Ischemia-stimulated neurogenesis is regulated by proliferation, migration, differentiation and caspase activation of hippocampal precursor cells

Brain Research 1058 (2005) 167 – 177 www.elsevier.com/locate/brainres

Research Report

Ischemia-stimulated neurogenesis is regulated by proliferation, migration, differentiation and caspase activation of hippocampal precursor cells Brendan Bingham, Danni Liu, Andrew Wood, Seongeun Cho* Neuroscience Discovery Research, Wyeth Research, CN 8000, Princeton, NJ 08543-8000, USA Accepted 29 July 2005 Available online 2 September 2005

Abstract A brief ischemic injury to the gerbil forebrain that caused selective damage in the CA1 region of the hippocampus also enhanced the production of new cells in the hippocampal neurogenic area. When evaluated 1 week after bromodeoxyuridine (BrdU) injection, approximately ten times more labeled cells were detected in the hippocampal dentate gyrus in ischemic animals than controls, indicating a stimulation of mitotic activity. To assess the temporal course of the survival and fate of these newborn cells, we monitored BrdU labeling and cell marker expression up to 60 days after ischemia (DAI). Loss of BrdU-positive cells was observed from both control and ischemic animals, but at 30 DAI and afterward, the ischemic group maintained more than 3 times as many BrdU-positive cells as the control group. In addition, ischemic injury also fostered the neuronal differentiation of these cells beyond the capacity observed in control animals and facilitated the migration of developing neurons to a neuronal cellular layer. The establishment of a temporal correlation between differentiation and migration provides evidence of the functional maturation of these cells. Surprisingly, we found that ischemic injury induced activation of caspase-3, not only in the CA1 region as expected, but also in the dentate subgranular zone (SGZ). Active caspase-3 immunoreactivity in the subgranular layer was co-localized with an early neuronal marker, suggesting that caspase-mediated apoptosis could mediate the loss of neurogenic cells in the SGZ. Inhibiting caspase-3 in the context of ischemia-induced neurogenesis might provide an opportunity for functional repair and a therapeutic outcome in the wake of ischemic injury. D 2005 Elsevier B.V. All rights reserved. Theme: Development and regeneration Topic: Cell differentiation and migration Keywords: Stroke; Caspase; Neuronal precursor cell; Survival; Beta-tubulin

1. Introduction In contradiction to the long-held belief that the cells of the adult mammalian brain cannot be regenerated after injury, recent findings have indicated not only the presence of stem cells in the brain throughout life, but also that neurogenesis is an ongoing process into adulthood [19,42]. Using several preclinical models of neuronal disease, it has been shown that brain injury, as well as some therapeutic regimens, can promote neurogenesis beyond the capacity

* Corresponding author. Fax: +1 732 274 4020. E-mail address: [email protected] (S. Cho). 0006-8993/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2005.07.075

present in naı¨ve animals [15,17,27,41,43]. In particular, acute ischemic stroke is known to increase the birth of new cells in the subgranular zone (SGZ) of the hippocampal dentate gyrus, as detected by increased incorporation of the thymidine analog bromodeoxyuridine (BrdU) [17,27]. Ischemia-induced stem cell division has been shown to be most prominent 7 to 10 days following a forebrain ischemic injury. Later, these newborn cells can mature, as evidenced by their expression of neuronal markers and their migration to the granule cell layer [27]. In addition, treatment with growth factors can markedly augment the neuroregenerative responses in the CA1 region by increasing cell proliferation and survival, resulting in replacement of cells in the damaged pyramidal neuronal layer [29]. The functional relevance of

168

B. Bingham et al. / Brain Research 1058 (2005) 167 – 177

this neuroregenerative response is supported by an improvement in memory-dependent cognitive performance in growth factor treated animals compared to controls [29]. Studies using high-resolution anatomical analysis have shown that newly developing neurons go on to receive synaptic inputs on their dendritic spines and to send axonal projections to their normal target cells in the brain [29,35,39]. Moreover, electrophysiological recordings and behavioral studies have confirmed the apparent normal functioning of these neurons [29]. These results suggest that neurons emerging in adulthood can incorporate into existing neuronal networks and contribute to functional synaptic circuitry, thus offering the possibility for CNS repair in neurodegenerative disease states. The birth of new neurons in adulthood has been shown to take place in discrete neurogenic regions of the brain, most notably in the hippocampus [14,24]. However, only a small fraction of newly emerged neurons survive. Most die before contributing to synaptic connections, essential for functional recovery, suggesting the existence of an endogenous regulation process that prevents massive and uncontrolled neuronal plasticity in the adult brain. Similar neuroanatomical remodeling is a normal event during early developmental stages, and the key molecular mechanisms regulating this process have been well characterized. During this period, active proliferation of embryonic stem cells is followed by programmed cell death, which takes place in a precisely controlled manner to form a well-organized, target-driven neural circuitry [34]. The cell death that occurs is believed to be largely apoptotic, involving typical morphological and biochemical features such as caspase activation and nuclear fragmentation. Caspases are a family of cysteine proteinases that play a central role in the control of apoptosis [12,28]. Activation of caspase-3 results in degradation of a number of intracellular and cytoskeletal protein substrates as part of an ordered process of cellular dissembly, leading to apoptotic cell death [7,32]. Caspase-3 expression in the CNS is developmentally regulated, with the highest levels of expression occurring in embryonic and neonatal brains, followed by an age-dependent decline in expression after birth [31]. Accordingly, caspase-3 knockout mice have abnormally large brains with an increased number of neurons [25], a result that reflects the important role of caspase-3 in the developmental refinement of brain circuitry. Although the factors that regulate neuromorphogenesis in adulthood are not well characterized, it is likely that some regulatory mechanisms involved during early development also continue to play important roles in shaping the mature brain. In adulthood, activation of neuronal caspase-3 is typically thought to be associated with apoptosis, and thus indicative of a neurodegenerative disease state. For example, in ischemic stroke, the activation of caspase-3 has been demonstrated in the damaged area of the brain by a variety of methods such as enzymatic activity assays, in situ hybridization, immunoblot analysis (both of caspases and

