Subventricular zone localized irradiation affects the generation of proliferating neural precursor cells and the migration of neuroblasts

Subjects

The subjects were 6–8-week- old male C57BL6/J mice from the National Cancer Institute maintained on a 12:12 hour light: dark cycle. Food and water were available ad libitum. All experiments described were performed with the approval of the Johns Hopkins Animal Care and Use Committee under standard protocols.

CT-based Localized Brain Irradiation

Mice were anesthetized with the injection of 100mg/kg ketamine+10mg/kg xylazine intraperitoneally. To directly visualize the ventricles and target the SVZ, 70µL of iodine contrast was injected intrathecally and CT images were obtained as described previously [32]. A single dose of 10Gy was delivered using CT-based tissue visualization. Previous studies have demonstrated that the overall geometric targeting accuracy of this technique is 0.2mm [32]. A radiation beam of 3×3mm was used to target right SVZ while left brain structures served as controls. For migration studies, a radiation beam of 1 mm diameter was used to target the anterior dorsal region of the SVZ. Additionally, a radiation beam of 5×9mm was used to target right olfactory bulb (OB), rostral migratory stream (RMS), and/or anterior SVZ (aSVZ). For sham irradiation, control animals were anesthetized, brought into the treatment room, and handled and positioned identically to irradiated animals without radiation delivery.

Immunohistochemistry

Mice (n=5/group) were deeply anesthetized and perfused transcardially with 0.9% saline followed by 4% paraformaldehyde (PFA) in 0.1M phosphate-buffered saline (PBS). Brains were removed from the skull and post-fixed in 4% PFA overnight at 4°C. The brains were then equilibrated in 30% sucrose and were frozen in Tissue-Tek OCT Compound. 10µm thick coronal slices were sectioned using a cryostat. Sections were treated with 0.01M sodium citrate at 95°C for 10 min for antigen retrieval. Sections were incubated in PBS containing 0.1% Triton-X-100, and 10% normal goat serum (NGS) for 30min, and then incubated with primary antibodies for 16hrs at 4°C. Primary antibodies used in this study were mouse anti-rH2Ax (Ser139) (1:700), rabbit anti-Ki67 (1:200), mouse anti-GFAP (1:500), rabbit anti-doublecortin (DCx) (1:500), mouse anti-nestin (1:100), mouse anti Mash-1 (1:100), cleaved caspase-3 (1:100) and mouse anti-CD31 (1:200). More details on the primary antibodies are summarized in Supplementary Table 1. The sections were washed with PBS and incubated with secondary antibodies conjugated with fluorophores for 1 hour at room temperature (RT). Anti-mouse, anti-rat and anti-rabbit Alexa 488 and/or 594 secondary antibodies (1:500, Invitrogen) were used. Nuclei were stained with DAPI (1:500). All the sections were air-dried and coverslipped with Aquamount mounting media (Vector Labs). Fluorescent images were taken with an ORCA II camera (Hamamatsu) connected to an Olympus IX81 inverted microscope.

SVZ Whole Mount Dissection and Immunostaining

Mice (n=3/group) were sacrificed by cervical dislocation and their brains were immediately extracted into L-15 Leibovitz medium (Gibco). A whole mount of SVZ tissue (2–4mm) with the lateral ventricle and striatum was dissected under microscope, and the hippocampus and septum were removed as described [24, 33]. The SVZ whole mounts were fixed in 4% PFA/0.1% Triton-X-100 overnight at 4° C. They were then washed with PBS and blocked with 10% NGS/0.5% Triton for 30min at RT. The tissues were incubated in primary and secondary antibodies diluted in blocking buffer for 48hr at 4°C. After staining, the SVZ whole mounts were further dissected from underlying striatum to 200–300µm thick tissue, transferred to a glass slide and coverslipped with Aquamount mounting media for imaging. A Nikon C1si True Spectral Imaging Confocal Laser Scanning Microscope was used to image the SVZ whole mounts.

Semi-Thin sections, Electron microscopy and Ultrastructural Analysis

Mice (n=5/group) were anesthetized and perfused with 0.9% saline, followed by 2% paraformaldehyde and 2.5% glutaraldehyde. Heads were removed and post-fixed in the same fixative overnight. After post-fixation, brains were dissected and washed in 0.1M phosphate buffer (PB) (pH 7.4), cut into 200µm coronal sections with a VT 1000M vibratome (Leica, Germany), and treated with 2% osmium tetraoxide in 0.1M PB for 2 hr. Sections were then rinsed, dehydrated through increasing ethanol solutions and stained with 2% uranyl acetate in 70% ethanol. Following dehydration, slices were embedded in araldite (Durcupan, Fluka BioChemika). To study the cellular organization of the SVZ and RMS, we cut serial 1.5µm coronal semithin sections with a diamond knife and stained them with 1% Toluidine blue. Sections were visualized under E200 light microscope (NIKON, Japan). To identify cell types, 60–70nm ultrathin sections were cut with a diamond knife, stained with lead citrate, and examined under a Spirit transmission electron microscope (FEI Tecnai, USA). The different cell types in the SVZ and RMS were identified based on their ultrastructural characteristics as described previously [25].

