IFN‐γ promotes radioresistant Nestin‐expressing progenitor regeneration in the developing cerebellum by augmenting Shh ligand production

Abstract Background Patients with brain tumors, especially pediatric brain tumors such as cerebellar medulloblastoma, always suffer from the severe side effects of radiotherapy. Regeneration of neural cells in irradiation‐induced cerebellar injury has been reported, but the underlying mechanism remains elusive. Methods We established an irradiation‐induced developing cerebellum injury model in neonatal mice. Microarray, KEGG analysis and semi in vivo slice culture were performed for mechanistic study. Results Nestin‐expressing progenitors (NEPs) but not granule neuron precursors (GNPs) were resistant to irradiation and able to regenerate after irradiation. NEPs underwent less apoptosis but similar DNA damage following irradiation compared with GNPs. Subsequently, they started to proliferate and contributed to granule neurons regeneration dependent on the sonic hedgehog (Shh) pathway. In addition, irradiation increased Shh ligand provided by Purkinje cells. And microglia accumulated in the irradiated cerebellum producing more IFN‐γ, which augmented Shh ligand production to promote NEP proliferation. Conclusions NEP was radioresistant and regenerative. IFN‐γ was increased post irradiation to upregulate Shh ligand, contributing to NEP regeneration. Our study provides insight into the mechanisms of neural cell regeneration in irradiation injury of the developing cerebellum and will help to develop new therapeutic targets for minimizing the side effects of radiotherapy for brain tumors.


| INTRODUC TI ON
Radiotherapy has become one of the main treatment methods for primary and metastatic brain tumors because their complex anatomical structure and the existence of the blood-brain barrier make it difficult to effectively control by surgery and drug treatment. 1wever, radiotherapy inevitably causes irradiation brain injury.The survival and function of normal neural cells are seriously affected, and sequelae such as cognitive impairment, ataxia, and endocrine dysfunction occur.The incidence of irradiation brain injury after radiotherapy for nasopharyngeal carcinoma was 1.9% ~ 5%, 2 that for low differentiated glioma and brain metastases was 1% ~ 24% 3 and 8% ~ 20%, 4,5 respectively, while that for pediatric brain tumors, such as cerebellar medulloblastoma, is even higher, more than 60% of radiotherapy patients, especially young children, will have severe radiotherapy sequelae due to the damaged cerebellum. 6These side effects affect the quality of life of patients for a long time, and unfortunately, there is currently a lack of effective treatment or strategies to alleviate the sequelae of this radiotherapy.In our current study, we will utilize a model of cerebellar irradiation injury in neonatal mice to mimic the radiotherapy-induced cerebellum dysfunction for mechanistic study.
The total volume of the cerebellum accounts for approximately 10% of the entire brain, yet it contains more than half of the total number of neurons in the whole brain.The cerebellum controls the balance and movement of the body and assists in the development of cognitive functions. 7The development of the cerebellum has a unique pattern, mainly in infancy, and has strict stratification; from the outside to the inside, the cerebellum is divided into an external granule layer (EGL) composed of Math1-expressing granule neuron precursors (GNPs), a molecular layer (ML), a Purkinje cell layer (PCL), an internal granule layer (IGL) composed of mature granule neurons (GNs) and the innermost white matter (WM). 8,9GNs are the most abundant and dominantly functional neurons in the cerebellum and are proliferated and differentiated by GNPs in late embryonic development and the early postnatal period under the control of the sonic hedgehog (Shh) signaling pathway. 10,11In mice, GNPs are located in the EGL during the first 3 weeks of life, and they sense the Shh ligand secreted by Purkinje cells, which activate the Shh pathway to promote GNP proliferation.3][14] After 21 days, all GNPs migrate to the IGL to complete differentiation, so the EGL disappears and the cerebellum is fully developed.Therefore, GNP proliferation and differentiation are essential for the normal development and functional formation of the cerebellum.GNPs are highly vulnerable to endogenous damage such as genetic mutations and exogenous damage, including irradiation.Studies have shown that a 2 Gy dose of irradiation can damage 70% of the cells in the cerebellum EGL of newborn rats. 15Studies have also shown that EGL can be reconstructed in a relatively short period of time and eventually form a normal adult cerebellum. 16However, how irradiation-damaged cells in the cerebellum regenerate and the mechanisms involved in the process have not been fully studied.
Nestin is a type VI intermediate filament protein 17 that is generally expressed in neural stem cells and glial cells and is mainly distributed in the cytoplasm and cell fibers. 18Recent studies have also shown that Nestin is expressed in progenitor cells of various lineages. 19,20In our previous studies, we identified a population of Nestin-expressing progenitors (NEPs) located in the deep part of the EGL.Unlike GNPs, NEPs present a quiescent status in the developing cerebellum of normal mice, but they are more tumorigenic in Shh-type medulloblastoma than GNPs based on Shh antagonizing receptor Pathed1 (Ptc1) gene deficiency.However, although the residence and gene expression profile are distinct in NEPs and GNPs, NEPs are neuronal lineage commitment cells that develop into mature GNs both in vitro and in vivo.Moreover, NEPs share the same ability to respond to the Shh ligand, activating Shh signaling to promote proliferation. 21More recently, NEPs showed reprogramming plasticity in neural cell differentiation. 22The properties of NEPs lead us to speculate that they might be resistant to radiotherapy-induced damage and regenerate in the radiotherapyinjured cerebellum.Therefore, we established a radiotherapyinduced cerebellar injury model in neonatal mice to identify the capacity of NEPs in the postinjury regeneration and investigate the underlying mechanisms.