their substrates) and immunohistochemistry [4,5,30,44], and indeed, caspase-positive neurons are lost through apoptosis following stroke. Much less known, however, is the function of caspase activation in the survival and differentiation of newborn cells. Controlling the fate of ischemiastimulated progenitor cells may provide a novel path forward to the management of stroke and traumatic brain injury. Therefore, a deeper understanding is required of the dynamic interplay between the proliferative and apoptotic processes that take place in the brain following ischemic injury. In the current study, we compare the survival rate of newborn cells in ischemic and non-ischemic control animals and evaluate the effect of ischemia on the differentiation and migration of surviving cells at multiple timepoints postischemia. We demonstrate an increased proliferation of hippocampal precursor cells following ischemia, as well as the differentiation of these newborn cells to a neuronal phenotype and the migration of these cells to a neuronal location within the hippocampus. More important, we provide clear evidence of the activation of caspase 3 in these ischemia-induced cells, thus raising the possibility that removal of excess premature cells in the adult brain may be a caspase-dependent process.

2. Materials and methods 2.1. Surgical procedures and BrdU injection Male Mongolian gerbils (Harlan, USA) weighing 60– 70 g received a transient forebrain ischemic injury by bilateral common carotid artery (CCA) occlusion. The animals were initially anesthetized with 2.5 –3% isoflourane and maintained with 2% isoflurane during the surgical procedure. After a small midline incision was made in the neck, the common carotid arteries were isolated from the nerves and surrounding tissues, and surgical silk was loosely placed around them. Both arteries were occluded with aneurysm clips for 5 min. The clips were then removed to restore cerebral blood flow, and the neck was sutured for recovery. Sham-operated animals underwent the same surgical procedure, except the carotid arteries were not occluded. To label proliferating cells, BrdU injection was given ip. Based on our preliminary results and those of Liu et al. [27], animals were given either one set of BrdU injections (on day 8, three times, 4 h apart) or 4 sets of injections (on consecutive days starting on day 8, three times daily, 4 h apart) as indicated in the results (see Fig. 1 for schematic illustration of treatment paradigm). 2.2. Tissue preparation, Nissl staining, and BrdU immunohistochemistry Animals were anesthesized with Nembutal and transcardially perfused with 200 ml of 0.9% saline followed by

B. Bingham et al. / Brain Research 1058 (2005) 167 – 177

169

Fig. 1. Time line of BrdU injection and tissue collection. Animals were subjected to ischemic occlusion or sham operation for 5 min followed by reperfusion. (A) BrdU was injected three times, at 4-h intervals, 8 DAI, and animals were sacrificed 16 and 30 DAI. (B) For long-term studies on survival and differentiation, animals were injected with BrdU three times a day at 4-h intervals on 4 consecutive days (8 – 11 DAI) and sacrificed at 30, 45, and 60 DAI. Isch, ischemic injury; Sac, sacrifice.

ice-cold 4% paraformaldehyde in 0.1 M PBS, pH 7.4. Brains were removed, post-fixed (4 -C for 16 h) and transferred to PBS containing 30% sucrose (4 -C for 48 h). Coronal sections of 35 Am thickness were cut using a freezing sliding microtome (Microm International, Walldorf, Germany), and free-floating sections were processed for immunohistochemistry. For detection of BrdU-labeled nuclei, tissue samples were pretreated with 0.1 M citric acid, pH 6.0 at 75 -C for 30 min, followed by 2 N HCl in 1 PBS for 30 min at room temperature. Before incubation with a primary antibody, sections were blocked with 3% normal goat serum (NGS; Vector Labs) in 1 PBS at room temperature for 60 min. The primary antibody used was mouse anti-BrdU (Becton Dickinson, 1:150) in 0.25% Tween-20, incubated overnight at room temperature. After washes in 1 PBS, sections were incubated with biotinylated goat anti-mouse secondary antibody (Sigma) in 1.5% NGS for 60 min and placed in avidin –peroxidase complex solution (Vector Labs) for 1 h. The horseradish peroxidase reaction was detected with 0.05% diaminobenzidine and 0.03% H2O2. Following counter-staining with Nuclear Fast Red for 30 s, sections were dehydrated through graded alcohols, treated in Clear solution (Stephens Scientific), and mounted onto Superfrost Plus slides (VWR Scientific). Sections were examined with an inverted Zeiss microscope (Carl Zeiss, Germany). BrdU-positive cells were counted blindly in 6 coronal sections of 35 Am thickness spaced 210 Am apart. All counting was performed at 40 magnification, and results were calculated as the average number of BrdU-positive cells per section and expressed as mean T SEM. Data were analyzed by Student’s t test, and P values were calculated. Separate brain cryosections were processed for Nissl staining to visualize the ischemia-induced neuronal loss as described in Ohsaki et al. [33]. 2.3. Double- or triple-label immunofluorescence staining Tissue sections were blocked in 3% NGS, 1% bovine serum albumin, and 0.05% Triton X-100 in PBS for 1 h at room temperature. For primary reactions, sections were