SVZ Neurosphere Cultures

Adult mouse brains were removed into sterile cold Neurocult basal medium (Stem Cell Technologies). The SVZ was dissected out from each hemisphere as described in Ferron S.R. et al. [34]. The left and right SVZ were collected separately and pooled from four different mice. The tissue was triturated gently until all the pieces were dissociated and cell suspension was homogenous. The cells were washed in the complete mouse Neurocult proliferation medium with Neurocult proliferation supplement, recombinant growth factors (10µg/mL epidermal growth factor [EGF] and 10µg/mL fibroblast growth factor [FGF]) and 2µg/mL heparin (Stem Cell Technologies). Viable cells were counted and plated at 2×10⁴ cells/cm². The cultures were incubated at 37°C in a 5% CO₂ humidified incubator adding fresh medium every 4 days. Neurospheres of 100–200µm size were passaged and replated.

Retrovirus Preparation and Stereotactic Injections

Engineered self-inactivating murine oncoretrovirus expressing green fluorescent protein (GFP), with the Woodchuck hepatitis virus post-transcriptional regulatory element under the Ubiquitin-C promoter, was produced as previously described [35]. A day after irradiation, mice (n=5/group) were anesthetized with ketamine (100mg/kg)/xylazine (10mg/kg) anesthesia and injected with buprenorphine (0.02mg/kg) for post-operative analgesia. The head was secured to a stereotaxic frame (David Kopf). The skull surface was exposed and retrovirus was injected (1µl per site) into the SVZ using a glass needle at the following two different coordinates bilaterally relative to bregma: anterior=1.0mm, lateral=±1.0mm, ventral=2.2mm; anterior =0. 0mm, lateral=±1.8mm, ventral=1.6mm. The scalp was sutured and the mice were returned to their standard housing. Mice were sacrificed a week after retrovirus injections, the brains were fixed, sectioned sagittally and immunostained against Ki67. Sections were visualized under a microscope for GFP-expressing migrating neuroblasts and Ki67⁺ cells. The experimental paradigm is shown in Figure 7A.

Cell Quantification and Statistical Analysis

In the coronal sections of the mouse brain, the quantification of Ki67⁺ cells in the SVZ, RMS and OB was performed on every 20^th section evenly spaced along the entire rostral-caudal length of the lateral ventricle of the SVZ using a fluorescent microscope. A total of 7, 9, and 10 sections were counted for SVZ, RMS and OB respectively for each animal. The cell counts for all mice in a given group were averaged per section and the standard error of the mean (SEM) was calculated. To quantify the number of cells within the SVZ under electron microscope, we considered the area comprised of the first 20µm adjacent to the ventricle lumen (0–1mm anterior to bregma). The counts were performed on 3 different levels per animal, and the average was shown in cells/mm. Blood vessel analysis was performed using semithin sections. We considered the area comprised of the first 100µm adjacent to the ventricle lumen (0–1mm anterior to bregma). The number and size of blood vessels were measured with Image Tool software (Evans Technology). Changes in RMS size were determined by measuring the area occupied by RMS in semithin sections using a light microscope, at 3 different levels per animal. The analysis was performed with Image Tool software. An unpaired t-test was performed using SigmaPlot 11.0 software to determine the difference between the left and right SVZ for all our data quantified. The differences were considered significant at a p-value less than 0.05.