| Irradiation
Postnatal Day 4 (P4) mouse pups were irradiated with a single dose of 4 Gy.The pups' bodies were covered by tinplate with head exposure.After irradiation, the cerebella of irradiated and nonirradiated control pups were collected for sectioning, quantitative PCR, and Western blotting assays.For in vitro cell irradiation, GNP and NEP cells were isolated and purified from the cerebella of naïve P4 Math1-GFP and Nestin-CFP mice, respectively, and cells were plated on PDL-coated coverslips or 24-well plates and then irradiated at 2 Gy.After irradiation, cell numbers were counted, and immunofluorescence and comet assays were performed.

| Immunofluorescence/immunohistochemistry and Western blotting
For immunofluorescence, cells were plated on PDL-coated coverslips with the treatments mentioned in the text and fixed with 4% paraformaldehyde (PFA).After permeabilization with 0.1% Triton X-100 in PBS (PBST) and blocking with 1% BSA in PBST, coverslips were incubated with primary antibodies overnight at 4°C and then incubated with fluorescence-conjugated secondary antibodies for 2 h at room temperature.DAPI was stained at the last step for 10 min at room temperature.Four washes with PBST were performed following each staining step.Then, coverslips were mounted with Fluoromount G before being visualized using a Nikon Eclipse Ti microscope.For immunohistochemistry, cerebella were harvested and fixed overnight in 4% PFA, cryoprotected gradiently in 15% and 30% sucrose, frozen in Tissue Tek-OCT (Sakura Finetek) and cut into 10-12 μm sagittal sections.Immunofluorescence staining for sections shared the same protocol for cytostaining described above.
Hematoxylin and eosin-stained paraffin sections were performed by Wuhan Servicebio Technology.
Western blotting was performed as regular, cerebellum tissues were lysed in RIPA buffer (Beyotime) supplemented with protease and phosphatase inhibitors (Beyotime).Total lysates containing equal amounts of protein were separated by SDS-PAGE and subsequently transferred onto PVDF membranes.Membranes were then probed with antibodies.Western blotting signals were detected by using SuperSignal West Pico Chemiluminescent substrate (Thermo).

| Comet assay
GNPs and NEPs were irradiated at 2 Gy and underwent a 30-min recovery.Then cells were harvested and subjected to an alkaline comet assay by using the Comet Assay Kit (Trevigen) according to the manufacturer's protocol.After single-cell electrophoresis on comet slides, slides were stained with ethidium bromide, and comet tail length was quantified and normalized to the nonirradiated control.A minimum of 100 cells per sample were counted.