incubated with rat anti-BrdU 1:150 (Oxford Biotechology), mouse anti-NeuN 1:50 (Chemicon), rabbit anti-GFAP 1:100 (Sigma), rabbit-anti-Tuj1 1:50 (Covance), and/or rabbit-anti active Caspase-3 1:50 (Cell Signaling), as indicated in figure legends, in 0.05% Triton X-100 overnight at 4 -C. After washing steps, sections were incubated for 1 h with a secondary antibody [Alexa 594-conjugated goat anti-rabbit 1:500, Alexa 488-conjugated goat anti-mouse 1:1000, or Alexa 350-conjugated goat anti-rabbit 1:2000 (Molecular probes)] in 1.5% NGS, and the resulting immunofluorescent signal was viewed under an Axiovert 135 fluorescence microscope (Carl Zeiss, Germany). 2.4. Image acquisition and data analysis Triple labeled (Brd-U, NeuN, and GFAP) brain sections were analyzed by confocal microscopy. Laser scanning of the dentate gyrus of the hippocampus was performed at 40 magnification on a Leica DM IRBE microscope, using the Leica Confocal Software package to set scanning parameters. Confocal images within the brain section were collected at approximately 1 Am intervals. Twelve to 20 stacked confocal images were typically assembled for each microscopic field collected from within the 35-Am sections. Stacked images were analyzed with the aid of Imaris software (Bitplane AG) to determine the presence of BrdU, NeuN, and GFAP staining. Data from approximately 100 BrdU-positive cells were collected from the hippocampus of each animal in the study. For each BrdU-positive cell imaged, the presence or absence of NeuN and GFAP staining was noted, and each cell was assigned to one of four categories: (1) NeuN-positive, GFAP-negative; (2) NeuN-negative, GFAP-positive; (3) NeuN-negative, GFAP-negative; or (4) NeuN-positive, GFAP-positive. Since the frequency of GFAP-positive cells was low (ranging from 2 to 10% per treatment group), for much of the analysis, the above categories were collapsed to two: NeuN-positive, and NeuN-negative. Within each timepoint (30, 45, and 60 DAI), ischemic animals were compared with non-ischemic controls for frequency of NeuN-positivity within the subset of BrdU-positive cells. Data were expressed

170

B. Bingham et al. / Brain Research 1058 (2005) 167 – 177

as mean T SEM and analyzed by Student’s t test, and P values were calculated.

3. Results A 5-min occlusion of the common carotid arteries was employed throughout this study to bring about global forebrain ischemia in gerbils. Nissl staining 7 days postinjury was used to verify the efficacy of this procedure. As expected, the ischemic injury resulted in the selective loss of

pyramidal neurons in the CA1 region of the hippocampus, as illustrated by a representative sample in Fig. 2A. In addition to promoting CA1 cell death, ischemia also increased proliferation of neuronal stem cells in discrete regions of the brain, including the subgranular zone of the dentate gyrus. To label mitotically active cells, BrdU was injected 8 DAI, at a time when ischemia-induced neurogenesis is reported to be active. BrdU exposure during this early post-ischemia timeframe establishes a population of labeled cells whose fate can be tracked weeks, or even months, later. As illustrated by representative pictures in

Fig. 2. Ischemia-induced neuronal death and neurogenesis. (A) Coronal sections of brain samples from animals subjected to ischemia or sham-operation were stained with thionin to visualize viable hippocampal neurons 7 DAI. Neurons in the CA1 subfield of the hippocampus were selectively lost, compared to an intact neuronal layer shown in a sham-operated animal. (B) Representative images taken from animals injected with BrdU 8 days after being subjected to ischemia or sham operation and sacrificed 8 days later (16 DAI). (C) Representative images taken from animals subjected to ischemia, injected with BrdU on 4 consecutive days (8 – 11 DAI), and sacrificed 30 and 60 DAI. (D) Fluorescence images from a representative triple-stained hippocampal section 45 DAI showing BrdU (green), NeuN (red), and GFAP (blue) immunoreactivities; BrdU and GFAP labeling are shown both in isolation and overlain with NeuN. (E) Confocal images of a representative triple-stained hippocampal section 45 DAI showing BrdU (green), NeuN (red), and GFAP (blue) immunoreactivities. Left: gallery of nine alternating confocal slices. Right: three-dimensional reconstruction of a confocal slice (bottom row, center, at the left) with the cross-hairs centered on a BrdU-positive, NeuN-positive cell in the GCL.

B. Bingham et al. / Brain Research 1058 (2005) 167 – 177

171

Fig. 2 (continued).

Fig. 2B and analyzed for the full data set in Table 1, ischemic hippocampal samples at 16 DAI had significantly more BrdU-positive cells than the non-ischemic samples, with the BrdU-labeled cells localized almost exclusively to the subgranular zone of the dentate gyrus. By 30 and 60 DAI (Fig. 2C), BrdU-positive cells in ischemic animals were located mainly in the granular cell layer (GCL) and had nuclei with a uniform oval shape, typical of granular neurons. In comparing the representative images in Figs. 2B and C, note that animals depicted in Fig. 2B received only one set of BrdU injections (at 8 DAI), whereas animals in Fig. 2C received 4 sets of BrdU injections (at 8– 11 DAI). Coupling BrdU labeling with staining for a neuronal marker (NeuN) and a glial marker (GFAP) is illustrated in Fig. 2D (fluorescence microscopy) and 2E (confocal microscopy). Most BrdU and NeuN signal was found in Table 1 Number of BrdU-immunoreactive cells in the dentate gyrus of hippocampus at 16 and 30 days following ischemia

the GCL, whereas GFAP signal was detected mainly in the hilar region. The majority of BrdU-positive cells showed evidence of having acquired a neuronal phenotype as measured by co-expression of NeuN; quantification of these results (Tables 1 – 3) is described below in detail. To estimate the survival of newborn cells during a shortterm and long-term recovery period, the number of hippocampal BrdU-positive cells was analyzed 8 and 22 days after cell labeling (16 and 30, DAI, respectively). As outlined in Table 1, at 16 DAI, the dentate gyrus of ischemic animals contained a total of 632 T 22 BrdU-positive cells whereas non-ischemic control animals had 65 T 7.8 cells (a statistically significant 9.7-fold increase). At 30 DAI, both the ischemic group and the non-ischemic control group had far fewer labeled cells, but the ischemic animals maintained 3.4 times more BrdU-positive cells than controls (68.7 T 5.7 vs. 20 T 1.5, P < 0.001). For an examination of the survival

