoechst 33258 staining of the same sections revealed that the nucleus was preserved with an intact nuclear membrane in these neurons (Fig. 3b). We also attempted to see whether or not the ischemic neurons that had completely lost their NeuN immunoreactivity (cytoplasmic as well as nuclear) retained their morphological integrity by visualizing their cellular boundaries with antibodies against the cleaved, active form (p20) of caspase-3. Caspase-3-p20 appears exclusively in neurons after ischemia and is found in the cytoplasm and nuclei, usually unveiling the image of the neuron that it is expressed in [2,16]. We found that 24 h following a 1-h ischemia, 47 F 12% of the caspase-3-p20 positive neurons were not labeled with NeuN, yet they preserved their morphological integrity although they were shrunken like the NeuN-positive ischemic neurons (Fig. 3e). The considerable reduction in colocalization rate was in contrast to complete overlap of NeuN and caspase-3-p20 immunoreactivities 1 h after recirculation in preliminary experiments. To determine whether the reduced NeuN immunoreactivity was caused by depletion of NeuN protein or by loss of its antigenicity, we detected NeuN protein levels in the ischemic brain. Western blot analysis of MCA territories prepared from the non-ischemic brains of naı¨ve control mice revealed four NeuN immunoreactive bands in agreement with McPhail et al.’s study [13] (Fig. 4). However, we found no change in the intensity of any of these bands in samples prepared from the ischemic MCA territory 6 or 24 h later.
4. Discussion Velier et al. [20] examined the temporal kinetics of neuronal cell loss and of cellular DNA damage after permanent MCA occlusion by using the neuron-specific marker NeuN and TUNEL. They concluded that there was an early loss of neurons (46% of the entire population) within the first 24 h that did not involve DNA damage and was presumably a form of necrotic cell loss. This conclusion was based on loss of NeuN immunoreactivity. Apparently, the authors considered loss of NeuN staining as a marker of neuronal loss, possibly induced by necrosis. Similarly, Sugawara et al. [19] demonstrated that 3 days after 10 or 5 min of global ischemia in rats, NeuN immunostaining showed a marked reduction in the CA1 pyramidal cell layer. These authors also used the NeuN positivity as a measure of the number of surviving neurons after ischemic insult. Davoli and his colleagues [2] investigated caspase-3 activation and DNA fragmentation following transient focal ischemia in the rat brain. After double labeling with TUNEL and NeuN, they showed that NeuN immunoreactivity was dramatically reduced by 24 h after ischemia and this was correlated with an increase in TUNEL labeling. Accordingly, they suggested that NeuN expression was decreased due to loss of neuronal viability. The authors also concluded that caspase-3 cleavage 3 and 6 h after ischemia was confined to the neuronal cells, which were still viable and thus immunoreactive for NeuN. Similar approaches were used to monitor neuronal loss after traumatic brain injury and in vitro oxygen-glucose deprivation [9,21].
¨ nal-Cßevik et al. / Brain Research 1015 (2004) 169–174 I. U
Fig. 4. (A) Western blot analysis of MCA territories prepared from the nonischemic controls revealed four NeuN immunoreactive bands in the mouse brain. No change was detected in the intensity of any of these bands in brain samples prepared from the ischemic (30 min) MCA territory 6 or 24 h after reperfusion. (B) Densitometric measurements of NeuN bands relative to h-actin values. The order of bands (1 – 4) are from the lowest to highest molecular weight.