| Microarray analysis
RNAs isolated from GNPs and NEPs of naïve mice (Figure 2H), as well as NEPs of irradiated and nonirradiated mice (Figure 4A), were hybridized to Affymetrix Mouse Genome 430 2.0 arrays.Microarray data were preprocessed using robust multichip analysis.Gene ontology analysis was carried out to examine the biological functions of the differentially expressed genes using NexusExp3 software.KEGG functional enrichment analysis was conducted based on the R package Cluster Profiler, and the results of enrichment analysis were visualized via the R package enrich plot.A p value <0.05 was set as the criterion.

| EdU incorporation and detection
Mice cerebella were irradiated at P4 and then collected for frozen section preparation at different hours post irradiation.Six hours before cerebellum collection, EdU (Beyotime) was administered by intraperitoneal injection to mice at 100 mg/kg, and then the sections were stained for EdU.

| ELISA (enzyme-linked immunosorbent assay)
Irradiated and nonirradiated cerebella were homogenized and centrifuged, and the supernatant was collected to perform ELISA.
The total protein concentrations of the samples were detected with a BCA Protein Assay Kit (Beyotime) and adjusted to be equal in each group.To detect IFNγ, a 96-well plate was precoated with purified anti-mouse IFNγ monoclonal antibody (1:500, clone R4-6A2, BD Pharmingen) overnight at 4°C.The plate was washed with PBST (PBS containing 0.05% Tween-20) and blocked with 1% BSA in PBST.Samples and standards (recombinant mouse IFNγ protein, BD Pharmingen) were added to the wells and incubated for 1 h at room temperature.Biotinylated anti-mouse IFNγ antibody (1:500, clone XMG1.2, BD Pharmingen) was later used to bind to IFNγ antigen for 1 h at room temperature.Then, streptavidin-HRP (1:1000, BD Pharmingen) was added and incubated for 30 min.Finally, the substrate TMB (Beyotime) was added for color development, and the reaction was stopped by 1 M H 2 SO 4 for 10 min.
The optical density (OD) in each well was read with a microplate reader at 450 nm.The IFNγ level was calculated according to the standard curve.

| Real-time quantitative PCR (qPCR)
Total RNA was extracted from cell lysate or tissue homogenate using TRIzol reagent (Sigma) in RNase-free conditions, and the purity of RNA was detected with a Nanodrop 2000 spectrophotometer (Thermo).cDNA was synthesized using oligo (dT) and Superscript II reverse transcriptase (Invitrogen).qPCR was performed in triplicate using SYBR qPCR Master Mix (Vazyme) and the ABI 7500 TaqMan Real-Time PCR Detection System.The differences in mRNA expression were calculated by the 2 −ΔΔCt method.The primers (all for mouse species) used in the experiments included GAPDH (forward:

| Slice culture
Cerebellum slice cultures were performed according to a previous publication. 23Irradiated or nonirradiated cerebella were embedded in 3% tissue culture grade agarose, and 300 μm sagittal slices were cut using a VT100S vibrating microtome (Leica).Slices were then transferred to a 0.4 μm Nuclepore membrane (Millicell) at the interface between air and culture medium (NB-B27 medium) in a 6-well culture plate and incubated at 37°C in 5% CO 2 .In certain experiments, cerebellum slices derived from nonirradiated mice were treated with 200 U/mL recombinant mouse IFNγ (Peprotech) for 24 h during culture, and then tissue lysates were prepared for Western blotting and qPCR.Cerebellum slices derived from irradiated mice 12 h post irradiation were treated with an IFNγ neutralizing antibody (clone R4-6A2, BioXcell) or isotype control rat IgG1 for 24 h at 10 μg/mL, and then tissue lysates and frozen sections were prepared for Western blotting and immunohistochemistry, respectively.