Days after ischemia (DAI)

Control (n = 5)

Ischemia (n = 5)

Ratio (Isch: Cont)

Table 2 Time course of the number of BrdU-immunoreactive cells in the dentate gyrus of hippocampus at 30, 45, and 60 days following ischemia

16 DAI 30 DAI

65.06 T 7.81 20.15 T 1.53

632.4 T 22.37 68.7 T 5.70

9.72 3.41

Days after ischemia (DAI)

Control

Ischemic

Ratio (Ischemic: Control)

30 DAI 45 DAI 60 DAI

34.9 T 2.6 35.3 T 3.2 25.8 T 1.8

107.9 T 6.3 129.4 T 24 85.4 T 7.7

3.09 3.65 3.31

All numbers are mean T SEM (number of BrdU-labeled cells per brain slice). Fold changes represent the ratio of BrdU positive cells in ischemic brain over the non-ischemic controls. A statistically significant difference was found between the ischemic and non-ischemic groups at both time points ( P < 0.001).

All numbers are mean T SEM (number of hippocampal BrdU-labeled cells per brain slice examined). N = 5 for all groups.

172

B. Bingham et al. / Brain Research 1058 (2005) 167 – 177

Table 3 Percentage of BrdU-immunoreactive cells that are also positive for NeuN, as detected by confocal microscopy Days after ischemia (DAI)

Control

Ischemia

P values (Isch vs. Cont)

30 DAI 45 DAI 60 DAI

68.26 T 3.78 76.71 T 2.0 70.61 T 4.96

84.30 T 1.68 94.56 T 1.18 84.18 T 3.04

0.0064 0.0001 0.0485

All numbers are mean T SEM (Percentage of BrdU-positive cells costaining for NeuN, based on counting a minimum of 100 BrdU-positive cells per hippocampal brain slice). P values were calculated by comparing ischemic group with the non-ischemic controls. N = 4 for the 30- and 45day control groups; = 5 for the 60-day control group and all ischemic groups.

of the newborn cells up to 60 DAI, animals were injected with BrdU on 4 consecutive days (8 –11 DAI). At 30 DAI, the ischemic group had 108 T 6 BrdU-positive cells, whereas the control group had 35 T 2.6 cells (Table 2). As expected, more BrdU-positive cells were present than was the case with one set of BrdU injections (30 DAI, Table 1 vs. 30 DAI, Table 2), but the difference between ischemic and control animals was similar in the two studies (3.4-fold vs. 3.1-fold). At 45 and 60 DAI, while there were fewer

BrdU-positive cells present than at 30 DAI, the difference between the ischemic and control groups remained essentially unaltered (Table 2). To assess neuronal differentiation of surviving newborn cells, brain sections from ischemic and non-ischemic animals were triple immunolabeled with BrdU, NeuN, and GFAP, and the co-localization of BrdU signals with these cell type markers was evaluated using laser scanning confocal microscopy. Since very few BrdU-positive cells showed immunoreactivity for GFAP, whether from ischemic or non-ischemic animals (Fig. 3), much of the analysis was limited to the issue of NeuN and BrdU co-localization (Table 3). In control animals, a majority of BrdU-positive cells also showed NeuN immunoreactivity (an average of 72% across the 30, 45, and 60 DAI samples). Following ischemia, there was not only an increase in the total number of newborn cells (Table 2), but also a statistically significant increase in the proportion of BrdU-positive cells co-stained with NeuN (84 – 95%), as summarized in Table 3. We found that there were no significant differences in the number of GFAP-positive cells between control and the ischemic groups. However, as shown in Fig. 3, the number of BrdU-positive cells that were stained with neither NeuN nor

Fig. 3. Confocal analysis cell marker immunoreactivity of BrdU-positive cells. Animals were injected with BrdU at 8 – 11 days following ischemia or shamoperation, and BrdU-positive hippocampal cells were analyzed for staining with NeuN and GFAP at 30, 45, and 60 DAI using confocal microscopy. For each time point, the proportions of NeuN+/GFAP (BrdU-immunoreactive cells positive for NeuN but not for GFAP), NeuN /GFAP+ (BrdU-immunoreactive cells positive for GFAP but not for NeuN), NeuN+/GFAP+ (BrdU-immunoreactive cells positive for both NeuN and GFAP), and NeuN /GFAP (BrdUimmunoreactive cells that are negative for NeuN and GFAP) are graphed. N = 4 for 30- and 45 DAI control groups; = 5 for 60 DAI control group and all ischemic groups.

B. Bingham et al. / Brain Research 1058 (2005) 167 – 177

GFAP was far fewer in ischemic animals (4 –14%) than in the non-ischemic control group (22 –24% of the total). These data suggest that while newborn cells might tend to differentiate toward a neuronal lineage even in the absence of ischemic insult, ischemic injury further primes these cells toward a neuronal fate without imposing a dramatic effect on the likelihood of differentiation to a glial phenotype. As progenitor cells become differentiated and mature, they migrate away from the SGZ to the GCL or the hilar region. To examine the temporal correlation of phenotypic expression of protein markers with cellular migration, BrdU-positive cells were analyzed for their localization at 30, 45, and 60 DAI (Fig. 4). In both ischemic animals and non-ischemic controls, we found a time-dependent increase in the proportion of cells in the GCL with a complementary decrease in the proportion of cells in the SGZ. At 30 and 45 DAI, there were no apparent differences between controls and ischemic animals, but at 60 DAI, ischemic animals had accumulated a greater proportion of BrdU-positive cells in the GCL than controls (72% vs. 48%) with a complementary depletion of BrdU-positive cells in the SGZ. Transient forebrain ischemia leads to prominent neuronal death in the CA1 area, while sparing granular neurons in the dentate gyrus (Fig. 2A). However, while examining ischemia-induced activation of caspase-3 in CA1 pyramidal neurons, we discovered that cells in the SGZ were also strongly positive for active caspase-3 immunoreactivity. The signal was evident at 6 and 12 h post-ischemia and returned to

