Proliferation dynamics of germinative zone cells in the intact and excitotoxically lesioned postnatal rat brain

Abstract

Background: The forebrain subventricular zone (SVZ)-olfactory bulb pathway and hippocampal subgranular zone (SGZ) generate neurons into adulthood in the mammalian brain. Neurogenesis increases after injury to the adult brain, but few studies examine the effect of injury on neural and glial precursors in the postnatal brain. To characterize the spatio-temporal dynamics of cell proliferation in the germinative zones, this study utilized a model of postnatal damage induced by NMDA injection in the right sensorimotor cortex at postnatal day 9. Dividing cell populations were labeled with 5-Bromodeoxyuridine (BrdU) in the intact and damaged postnatal brain. Identity of proliferating cells was determined by double immunolabeling with nestin, GFAP, NeuN and tomato lectin (TL).

Results: In the control brain, grouped BrdU+ cells were observed in the Rostral Migratory Stream (RMS), SVZ and SGZ. Maximal proliferation was seen at P12, persisted until P23 and diminished by P49. After injury, a striking reduction in the number of BrdU+ cells was observed in the ipsilateral SVZ from 10 hours (58% decrease) until 14 days post-lesion (88% decrease). In contrast, an increase in grouped BrdU+ cells was seen in the striatum adjacent to the depleted SVZ. Significantly reduced numbers of BrdU+ cells were also seen in the RMS until 3 days post-lesion. No changes were noted in the SGZ. Both in controls and lesioned hemispheres, BrdU+ cells located in the germinal zones were mostly nestin positive and negative for GFAP, NeuN, and TL. In the SVZ area lining the ventricle, BrdU+/nestin+ cells were mainly located between TL+ ependyma and parenchymal GFAP+ astrocytes. After excitotoxicity, a decrease in the number and orientation of GFAP/nestin+ prolongations leaving the SVZ to the cortex, corpus callosum and striatum was noted until 5 days post-lesion.

Conclusion: Postnatal excitotoxic injury differentially affects proliferating cells in the germinative zones: no change is observed in the dentate gyrus whereas excitotoxicity causes a significant decrease in proliferating cells in the SVZ and RMS. Depletion of BrdU+ cells in the postnatal SVZ and RMS differs from previous studies after adult brain injury and may affect the SVZ-RMS migration and is suggestive of progenitor recruitment to injured areas.

Figure 1

Camera lucida drawings of anterior coronal brain sections. Dots represent BrdU+ cells. No difference in number or spatial and temporal distribution of BrdU+ cells was seen in the contralateral and control hemispheres at corresponding time points (corresponding time points: 10 h = P9, 3d = P12, 5d = P14, 7d = P16, 14d = P23, 40d = P49). Ipsilateral (ip) hemispheres show BrdU+ cell distribution following NMDA-induced excitotoxicity in coronal brain sections at the level of the neurodegenerating area (outlined in black) and SVZ (BrdU+ cells of the SVZ are represented in red). Proliferation was observed at 10 h (P9), peaked at 3d (P12) and diminished by 40d (P49).

Figure 2

Temporal dynamics of proliferation in the germinative zones of control brains. Trends of cell proliferation in the postnatal rat brain, from P9 to P49, in the SVZ (a), RMS (b), and DG (c) are shown. Peak proliferation was always seen at P12. Curves were calculated using a polynomial regression line (order = 5, period = 2).

Figure 3

Quantitative analysis of excitotoxicity-induced changes in the germinative zones. Mean values of the number of BrdU+ cells in the SVZ (a), RMS (b), and DG (c) are shown and compared between intact controls, and lesioned (NMDA injected) contralateral (cl) and ipsilateral (ip) hemispheres. No difference was seen between intact controls and contralateral hemispheres at any time. In the SVZ, there is a significant decrease (asterisk) between the total number of BrdU+ cells in ip hemispheres when compared to cl and control hemispheres at all time points except 40 days post lesion. In the RMS, a significant decrease (asterisk) was seen at 10 hours and 3 days post lesion in NMDA-injected brains when compared to controls. In the DG, no significant difference in BrdU+ cell number was observed at any time point in ip hemispheres were compared to cl or control hemispheres.