Localized radiation eliminates actively proliferating cells in the SVZ

A series of pilot experiments was performed to validate the localized nature of the radiation beams from the SARRP. A 10-Gy, 3×3mm field of localized radiation was delivered to the right ventricle while the non-irradiated left ventricle served as a control (Figure 1A). The accuracy of localized radiation delivery in the targeted right SVZ region was validated by immunohistochemical staining of γH2Ax, a marker of DNA double strand breaks, one hour post irradiation. The irradiated right ventricle of the mouse brain showed restricted expression of γH2Ax (Figure 1B). Cell proliferation in the SVZ was assessed by immunostaining against Ki67 in response to the localized radiation. A strong lateral asymmetry in cell proliferation was observed on every coronal section one-day post irradiation (Figure 1C) and on SVZ whole mounts one-month post irradiation (Figure 1D). There was a significant decrease in the number of Ki67⁺ cells in the right SVZ at one day (left SVZ: 85.05±12.84 cells/section vs. right SVZ: 14.27±2.35 cells/section) (Figure 1E) and one month (left SVZ: 56±5.26 cells/section vs. right SVZ: 7±1.67cells/section) (Figure 1F) after irradiation (p<0.001). Remarkably, the number of proliferating cells in the RMS and OB was not affected by localized radiation of the SVZ at either one day or one month after irradiation (Figure 1E–F). These results indicate that the vast majority of proliferating cells in the mouse SVZ were permanently eliminated in a highly localized fashion using SARRP technology.

Localized radiation of SVZ decreases NPCs and neuroblasts

We then explored the effect of localized radiation on different cell types in the SVZ, such as NSCs/type B1 cells, NPCs/type C cells, and migrating neuroblasts/type A cells, using cell specific markers and electron microscopy one-month post irradiation. Nestin and GFAP immunostaining was used to label type B cells, Mash-1 for type C cells, and doublecortin (DCx) for migrating neuroblasts [24]. While GFAP and nestin expression appeared to remain stable (Figure 2A, and 2C), there was a strong qualitative reduction of Mash-1⁺ cells (Figure 2B) and DCx⁺ cells (Figure 2C). The GFAP⁺ type B1 cells in the coronal sections were distinguished by their contact with the ventricular surface (arrowheads in Figure 2A) and long distal processes (arrows in Figure 2A) from the parenchymal astrocytes or type B2 cells that form glial tubes and are also positive for GFAP [24]. GFAP and DCx co-staining of lateral ventricle whole mounts further confirmed the presence of GFAP⁺ cells and the absence of DCx⁺ cells in the irradiated SVZ (Figure 2D). These findings from SVZ immunostaining were further confirmed with electron microscopy, and quantitative analysis of different cell types was performed using electron microscopy (Figure 3).

Light and electron microscope examination of the SVZ showed a clear disruption in both the antero-posterior and dorso-ventral axes in irradiated hemisphere one month after irradiation. While the cell organization in non-irradiated SVZ appeared normal, with a monolayer of ependymal cells and typical chains of migratory neuroblasts (dark cells) surrounded by type B (light cells), SVZ in irradiated hemisphere was reduced to ependymal cells with dispersed astrocytes and neuronal bodies close to the ventricle (Figure 3A and 3C). There were no migratory chains after irradiation (Figure 3A and 3C), but dispersed type A cells were occasionally observed. Consequently, cell density was drastically decreased in irradiated SVZ (left SVZ: 236.1±26.2 cells/mm vs. right SVZ: 97.9±6 cells/mm; p<0.005) (Figure 3B). This loss of cell density was attributed to a decrease in the number of type B cells (left SVZ: 76.8±7.2 cells/mm vs. right SVZ: 38.2±6.6 cells/mm; p=0.004), type C cells (left SVZ: 25.5±5.9 cells/mm vs. right SVZ: 0.5±0.3 cells/mm; p=0.01) and type A cells (left SVZ: 77.0±17.0 cells/mm vs. right SVZ: 3.5±0.6 cells/mm; p=0.01), while ependymal cells remained unaltered (left SVZ: 46.9±2.9 cells/mm vs. right SVZ: 46.8±2.3 cells/mm) (Figure 3D and Supplementary Table 2). Most importantly, type B1 cells, identified by their apical process extending into the lateral ventricle and primary cilia, were observed in the right SVZ one month after irradiation (Figure 3E-E’, and Table 2). Extensive portions of basal lamina were found frequently between astrocytes and/or ependymal cells in the irradiated SVZ, compare to non-irradiated hemisphere (Figure 3F). As a result of the decreased cell density, neuropil structures were shifted medially towards the ependymal layer in the right SVZ following irradiation. Mylenic and amylenic axons, including synaptic contacts, were observed adjacent to the ependymal layer, establishing direct contact with ependymal and/or astrocytic cells in some cases (Figure 3G). Together, these results indicate that the migrating neuroblasts were not replenished even one month after irradiation, despite the presence of type B cells in the SVZ post-irradiation (Figure 2 and 3E).