173

the basal level by 24 h (see Fig. 5A for a representative image taken at 6 h post-ischemia; see also Fig. 5 in Cho et al. [6]). To confirm that these active caspase-3-positive cells in the SGZ are of neuronal progenitor origin, we examined their expression of beta-tubulin, an early marker expressed in dividing neuronal precursor cells. Results from dual immunostaining as illustrated in Fig. 5B indicated clear evidence of co-localization of beta-tubulin and caspase-3 immunoreactivity. However, not all caspase-positive cells were betatubulin-positive (¨30% co-localization), possibly due to the heterogeneous nature of the population of precursor cells at different developmental stages. We also attempted to quantify the co-localization of caspase-3 immunoreactivity with BrdU signal, but few doubly labeled cells could be found, since the optimal timing for detection of caspase-3 activation is within hours following ischemia, which is a time when very few BrdU-labeled cells can be detected. Nonetheless, these data suggest that ischemia may have induced apoptotic cell death of neurogenic stem cells in the SGZ of the hippocampus via a caspase-dependent mechanism.

4. Discussion The developing CNS is well known to regulate morphogenesis by apoptosis and neurogenesis. While apoptosis has long been known to occur in the adult brain, recent studies have detected neurogenesis as well, suggesting that similar

Fig. 4. Localization of BrdU-positive cells within the dentate gyrus following ischemia. Animals were injected with BrdU at 8 – 11 days following ischemia or sham-operation, and the locations of BrdU-immunoreactive cells were assessed under fluorescence microscopy 30, 45, and 60 DAI. For each time point, the proportion of BrdU-positive cells (mean T SEM) is reported for the granular cell layer (GCL), the subgranular zone (SGZ), and the hilar region. N = 4 for 30and 45 DAI control groups; = 5 for 60 DAI control group and all ischemic groups.

174

B. Bingham et al. / Brain Research 1058 (2005) 167 – 177

Fig. 5. Co-localization of immunoreactivity of caspase-3 and TuJ1 in the dentate gyrus following ischemia. Representative images taken from samples collected 6 h following global ischemic injury. Coronal sections containing the hippocampus were (A) stained with anti-active caspase-3 or (B) double-labeled with antiactive caspase-3 (red) and anti-beta tubulin (TuJ1, green). In images (A), the entire dentate gyrus is shown. In images (B), only the subgranular zone is shown.

processes regulate neuronal cell number and plasticity in adults as in perinatal animals. Notably, in hippocampal regions, where neuronal cell turnover is robust, precursor cells in the dentate gyrus SGZ appear to proliferate continuously and migrate into the granule cell layer, ultimately differentiating into hippocampal granule cell neurons in both rodents and primates [1,15,23]. While several types of brain injuries, including epilepsy, kainateinduced excitotoxicity, and ischemia, have been shown to reduce neuronal cell numbers as a consequence of cell death, injury-generated stimuli also appear to up-regulate the synthesis of new neurons. For example, significant stimulation of neurogenesis has been demonstrated in the adult forebrain and hippocampus following ischemic insults in rat, mouse, and gerbil models [3,16,17,27,35,36]. Post-stroke cognitive recovery has also been found to be impaired in animals in which the ischemia-stimulated neurogenesis has been attenuated by radiation treatment, further suggesting that neurogenesis may play a critical role in stroke recovery [38]. While the possible long-term benefit of injury-induced neuromorphogenesis bears further investigation, the findings that some stroke victims undergo a substantial spontaneous recovery from ischemic injury suggest the functional operation of neuromorphogenic processes that at least in some circumstances can provide a repair capacity sufficient to compensate for the significant brain dysfunction that results from ischemic injury. A clearer understanding of cell survival, migration, and phenotypic fate, processes that

together govern the final outcome of regenerative repair, may lead to the ability to manipulate these processes for therapeutic benefit. In the present study, we have compared genesis, survival, differentiation, and migration of neuronal progenitors in control and ischemically treated gerbils and have also demonstrated caspase 3 activation in these cells. Substantial numbers of recently divided cells were detected in the dentate gyrus in control animals, indicating that neurogenesis is an ongoing process during adulthood. Compared with controls, ischemia stimulated animals displayed a large increase in the rate of production of new cells as indicated by BrdU labeling. Tracking these BrdU-labeled cells in the weeks post-ischemia, we found a decrease in their numbers over time, which is consistent with loss due to cell death. However, alternative explanations, such as the dilution of BrdU label through repeated rounds of mitosis and the migration of labeled cells outside the field of detection are also possibilities. Irrespective of the reason for the decrease in the numbers of labeled cells, those remaining beyond the 30-day post-ischemia time point were detected as having localized to the granular neuronal layer. Despite the fact that newborn cells from several brain regions, including the hippocampus, appear to have the capacity to form both neuronal and astroglial lineages [20], the vast majority of the surviving cells in our studies had differentiated toward a neuronal phenotype, as evidenced by NeuN/BrdU dualpositivity in the ischemic as well as control groups. These