| Statistical analysis
Experimental data were analyzed using GraphPad Prism software.
The Shapiro-Wilk test was used to test for data distribution normality.Unpaired Student's t test was used to calculate the difference when data exhibit Gaussian distribution.Data that do not exhibit Gaussian distribution was analyzed via a non-parametric Student's t test.Differences were considered to be significant when their value was less than 0.05 (p < 0.05).Data are expressed as the mean ± SEM.

| NEPs were more resistant to irradiation than GNPs
Our previous studies showed that NEPs are quiescent in normal cerebellum development but are more tumorigenic in Shhmedulloblastoma than GNPs in Ptc1 gene deficient mice. 21These findings lead us to speculate that NEPs may react to radiotherapyinduced damage differently than GNPs.To address this question, Nestin-CFP transgenic mice, in which the mouse Nestin promoter directs the expression of a cyan fluorescent protein (CFP) with a nuclear localization signal, and Math1-GFP mice were used to establish a radiotherapy-induced cerebellum injury model.In the cerebellum of Nestin-CFP mice, NEPs reside in the deep part of the EGL (Figure 1A), and there are also Nestin-CFP-positive cells located in the ML and WM, which are astrocytes and neural stem cells, respectively. 21Mouse cerebella were irradiated with a single dose of 4 Gy on P4, and at P6, the irradiated cerebella were harvested for immunohistochemistry analysis.Cerebella from Nestin-CFP mice at P6 without irradiation were used as controls.The results showed that irradiation eliminated the majority of GNPs in the cerebellum, resulting in a much thinner EGL in the irradiated cerebellum than in the naïve cerebellum (Figure 1A,B).This finding is consistent with previous reports that cerebellar GNPs are very sensitive to irradiation.However, cells that survived irradiation were enriched by CFP+ cells, suggesting that NEPs may be more radioresistant than conventional GNPs.To further confirm this, we also irradiated the cerebella of Math1-GFP mice in which conventional GNPs (specifically expressing Math1) are positive for GFP at P4.We then isolated cells from the cerebella at 1 and 2 days following irradiation and analyzed the number of GNPs (GFP+) and NEPs (CFP+ CD133-ACSA-

| NEPs resembled GNPs in DNA damage following irradiation but downregulated DNA damage response and repair (DDR) genes
To determine the reason for the radioresistance of NEPs, we examined the extent of DNA damage in GNPs and NEPs after irradiation by comet tailing assay.Purified GNPs and NEPs from nonirradiated mice were plated in vitro and treated with 2 Gy irradiation.Irradiated GNPs and NEPs were harvested for the comet tailing assay.
As shown in Figure 2A-E All these data suggest that in response to irradiation, a downregulated DDR probably render NEPs resistant to irradiation.

| NEPs regenerated to repopulate cerebellar neuronal populations after irradiation
Next, to investigate whether radioresistant NEPs have regeneration capacity after irradiation, we evaluated their proliferation in the cer- controls.As expected, conventional GNPs (GFP+) in the EGL were highly proliferative (Ki67+) in nonirradiated Math1-GFP cerebella (Figure 3A).The majority of GNPs were ablated by irradiation, resulting in almost no GFP-positive cells remaining in the EGL following irradiation, and they did not proliferate after irradiation (Figure 3B).Consistent with our previous report, NEPs were predominantly quiescent in naïve Nestin-CFP cerebellum at P8, which resided in the deep part of the EGL (Figure 3C).However, the majority of NEPs were positive for Ki67 4 days post irradiation (Figure 3D), suggesting that NEPs initiated their proliferation in response to irradiation.
Then, we checked the size of the cerebella with/without irradiation at P7 and P21 by cerebellum appearance and hematoxylin and eosinstained mid sagittal sections.We found that the irradiated cerebella were much smaller than the nonirradiated ones 3 days after irradiation, while the size of the irradiated cerebella almost recovered to that of the nonirradiated cerebella at P21 (Figure 3E).The results suggest that NEPs proliferated and the regeneration of the cerebellum occurred after irradiation.To determine whether those proliferative NEPs contribute to neurogenesis in the irradiated cerebellum, we crossed Nestin-CreER T2  in the situation of irradiation-induced cerebellar injury, NEPs proliferate and differentiate into mature granule neurons.These data showed increased granule neurons derived from NEPs after irradiation, indicating that NEPs participate in the regeneration of the irradiated cerebellum.