Figure 4

Immunohistochemical staining for BrdU in the SVZ, RMS and SGZ. The significant decrease in BrdU+ cell number in the ipsilateral (ip) SVZ following NMDA-induced excitotoxicity is shown in a coronal brain section at 3 days post lesion (3d; compare a, b). BrdU+ cells at all time points in the ip SVZ showed a sharp decrease in number when compared to control hemispheres until 40 days post lesion (40d), the last time point studied, where no difference was seen (c, d). In the ip RMS, a decrease in BrdU+ cells was seen at 10 hours post lesion (10 h) and at 3 days post lesion (e) when compared to the corresponding control RMS (P9 and P12 brains, respectively; f). Thereafter, the ip RMS (7–40 days post lesion) when compared to controls (P16-P49, respectively), showed no difference in BrdU+ cell numbers (g, h). The dentate gyrus of control and NMDA-injected hemispheres showed no differences in cell number at any time point, as shown in a coronal section of the DG at 3 days post lesion (i, j). Note the hippocampal shrinkage in the ipsilateral hemisphere (j).

Figure 5

Phenotypes of proliferating cells in the SVZ. Optical microscope studies of coronal brain sections after double-labeled immunofluorescence showed that there is a characteristic placement of cell phenotypes along the entire length of the ventricular wall (vw). Photographs show TL+ ventricular ependymal cells (arrow, a) bordered BrdU+/nestin+ cells (arrows, b), which were found next to GFAP+ cells (arrow, c). Photographs of confocal imaging in the SVZ (more specifically, in zone a of the SVZ) revealed a similar placement where TL+ ependymal cells (d) of the ventricular wall are seen next to BrdU+/nestin+ progenitor cells (e; arrow, f) and GFAP+ cells (g). BrdU+/nestin+ cells outlined in white in photograph e can be seen at greater magnification in f. This patterning was seen at all time points except at postnatal day 9 (P9). At this time, in both control (P9) and lesioned brains (10 h), BrdU+/GFAP+ cells were also seen next to the ventricular wall (asterisk, h). At 10 hours post lesion, a decrease in the number of GFAP+ filaments in the ipsilateral (ip; i) hemisphere was also noted when compared to contralateral (cl; h) and control hemispheres. BrdU (red); nestin, TL, GFAP (green).

Figure 6

Nestin+ and GFAP+ Projections leaving the SVZ. In control brains at P12 and P14 GFAP+ nestin+ projections were seen leaving the SVZ extending towards the cortex (cx; arrows, a). After NMDA-induced excitotoxicity at 3 and 5 days post lesion, the presence of mature astrocytes with a star-like morphology were noticed next to the prolongations (b; asterisks, d). There was also a decrease observed in the number of prolongations (c, d) and orientation seemed disturbed such that many prolongations did not extend towards the cortex or were shorter (arrows, c) and did not reach the outer layers of the cortex (b-d). At 3 and 5 days post lesion note that the GFAP+ and nestin+ prolongations were also seen to extend into the corpus callosum (cc; e) and towards the striatum (see figure 5, b-c) in control brains at P12 and P14. BrdU (red); nestin, GFAP (green).

Figure 7

Double immunohistochemical staining of BrdU+ cell clusters in the striatum. In lesioned animals, from 3 days post lesion until 7 days post lesion grouped BrdU+ cells were found in the parenchyma a short distance away from the ventricular wall. Double labeling immunofluorescence studies followed by confocal imaging revealed that striatal BrdU+ cells colocalized with GFAP (arrow, a, b-c) and nestin (d, e-g). Within these groups it was possible to located cells with unipolar (arrow, a), bipolar (arrows, b), and typical mature astrocyte star-like morphology (b, c). Few BrdU+ cells were seen to colocalize with TL (h). BrdU (red); nestin, TL, GFAP (green).

Figure 8

Phenotypes of proliferating cells in the RMS and DG. Double-labeled immunofluoresence studies showed that in the RMS most cells were BrdU+/nestin+ (arrow, a) and revealed the presence of GFAP+ filaments (arrow, b) surrounding BrdU+ cells (asterisk, b). In the DG, BrdU+/nestin+ cells (c) were seen and a few BrdU+/GFAP+ cells could also be found (arrow, d, e). BrdU (red); nestin, GFAP (green).