B. Bingham et al. / Brain Research 1058 (2005) 167 – 177

data were generated using confocal imaging, which allows the assessment of individual cell phenotypes within a brain slice, without the requirement of removing those cells from their three-dimensional context. Furthermore, that a greater proportion of BrdU-positive cells tested positive for NeuN in ischemic animals vs. controls at all timepoints indicates that ischemic injury contributes to the neuronal differentiation of hippocampal stem cells. This increase in the proportion of NeuN positive cells came about not due to a decrease in the proportion of BrdU/GFAP positive cells, whose numbers remained relatively constant, especially at the 45- and 60-DAI timepoints, but rather by a decrease in the proportion of NeuN-negative/GFAP-negative cells. This finding suggests that stimuli generated as a result of ischemic insult can promote a neurogenic fate by increasing the differentiation of non-committed precursor cells to a neuronal lineage. Of importance to note, some molecular events known to affect neurogenesis and neuronal differentiation are also involved in stroke pathology. For example, activation of glutamate receptors by low concentrations of NMDA is known to induce neurogenesis, and exogenous inflammatory stimuli have been shown to down regulate the numbers of new neurons differentiating in response to ischemia [11]. More studies are required to develop an understanding of the sophisticated biochemical interplay controlling cell death and neuronal replacement in order to promote the selective augmentation of the desired responses following ischemic injury. Although we have not attempted to ascertain whether ischemia-induced neurons can become functionally integrated into brain circuitry, our findings are consistent with this eventual outcome. By tracking neuronal and glial markers at multiple time points up to two months postischemia, we have demonstrated that ischemic injury can prime the fate of previously uncommitted progenitor cells. Moreover, we have established a temporal correlation between the differentiation of newborn cells toward a neuronal fate and their migration to the GCL, which is a further indicator of their functional maturation. It was previously shown that highly polysialylated neuronal cell adhesion molecule (NCAM-H) is specifically expressed during the migration stage of stem cell development [40]. In addition, expression of polysialic acid (PSA) promotes migration of neural progenitor cells out of the subventricular zone (SVZ) and orchestrates the temporal regulation inducing differentiation of these cells to a neuronal phenotype [37]. Therefore, it will be interesting to examine whether and how the increased migration of precursor cells into the GCL by ischemic injury that we see in our study accompanies the stimulation of PSA expression. In the well-documented neurogenic response to brain injury, neuroblast proliferation is triggered in damaged and non-damaged areas [9]. Why non-damaged regions react to injury by upregulating neurogenesis is not yet understood, but several studies provide evidence for the formation of new neurons that mature in a phenotypically appropriate

175

manner and persist to replace neurons lost from injured tissue. For example, in the rat, neurogenesis can be stimulated following intrastriatal infusion of quinilinic acid, producing new DARPP-32 positive medium spiny neurons and parvalbumin positive interneurons, phenotypes that are specifically lost during the injury [8]. Forebrain SVZ neurogenesis can also be stimulated by ischemic insult to increase the contribution of new cells and the migration of neuroblasts to peri-infarcted striatum and cortical sites of injury; these cells appear to persist as medium spiny neurons in the neostriatum [2,18,36]. Although neurogenesis reduces detectably as animals age, the persistence of basal neurogenesis in both adult rodent and primate brains and their specific contribution to damaged regions that normally are neurogenically quiescent provides support for the activation of appropriate endogenous repair mechanisms in the adult [26]. Functional neuronal replacement has also been shown in a study where the normal capacity of endogenous stem cell recovery post-ischemia was enhanced by administration of growth factors. EGF and FGF, administered by intraventricular infusion, were shown to produce regeneration of hippocampal CA1 projection neurons by recruitment from the periventricular region adjacent to the hippocampus [30]. The observation that these treatments also ameliorated the ischemia-induced deficit in hippocampal cognitive functions further supports a therapeutic strategy involving manipulation of progenitor cells to enhance brain repair and functional recovery. While the loss of new neurons during development may be a normal process, in the adult and aging brain reducing attrition of recently born neurons will likely be critical to achieving meaningful repair. Optimizing trophic support and cell survival of newly integrating cells may therefore enhance the ability of this intrinsic mechanism to sustain more cells and enhance functional regeneration. As noted above, despite a surge of new cell birth following ischemia, we found that the majority of these newborn cells do not survive longer than 3 weeks as tracked by BrdU-labeling. We proposed that such a decrease in the number of BrdU labeled cells may be due to apoptosismediated cell death, and thus investigated whether cells in the neurogenic SGZ could be counterstained with an antibody recognizing active caspase-3. Coupled with the identification of neuronal precursor cells by Tuj1 immunostaining to detect beta-tubulin expression, we observed a substantial overlap of the expression of beta-tubulin and active caspase-3 following ischemia, a result that is suggestive of these early stage neurons having initiated the process of apoptosis. Activation of caspase 3 is a hallmark of apoptotic cell death and plays a key role in many neuropathological conditions, including stroke, spinal cord injury, and epilepsy. In various animal models of neurodegeneration, treatment with caspase inhibitors has been shown to prevent neuronal apoptosis, thus providing therapeutic potential [13,21,22]. Our observation that caspase-3 is activated in the dentate subgranular layer follow-

176

B. Bingham et al. / Brain Research 1058 (2005) 167 – 177

ing ischemic injury lends additional therapeutic importance in that it raises the possibility that modulation of the survival parameters of neuronal progenitor cells could provide a significant increase in the number of new neurons available for regeneration. Caspase inhibitors have been reported to modulate seizure-induced neurogenesis in the dentate gyrus by decreasing the number of terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL)positive cells and by increasing the number of BrdU positive cells 1 week post-injury [10]. However, in that study, the caspase inhibitor was not administered beyond the first week following injury and thus the survival promoting effects did not last long enough to translate into a repopulation of the damaged dentate gyrus. In addition, it is possible that additional growth cues might be required to promote long-term survival of these newborn neurons. Since modulating neurogenesis by increasing proliferation may carry more risk than preventing the death of existing cells, a greater therapeutic potential for functional repair of the injured brain may be achieved by providing the conditions that promote the survival of those newly emerging cells following an ischemic injury, such as by inhibiting activated caspase-3 and promoting the propagation of appropriate trophic signals.