| Regeneration of NEPs relied on the Shh signaling pathway
To further investigate the molecular basis for NEP regeneration after irradiation, we harvested NEPs from Nestin-CFP cerebella 2 days after irradiation at P4, since at this time point, NEPs started to proliferate in response to irradiation.Naïve NEPs from P6 Nestin-CFP mice without irradiation were collected as controls.Then, mRNA was extracted from proliferative NEPs and naïve NEPs, which were used for microarray analysis.Then, KEGG functional enrichment analysis was performed.The top 15 activated signaling pathways were enriched (Figure 4A).Several signaling pathways involved in cell proliferation were activated in NEPs derived from irradiated mice compared with their naïve counterparts, such as the MAPK and WNT pathways.Among them, the Shh signaling pathway was also activated.To confirm whether Shh signaling is involved in NEP regeneration, we next tested whether such regeneration can be re-

| Augmented Shh ligand was provided by Purkinje cells following irradiation
Since the Shh signaling pathway is involved in NEP regeneration after irradiation, we wondered which step of the pathway was altered to contribute to NEP regeneration.To address this question, we first tested the trigger of the Shh signaling pathway, Shh ligand, in the cerebella of irradiated mice 2 days post irradiation.The Western blotting results showed that the protein level of active Shh ligand (Nterminal of full-length Shh, Shh-N) was augmented in the irradiated cerebella compared with that in the naïve cerebella (Figure 5A,B).
The mRNA level of Shh was also evaluated and shared a similar pattern (Figure S1).These results indicate that Shh ligand expression was upregulated in the cerebellum post irradiation.To confirm whether NEPs can respond to the Shh ligand to proliferate, NEPs were isolated from irradiated Nestin-CFP mice at 2 days post irradiation and cultured in vitro for 24 h with recombinant Shh-N protein or vehicle control.The results showed that NEPs proliferated with Shh-N incubation but not vehicle control (Figure S2), suggesting that NEP proliferation in vivo via Shh signaling after irradiation probably due to their responding to the enhanced Shh ligand.Then, considering that Purkinje cells and astrocytes are the main source of Shh ligand in normal cerebellum development and Shh-medulloblastoma tumorigenesis, respectively, 13,24 we then identified the cell source of Shh ligand after irradiation in our model.For this purpose, Shh-Cre-GFP transgenic mice, in which cells expressing Shh are GFP positive, were irradiated on cerebella at P4, and the cerebella were collected at P6 for immunohistochemistry analysis.Cerebella from nonirradiated P6 Shh-Cre-GFP mice were also collected as controls.We observed that Shh-producing cells (GFP+ cells) were significantly increased in the irradiated cerebella compared with the nonirradiated ones, and the Shh-producing cells resided in the Purkinje cell layer (Figure 5C,D).
Then, the cerebellar sections were counterstained with the Purkinje cell marker, Calbindin (Figure 5E) and astrocyte marker GFAP (data not shown), and it was confirmed that Purkinje cells were the main source of the Shh ligand after irradiation.Meanwhile, as shown in Figure 5E, The number of Purkinje cells was increased following cerebellar irradiation.To confirm this, Purkinje cell numbers in the mid sagittal cerebellum sections were counted according to Calbindinpositive staining.As shown in Figure 5F-H, Purkinje cell numbers in cerebellar lobules II-IV were comparable between the irradiated and nonirradiated cerebella, while their numbers in lobules V-X were significantly higher in the irradiated cerebella than in the naïve controls, although the size of the irradiated cerebella was smaller.Then, we wondered if Purkinje cells proliferated after irradiation.To confirm this, WT mice were irradiated and then collected for section preparation at 12, 24, 48, and 72 h post irradiation.EdU was administered to mice 6 h before cerebellum collection, and then the sections were stained for EdU and Calbindin.A parallel experiment was performed with naïve mice as a control.As shown in Figure 5I-L and Figure S3, the number of Purkinje cells increased significantly beginning 24 h post irradiation, but they did not show signs of proliferation, as evaluated by EdU incorporation at these time points.These results led us to speculate that the increase in Purkinje cells might be due to differentiation from their precursors.We then immunostained the sections above for the Purkinje cell precursor marker, Lhx1 and found that the number of Purkinje cell precursors in the irradiated cerebellum increased at 12 h and decreased beginning 24 h after irradiation (Figure S3 and Figure 5K,L).All the data above indicate that Shh ligand was upregulated in the Purkinje cells of the irradiated cerebellum.And the number of Purkinje cells was increased in lobule V-X of the irradiated cerebellum, which might be differentiated from their precursors.