In all of the examples above, the reduction in NeuN immunostainability was considered as an evidence of neuronal loss by the authors. Although there is no established evidence for this contention, it is conceivable that irreparably damaged neurons that are destined to die due to severe metabolic perturbations will lose the NeuN immunoreactivity. Our data showing presence of caspase activity in neurons that have lost NeuN staining support this idea. Hence, the loss of NeuN immunoreactivity may be considered as a surrogate marker of irreversible neuronal injury and a predictor of neuronal death. However, loss of NeuN labeling does not necessarily indicate disintegration (loss) of neurons; instead, it may also be caused by depletion of the protein or by a change in its antigenicity. In other words, an injured but still viable neuron may lose the NeuN protein due to suppression of protein synthesis and/or to its enhanced consumption [7]. A similar depletion has also been reported for several other proteins including another neuronspecific protein, the microtubule-associated protein-2 (MAP-2) [1,11,12,14]. The cerebral blood flow threshold for inhibition of protein synthesis is well above the threshold for irreversible neuronal injury [8]. Hence protein synthesis
173
decreases even after mild perturbations before irreversible damage takes place although some proteins are preferentially expressed in injured neurons [6]. To test the above possibility, we used a mild ischemia model (30 min ischemia). Contrary to transient focal ischemia with longer duration or to permanent ischemia models used in the above studies, neuronal death progresses more slowly in this model [3,4,16,22]. We have demonstrated by several means that neurons can still maintain the integrity of their nuclear and cytoplasmic morphology while NeuN immunoreactivity decreases. For example, NeuN immunoreactivity was significantly depressed 6 h after 30 min of MCA occlusion even in the peri-infarct areas that were supported by collateral blood supply during occlusion. H&E staining showed that neurons in these areas were at an early stage of neurodegeneration. Similarly, Hoechst 33258 staining revealed that the nucleus was preserved and had an intact membrane in neurons with no nuclear NeuN immunoreactivity. We have also shown that even ischemic neurons that completely lost their NeuN immunoreactivity maintained their cellular integrity, as demonstrated by a surrogate neuronal marker, caspase-3-p20 under ischemic conditions. We used caspase-3-p20 instead of other established markers such as MAP-2 (which are also depleted during ischemia) [1,11,12,14] because this cleaved, active form of caspase-3 is not constitutively present but emerges exclusively in neurons after ischemia [2,16,22]. Demonstration of the presence of morphologically intact neurons that have completely lost NeuN immunoreactivity points to an important potential source of error in interpreting NeuN labeling of the ischemic tissue. For example, although TUNEL staining is known to be observed almost exclusively in neurons in the ischemic brain [2,3,10]. Namura et al. [16] reported that about 60% of the TUNEL positive nuclei was not colocalized with NeuN labeling. This discrepancy was possibly caused by the decreased NeuN immunoreactivity of neurons, but not because of them being non-neuronal cells [16,22]. Interestingly, the NeuN protein level did not show any appreciable decrease even 24 h after reperfusion, strongly suggesting that loss of NeuN antigenicity—rather than reduced protein levels—accounts for loss of immunoreactivity in this relatively mild brain injury model. Indeed, the number of NeuN-labeled neurons was partially restored after antigenic retrieval, at least in the penumbra. However, NeuN protein levels may significantly decrease after a more severe injury as has been reported following axotomy [13]. In conclusion, since NeuN antigenicity is readily reduced in neurons after brain injury induced by metabolic perturbations, loss of NeuN immunoreactivity should not be taken as a definite indicator of neuronal loss and, quantitative analysis of NeuN positive neurons should be used with caution after central nervous tissue injury. Although severely injured neurons that have lost their NeuN immunoreactivity are very likely to be destined to die, there is as yet no evidence implying the loss of NeuN immunoreactivity as a
174
¨ nal-Cßevik et al. / Brain Research 1015 (2004) 169–174 I. U
specific and early marker of irreversible injury. On the other hand, erroneous conclusions may be reached due to falsenegative identification of neurons after CNS damage and also during evaluation of cell death after mild and potentially reversible injuries.