Acknowledgments The authors wish to recognize the technical assistance provided by Dr. Jessica E. Malberg, Dr. Myles Fennell, and Ms. Smita Kotnis. In addition, we thank Dr. Jeffrey D. Kennedy for a critical reading of the manuscript.

References [1] J. Altman, G.D. Das, Autoradiographic and histological evidence of postnatal hippocampal neurogenesis in rats, J. Comp. Neurol. 124 (1965) 319 – 335. [2] A. Arvidsson, T. Collin, D. Kirik, Z. Kokaia, O. Lindvall, Neuronal replacement from endogenous precursors in the adult brain after stroke, Nat. Med. 8 (2002) 963 – 970. [3] A. Arvidsson, Z. Kokaia, O. Lindvall, N-methyl-d-aspartate receptormediated increase of neurogenesis in adult rat dentate gyrus following stroke, Eur. J. Neurosci. 14 (2001) 10 – 18. [4] A. Benchoua, C. Guegan, C. Couriaud, H. Hosseini, N. Sampaio, D. Morin, B. Onteniente, Specific caspase pathways are activated in the two stages of cerebral infarction, J. Neurosci. 21 (2001) 7127 – 7134. [5] J. Chen, T. Nagayama, K. Jin, R.A. Stetler, R.L. Zhu, S.H. Graham, R.P. Simon, Induction of caspase-3-like protease may mediate delayed neuronal death in the hippocampus after transient cerebral ischemia, J. Neurosci. 18 (1998) 4914 – 4928. [6] S. Cho, D. Liu, C. Gonzales, M.M. Zaleska, A. Wood, Temporal assessment of caspase activation in experimental models of focal and global ischemia, Brain Res. 982 (2003) 146 – 155. [7] G.M. Cohen, Caspases: the executioners of apoptosis, Biochem. J. 326 (Pt. 1) (1997) 1 – 16. [8] T. Collin, A. Arvidsson, Z. Kokaia, O. Lindvall, Quantitative analysis of the generation of different striatal neuronal subtypes in the adult brain following excitotoxic injury, Exp. Neurol. 195 (2005) 71 – 80.

[9] V. Darsalia, U. Heldmann, O. Lindvall, Z. Kokaia, Stroke-induced neurogenesis in aged brain, Stroke 36 (2005) 1790 – 1795. [10] C.T. Ekdahl, P. Mohapel, E. Elmer, O. Lindvall, Caspase inhibitors increase short-term survival of progenitor-cell progeny in the adult rat dentate gyrus following status epilepticus, Eur. J. Neurosci. 14 (2001) 937 – 945. [11] C.T. Ekdahl, J.H. Claasen, S. Bonde, Z. Kokaia, O. Lindvall, Inflammation is detrimental for neurogenesis in adult brain, Proc. Natl. Acad. Sci. U. S. A. 100 (2003) 13632 – 13637. [12] B.A. Eldadah, A.I. Faden, Caspase pathways, neuronal apoptosis, and CNS injury, J. Neurotrama 17 (2000) 811 – 829. [13] M. Endres, S. Namura, M. Shimizu-Sasamata, C. Waeber, L. Zhang, T. Gomez-Isla, B. Hyman, M. Moskowitz, Attenuation of delayed neuronal death after mild focal ischemia in mice by inhibition of the caspase family, J. Cereb. Blood Flow Metab. 18 (1998) 238 – 247. [14] P.S. Eriksson, E. Perfilieva, T. Bjork-Eriksson, A.M. Alborn, C. Nordborg, D.A. Peterson, F.H. Gage, Neurogenesis in the adult human hippocampus, Nat. Med. 4 (1998) 1313 – 1317. [15] E. Gould, P. Tanapat, Lesion-induced proliferation of neuronal progenitors in the dentate gyrus of the adult rat, Neuroscience 80 (1997) 427 – 436. [16] W.P. Gray, L.E. Sundstrom, Kainic acid increases the proliferation of granule cell progenitors in the dentate gyrus of the adult rat, Brain Res. 790 (1998) 52 – 59. [17] K. Jin, M. Minami, J.Q. Lan, X.O. Mao, S. Batteur, R.P. Simon, D.A. Greenberg, Neurogenesis in dentate subgranular zone and rostral subventricular zone after focal cerebral ischemia in the rat, Proc. Natl. Acad. Sci. U. S. A. 98 (2001) 4710 – 4715. [18] K. Jin, Y. Sun, L. Xie, A. Peel, X.O. Mao, S. Batteur, D.A. Greenberg, Directed migration of neuronal precursors into the ischemic cerebral cortex and striatum, Mol. Cell. Neurosci. 24 (2003) 171 – 189. [19] C.B. Johansson, S. Momma, D.L. Clarke, M. Risling, U. Lendahl, J. Frisen, Identification of a neural stem cell in the adult mammalian central nervous system, Cell 96 (1999) 25 – 34. [20] F.P. Jori, U. Galderisi, E. Piegari, M. Cipollaro, A. Cascino, G. Peluso, R. Cotrufo, A. Giordano, M.A. Melone, EGF-responsive rat neural stem cells: molecular follow-up of neuron and astrocyte differentiation in vitro, J. Cell. Physiol. 195 (2003) 220 – 233. [21] S. Knoblach, X. Huang, J. VanGelderen, D. Calva-Cerqueira, A. Faden, Selective caspase activation may contribute to neurological dysfunction after experimental spinal cord trauma, J. Neurosci. Res. 80 (2005) 369 – 380. [22] A. Kondratyev, K. Gale, Intracerebral injection of caspase-3 inhibitor prevents neuronal apoptosis after kainic acid-evoked status epilepticus, Brain Res. Mol. Brain Res. 75 (2000) 216 – 224. [23] D.R. Kornack, P. Rakic, Continuation of neurogenesis in the hippocampus of the adult macaque monkey, Proc. Natl. Acad. Sci. U. S. A. 96 (1999) 5768 – 5773. [24] H. Kuhn, T. Palmer, E. Fuchs, Adult neurogenesis: a compensatory mechanism for neuronal damage, Eur. Arch. Psychiatry Clin. Neurosci. 251 (2001) 152 – 158. [25] K. Kuida, T.S. Zheng, S. Na, C. Kuan, D. Yang, H. Karasuyama, P. Rakic, R.A. Flavell, Decreased apoptosis in the brain and premature lethality in CPP32-deficient mice, Nature 384 (1996) 368 – 372. [26] R.J. Lichtenwalner, J.M. Parent, Adult neurogenesis and the ischemic forebrain, J. Cereb. Blood Flow Metab., in press. [27] J. Liu, K. Solway, R.O. Messing, F.R. Sharp, Increased neurogenesis in the dentate gyrus after transient global ischemia in gerbils, J. Neurosci. 18 (1998) 7768 – 7778. [28] N. Marks, M.J. Berg, Recent advances on neuronal caspases in development and neurodegeneration, Neurochem. Int. 35 (1999) 195 – 220. [29] H. Nakatomi, T. Kuriu, S. Okabe, S. Yamamoto, O. Hatano, N. Kawahara, A. Tamura, T. Kirino, M. Nakafuku, Regeneration of hippocampal pyramidal neurons after ischemic brain injury by recruitment of endogenous neural progenitors, Cell 110 (2002) 429 – 441.