| IFNγ stimulated Shh ligand expression following irradiation, contributing to NEP proliferation
We next sought to investigate the mechanisms by which the Shh ligand was upregulated in the irradiated cerebellum.It was reported by some publications that Shh ligand expression can be induced by IFNγ in neural cells under certain conditions. 25,26This led us to speculate that IFNγ may also contribute to Shh ligand production by Purkinje cells following irradiation.First, we evaluated the mRNA and protein levels of IFNγ in the cerebella with or without irradiation.As shown in Figure 6A,B, the mRNA level of IFNγ in the irradiated cerebellum 24 h post irradiation was significantly upregulated compared with that in the nonirradiated cerebellum, while IFNγ expression showed no difference between the irradiated cerebellum and the control cerebellum when they were collected 72 h after irradiation, indicating that IFNγ expression was induced transiently by irradiation.Meanwhile, the protein level of IFNγ in the cerebella 24 h after irradiation was also increased compared with the control.
Then, we tested whether the downstream signal of IFNγ was activated in the irradiated cerebellum.As shown in Figure 6C,D, the phosphorylation of STAT1 was increased in the irradiated cerebellum 24 h after irradiation, indicating that the downstream signal of IFNγ was activated in this circumstance.Next, we tried to identify the cell source of the increased IFNγ in the irradiated cerebellum.
Since T cells are known as the main source of IFNγ, we tested the Tcell composition in the cerebellum by FACS.As shown in Figure S4A, T cells (CD3+) constituted less than 1% of the cerebellum even with irradiation.T cells from irradiated and nonirradiated cerebella 24 h after irradiation were sorted for RNA extraction to evaluate IFNγ mRNA expression.And we found that IFNγ expression was not altered in T cells based on irradiation (Figure S4B).Considering that microglia, the immune cells in the brain, are reported to produce IFNγ in inflammatory status [27][28][29] and become activated in response to irradiation, 30,31 we then checked the presence of microglia in the irradiated cerebellum.Cerebella of Nestin-CFP mice were irradiated at P4 and harvested for frozen section preparation 24 h after irradiation, and the sections were immunostained with the microglia marker, Iba-1.The results showed that more microglia accumulated around the EGL region after irradiation (Figure 6E,F).To confirm whether the accumulated microglia in the irradiated cerebellum could produce more IFNγ, microglia were sorted from the irradiated and nonirradiated cerebella 24 h post irradiation, and then qPCR was performed to determine IFNγ expression.The results showed that IFNγ expression in microglia derived from the irradiated cerebellum increased significantly compared with that in nonirradiated controls (Figure 6G).Meanwhile, microglial activation markers such as IL-1α and IL-6 were also upregulated in microglia derived from the irradiated cerebellum.These results indicate that microglia in the cerebellum were activated by irradiation and expressed more IFNγ.We then further studied the contribution of IFNγ to Shh ligand production in the irradiated cerebellum.Since we demonstrated that the augmented Shh ligand was mainly from Purkinje cells, we first detected the presence of IFNγ receptors on Purkinje cells in the naïve cerebellum (Figure S5).Because it is infeasible to isolate primary Purkinje cells from the murine cerebellum, we chose to use a semi in vivo method, slice culture, which maintains the cell and molecule bioactivity during culture in plates, to perform the investigation.Naïve cerebella were collected from P4 WT mice and subjected to slice culture with exogenous IFNγ (200 U/mL) for 24 h.Then, the slice tissue was collected to determine the protein levels of the Shh ligand by Western blotting.As shown in Figure 6H, IFNγ treatment significantly increased the Shh ligand protein level compared with vehicle treatment.Meanwhile, the treatment induced the expression of the Shh signal pathway target genes including Gli1, Ptc2, and Sfrp1, indicating that the Shh signal pathway is activated by IFNγ in current system (Figure S6).We also performed local IFNγ injection into the cerebellum of Shh-Cre-GFP mice and found that IFNγ injection increased Shh/GFP+ cell numbers compared with vehicle