Localized radiation affects the generation of NPCs and migrating neuroblasts from NSCs

The presence of type B1 cells after irradiation raised the question if their proliferation was intact. Double labeling of the SVZ against nestin and Ki67 one month after radiation showed proliferation in few of the residual type B cells (Figure 4A). These proliferating type B cells were further found to be in close proximity to blood vessels (Figure 4B). CD31 immunostaining for endothelial cells in the SVZ appeared to be no different between the control and irradiated SVZ (Supplementary Figure 1A). These results were also confirmed by electron microscopy, where the basal lamina and endothelial cells were no different between the left (unirradiated) SVZ and the right (irradiated) SVZ (Supplementary Figure 1B). There was no difference in the number of blood vessels (left SVZ: 66±4.93 vs. right SVZ: 51±7.57; p=0.172) and in the size of the blood vessels (left SVZ: 50.93±11.6 vs. right SVZ: 45.09±7.29; p=0.692) as quantified in the semithin section (Supplementary Figure 1C–D). Interestingly, when the dissected SVZ tissue was dissociated and cultured in vitro, the cells from the left SVZ gave rise to neurospheres whereas the cells from the irradiated right SVZ failed to form neurospheres (Figure 4C). Taken together, these results indicate that the residual type B cells are able to proliferate after irradiation but lack the ability to generate NPCs and migrating neuroblasts.

Localized radiation of SVZ affected the size of the RMS

Although there was no significant difference in the number of proliferating cells in the RMS (Figure 1F), there were fewer DCx⁺ migrating neuroblasts in the right RMS one month after localized right SVZ radiation as compared to the left RMS (Figure 5B). In addition, there was a significant reduction in cross-sectional size of the RMS (left RMS: 9892±737mm² vs. right RMS: 7202±474mm²) (Figure 5C). Ultrastructural analysis with electron microscopy revealed no morphological alterations in cell populations of the RMS despite the reduction in size (Figure 5D). Further, we noticed few cells in the irradiated RMS in their mitotic phase, which based on their ultrastructural features, were identified as neuroblasts (Figure 5D, inset).

Localized radiation disrupts the migration of neuroblasts

The anterior dorsal SVZ is one of the regions with the largest NSC population [24]. In this experiment, we irradiated the anterior dorsal region of the SVZ with a single 10-Gy lateral beam of 1 mm diameter (Supplementary Figure 2A–B). Whole mount staining of SVZ against GFAP and DCx showed no neuroblasts in the irradiated region after one week (Supplementary Figure 2C). Additionally, no neuroblasts from the posterior dorsal SVZ were found to migrate into the irradiated region indicating that radiation may have affected the migration of neuroblasts in the SVZ (Supplementary Figure 2C’-C”). We then examined the effect of irradiation on anterior SVZ (aSVZ) while sparing the posterior SVZ. A 5×9mm beam of 10Gy was used to deliver radiation to the aSVZ, RMS and OB (Figure 6A). Immunostaining of SVZ whole mount against γH2Ax, one-hour post irradiation, showed a precise radiation field confined only to the aSVZ, sparing the posterior SVZ (Figure 6B). DCx⁺ neuroblasts were seen one hour post-irradiation, but were absent after one day and one month post-irradiation (Figure 6C). The neuroblasts from the posterior SVZ did not migrate through the irradiated aSVZ, consistent with our results in Supplementary Figure 2. Further, there were no apoptotic cells observed one week after localized irradiation of the anterior SVZ, as assessed by immunohistochemistry against activated Caspase-3 (data not shown).

To evaluate if the migratory properties of neuroblasts were affected by irradiation, we injected GFP-expressing retrovirus into the SVZ to label NSCs and track the migratory pattern of newly generated neuroblasts, one day after radiation treatment. Three separate groups of mice received either no radiation or a single 10-Gy dose to their RMS/OB (sparing the entire SVZ) or to their aSVZ/RMS/OB. One day after irradiation, mice were injected with GFP-expressing retrovirus in the SVZ to label dividing NSCs and track the migration of neuroblasts (Figure 7A). As predicted, sham-irradiated mice had GFP⁺ cells in the SVZ, RMS and OB one week after injection (Figure 7B, insets 1, 2, 3 and 4). Mice that received radiation to the RMS/OB showed GFP⁺ cells in their SVZ but not in RMS and OB (Figure 7C, inset 5). Mice that received radiation to the aSVZ/RMS/OB showed no GFP⁺ cells (Figure 7D). Quantification of Ki67⁺ cells further revealed that the RMS/OB irradiation did not affect the SVZ proliferation (Supplementary Figure 3). These results indicate that irradiation of the RMS alone (sparing SVZ) can inhibit the migration of neuroblasts. Our findings suggest that irradiation did not disrupt the intrinsic mechanism of neuroblast migration, but the local microenvironment permissive of cell migration.