[10]
[11]
Acknowledgements
[12]
Dr. T. Dalkara’s work is supported by the Turkish Academy of Sciences.
[13]
References [1] C.J. Chen, S.L. Liao, W.Y. Chen, J.S. Hong, J.S. Kuo, Cerebral ischemia/reperfusion injury in rat brain: effects of naloxane, NeuroReport 12 (6) (2001 May 8) 1245 – 1249. [2] M.A. Davoli, J. Fortunis, J. Tam, S. Xanthoudakis, D. Nicholdon, G.S. Ng, G.Y.K. Ng, D. Xu, Immunohistochemical and biochemical assessment of caspase-3 activation and DNA fragmentation following transient focal ischemia in the rat, Neuroscience 115 (2002) 125 – 136. [3] M. Endres, S. Namura, M. Shimizu-Sasamata, C. Waeber, L. Zhang, T. Gomez-Isla, B.T. Hyman, M.A. Moskowitz, Attenuation of delayed neuronal death after mild focal ischemia in mice by inhibition of the caspase family, J. Cereb. Blood Flow Metab. 18 (3) (1998) 238 – 247. [4] K. Fink, J. Zhu, S. Namura, M. Shimizu-Sasamata, M. Endres, J. Ma, T. Dalkara, J. Yuan, M.A. Moskowitz, Prolonged therapeutic window for ischemic brain damage caused by delayed caspase activation, J. Cereb. Blood Flow Metab. 18 (10) (1998) 1071 – 1076. ¨ zdemir, H. Bolay, O. Saribas, T. Dalkara, Role of endo[5] Y. Gu¨rsoy-O thelial nitric oxide generation and peroxynitrite formation in reperfusion injury after focal cerebral ischemia, Stroke 31 (8) (2000 Aug) 1974 – 1980 (discussion 1981). [6] D.M. Hermann, E. Kilic, R. Hata, K.A. Hossmann, G. Mies, Relationship between metabolic dysfunctions, gene responses and delayed cell death after mild focal cerebral ischemia in mice, Neuroscience 104 (2001) 947 – 955. [7] K.A. Hossmann, Disturbances of cerebral protein synthesis and ischemic cell death, Prog. Brain Res. 96 (1993) 161 – 177. [8] K.A. Hossmann, Viability thresholds and the penumbra of focal ischemia, Ann. Neurol. 36 (1994) 557 – 565. [9] T. Igarashi, T.T. Huang, L.J. Noble, Regional vulnerability after traumatic brain injury: gender differences in mice that over express hu-
[14]
[15] [16]
[17] [18]
[19]
[20]
[21]
[22]
man copper, zinc superoxide dismutase, Exp. Neurol. 172 (2001 Dec) 332 – 341. Y. Li, M. Chopp, N. Jiang, F. Yao, C. Zaloga, Temporal profile of in situ DNA fragmentation after transient middle cerebral artery occlusion in the rat, J. Cereb. Blood Flow Metab. 15 (3) (1995) 389 – 397. Y. Li, N. Jiang, C. Powers, M. Chopp, Neuronal damage and plasticity identified by microtubule-associated protein 2, growth-associated protein 43, and cyclin D1 immunoreactivity after focal cerebral ischemia in rats, Stroke 29 (9) (1998) 1972 – 1980. D.F. Matesic, R.C. Lin, MAP-2 protein as an early indicator of ischemia induced neurodegeneration in gerbil forebrain, J. Neurochem. 63 (3) (1994 Sep) 1012 – 1020. L.T. McPhail, C.B. McBride, J. McGraw, J.D. Steeves, W. Tetzlaff, Axotomy abolishes NeuN expression in facial but not rubrospinal neurons, Exp. Neurol. 185 (2004) 182 – 190. S.L. Minger, J.W. Geddes, M.L. Holtz, S.D. Craddock, S.W. Whiteheart, R.G. Siman, L.C. Pettigrew, Glutamate receptor antagonists inhibit calpain-mediated cytoskeletal proteolysis in focal cerebral ischemia, Brain Res. 810 (1998) 181 – 199. R.J. Mullen, C.R. Buck, A.M. Smith, NeuN, a neuronal specific nuclear protein in vertebrates, Development 116 (1992) 201 – 211. S. Namura, J. Zhu, K. Fink, M. Endres, A. Srinivasan, K.J. Tomaselli, J. Yuan, M.A. Moskowitz, Activation and cleavage of caspase-3 in apoptosis induced by experimental cerebral ischemia, J. Neurosci. 18 (10) (1998) 3659 – 3668. A. Relo, J. Feldon, Specific neuronal protein: a new tool for histological evaluation of excitotoxic lesions, Physiol. Behav. 76 (2002) 449 – 456. H.B. Sarnat, D. Nochlin, D.E. Born, Neuronal nuclear antigen (NeuN): a marker of neuronal maturation in the early human fetal nervous system, Brain Dev. 20 (1998) 88 – 94. T. Sugawara, A. Lewen, N. Noshita, Y. Gasche, H.P. Chan, Effects of global ischemia duration on neuronal, astroglial, oligodendroglial and microglial reactions in the vulnerable hippocampal CA1 subregion in rats, J. Neurotrauma 19 (1) (2002) 85 – 98. J.J. Velier, J.A. Ellison, K.K. Kikly, A.P. Spera, F.C. Barone, G.Z. Feuerstein, Caspase-8 and caspase-3 are expressed by different populations of cortical neurons undergoing delayed cell death after focal stroke in the rat, J. Neurosci. 19 (14) (1999) 5932 – 5941. G.P. Xu, K.R. Dave, R. Vivero, R. Schmidt-Kastner, J.T. Sick, M. Perez-Pinzon, Improvement in neuronal survival after ischemic preconditioning in hippocampal slice cultures, Brain Res. 952 (2002) 153 – 158. C. Zhu, X. Wang, H. Hagberg, K. Blomerg, Correlation between caspase-3 activation and three different markers of DNA damage in neonatal cerebral hypoxia-ischemia, J. Neurochem. 75 (2) (2000 Aug) 819 – 829.