| DISCUSS ION
As one of the main therapeutic methods for brain tumors, radiotherapy shows potency but always causes severe side effects, especially in the treatment of pediatric brain tumors.For example, in the most common pediatric brain tumor, cerebellar medulloblastoma, radio-and chemotherapy will induce cognitive disorders, endocrine dysfunction and ataxia in patients due to cerebellar injury. 32,33The high proliferation of GNPs in the developing cerebellum makes them sensitive to radiotherapy, 34 and less mature GNs are differentiated to fulfill the physiological functions of the cerebellum as a consequence.Although in early studies it had been found that the injured cerebellum could regenerate in animal models, the mechanisms have not been fully demonstrated.More In our previous studies, NEPs were found to be quiescent in normal status but were more tumorigenic in Shh-medulloblastoma.
Then, we found that these NEPs showed resistance to irradiation and started to proliferate after irradiation.The regeneration of NEPs in the injured cerebellum was also reported by Dr. Joyner's group, who used an in vivo mouse model to demonstrate that NEPs have the plasticity to generate GNs postinjury to recover cerebellar functions. 22,35 our current study, we focused on the events and underlying mechanisms that contribute to NEP regeneration in the early stage of irradiation-induced cerebellar injury.First, we found that NEPs were more resistant to irradiation both in vivo and in vitro, while they were subjected to similar DNA damage but lower DDR responses than GNPs, suggesting that quiescence and low DDR-related gene expression probably render NEPs more tolerant to irradiation-induced DNA damage.Then, the radioresistant NEPs switch their quiescent status to proliferative status under the regulation of the Shh signaling pathway and further differentiate into mature GNs in the irradiated cerebellum.These results were consistent with the findings of Dr. Joyner's group.Next, to investigate the triggers to initiate NEP proliferation post irradiation, we detected augmented Shh ligand at both the protein level and the mRNA level in the irradiated cerebellum, suggesting that NEPs accessed more Shh ligand in the local environment post irradiation and responded to it.In the normal developing cerebellum, Purkinje cells provide Shh ligand to support GNP proliferation. 13,14ter, Purkinje cells were also found to provide Shh ligand to control astrocyte differentiation; thus, they play a central role in the cerebellum development. 36Therefore, to confirm whether the augmented Shh ligand post irradiation was provided by Purkinje cells, Shh-Cre-GFP transgenic mice were used to identify Shh-producing cells, and the results confirmed that Purkinje cells were the main source of Shh ligand in the irradiated cerebellum.Our data indicate that not only in normal cerebellar development but also in injured conditions, Purkinje cells contribute to neural cell proliferation through Shh ligand production.Surprisingly, we found that the number of Shhproducing Purkinje cells was increased in the irradiated cerebellum.
Further measurement showed that the numbers of Purkinje cells in lobules V-X of the irradiated cerebellum were increased.Our results were inconsistent with some early studies that demonstrated that Xirradiation affected the alignment and morphology but not the numbers of Purkinje cells in infant rats. 37,38We observed that not only the layers in all lobules but also the numbers in lobules V-X of Purkinje cells were altered in the irradiated mouse cerebellum, which might be due to the difference in animal irradiation models.However, irradiation did not induce the proliferation of Purkinje cells.Considering that Purkinje cells are