B. Bingham et al. / Brain Research 1058 (2005) 167 – 177 [30] 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 (1998) 3659 – 3668. [31] B. Ni, X. Wu, Y. Su, D. Stephenson, E. Smalstig, J. Clemens, S. Paul, Transient global forebrain ischemia induces a prolonged expression of the caspase-3 mRNA in rat hippocampal CA1 pyramidal neurons, J. Cereb. Blood Flow Metab. 18 (1998) 248 – 256. [32] D.W. Nicholson, N.A. Thornberry, Caspases: killer proteases, Trends Biochem. Sci. 22 (1997) 299 – 306. [33] K. Ohsaki, N. Osumi, S. Nakamura, Altered whisker patterns induced by ectopic expression of Shh are topographically represented by barrels, Brain Res. Dev. Brain Res. 137 (2002) 159 – 170. [34] R.W. Oppenheim, D. Prevette, M. Tytell, S. Homma, Naturally occurring and induced neuronal death in the chick embryo in vivo requires protein and RNA synthesis: evidence for the role of cell death genes, Dev. Biol. 138 (1990) 104 – 113. [35] J.M. Parent, T.W. Yu, R.T. Leibowitz, D.H. Geschwind, R.S. Sloviter, D.H. Lowenstein, Dentate granule cell neurogenesis is increased by seizures and contributes to aberrant network reorganization in the adult rat hippocampus, J. Neurosci. 17 (1997) 3727 – 3738. [36] J.M. Parent, Z.S. Vexler, C. Gong, N. Derugin, D.M. Ferriero, Rat forebrain neurogenesis and striatal neuron replacement after focal stroke, Ann. Neurol. 52 (2002) 802 – 813. [37] A. Petridis, A. El-Maarouf, U. Rutishauser, Polysialic acid regulates cell contact-dependent neuronal differentiation of progenitor cells from the subventricular zone, Dev. Dyn. 230 (2004) 675 – 684.

177

[38] J. Raber, Y. Fan, Y. Matsumori, Z. Liu, P.R. Weinstein, J.R. Fike, J. Liu, Irradiation attenuates neurogenesis and exacerbates ischemiainduced deficits, Ann. Neurol. 55 (2004) 381 – 389. [39] T. Saegusa, S. Mine, H. Iwasa, H. Murai, T. Seki, A. Yamaura, S. Yuasa, Involvement of highly polysialylated neural cell adhesion molecule (PSA-NCAM)-positive granule cells in the amygdaloidkindling-induced sprouting of a hippocampal mossy fiber trajectory, Neurosci. Res. 48 (2004) 185 – 194. [40] T. Seki, Y. Arai, Highly polysialylated neural cell adhesion molecule (NCAM-H) is expressed by newly generated granule cells in the dentate gyrus of the adult rat, J. Neurosci. 13 (1993) 2351 – 2358. [41] E.Y. Snyder, C. Yoon, J.D. Flax, J.D. Macklis, Multipotent neural precursors can differentiate toward replacement of neurons undergoing targeted apoptotic degeneration in adult mouse neocortex, Proc. Natl. Acad. Sci. U. S. A. 94 (1997) 11663 – 11668. [42] H.J. Song, C.F. Stevens, F.H. Gage, Neural stem cells from adult hippocampus develop essential properties of functional CNS neurons, Nat. Neurosci. 5 (2002) 438 – 445. [43] S. Yoshimura, Y. Takagi, J. Harada, T. Teramoto, S.S. Thomas, C. Waeber, J.C. Bakowska, X.O. Breakefield, M.A. Moskowitz, FGF-2 regulation of neurogenesis in adult hippocampus after brain injury, Proc. Natl. Acad. Sci. U. S. A. 98 (2001) 5874 – 5879. [44] W.R. Zhang, K. Sato, M. Iwai, I. Nagano, Y. Manabe, K. Abe, Therapeutic time window of adenovirus-mediated GDNF gene transfer after transient middle cerebral artery occlusion in rat, Brain Res. 947 (2002) 140 – 145.