differentiated from their precursors, 39,40 we next observed the accumulation of Lhx1+ Purkinje cell precursors in the irradiated cerebellum.These results suggest that the increase in Purkinje cell numbers post irradiation resulted from their differentiation but not proliferation.Taken together, the increased Purkinje cells are Shh-producing cells in the irradiated cerebellum; thus, the augmented Shh level should be the combined result of both increased Purkinje cell numbers and their Shh-producing activity.Furthermore, we investigated the regulatory mechanisms of augmented Shh ligand production post irradiation.2][43] However, the mechanism of its expression is less well understood.It was reported that IFNγ, an inflammatory cytokine in the immune system, could induce the expression of Shh through direct targeting of its transcription in GNPs. 25,26Here, we observed increased IFNγ expression in the irradiated cerebellum, prompting us to hypothesize that it would account for the augmented Shh ligand post irradiation.To address this question, we performed in vivo administration of recombinant IFNγ by intracerebellum injection into naïve Shh-Cre-GFP and Nestin-CFP mice and found that IFNγ administration increased the Shh-producing cell
2-) per cerebellum by FACS.CD133 and ACSA-2 are the cell surface markers for neural stem cells and astrocytes, respectively, and they were used here to exclude these two Nestin/CFP+ populations in Nestin-CFP mice.Compared with that in the nonirradiated cerebella, the number of GNPs was markedly reduced in the cerebella following irradiation.As a comparison, the number of NEPs remained indistinguishable between the irradiated cerebellar and nonirradiated ones, indicating that NEPs are more resistant to irradiation than GNPs (Figure 1C,D).In addition, we purified GNPs and NEPs from the cerebella of nonirradiated Math1-GFP and Nestin-CFP mice, respectively, and treated both cell populations with 2 Gy irradiation in vitro.Forty-eight hours after irradiation, cells were harvested to detect apoptosis and differentiation by immunocytochemistry.As shown in Figure 1E-G, most GNPs were positive for cleaved caspase-3 and exhibited condensed nuclei, suggesting that they were undergoing apoptosis.However, only a proportion of NEPs were found to be apoptotic; instead, the majority of NEPs started their differentiation in vitro, as reflected by the intensive expression of Tuj1, a maker of granule neurons.The above data demonstrate that NEPs represent a radioresistant neuronal population in the developing cerebellum.F I G U R E 1 Nestin-expressing progenitors (NEPs) were more resistant to irradiation than granule neuron precursors (GNPs).(A, B) The cerebella of Nestin-CFP mice were irradiated with a 4 Gy dose (IR) or not (Non-IR) at P4, and the cerebellum sections were collected 2 days later to perform immunostaining for GFP or CFP.DAPI counterstained the nuclei.(C, D) Math1-GFP (C) and Nestin-CFP (D) mice were irradiated on the cerebellum with a 4 Gy dose at P4, 1 or 2 days post irradiation, the numbers of GNPs (GFP+) and NEPs (CFP+ CD133-ACSA-2-) were quantified by FACS.n = 5 mice in panel (A-D).(E, F) GNPs and NEPs were sorted from the EGL of nonirradiated Math1-GFP and Nestin-CFP mice at P4, respectively, and seeded in culture plates.Then, the cells were irradiated with a single dose of 2 Gy.Twenty-four hours after irradiation, immunostaining was performed to detect the cell apoptosis marker cleaved caspase-3 (CC3) and differentiation marker Tuj1.(G) Statistical quantification of the percentage of CC3+ and Tuj1+ cells in panels (E, F). **p < 0.01; ***p < 0.001.
, comparable lengths of DNA tails were observed in GNPs and NEPs at 6 and 24 h following irradiation, suggesting that irradiation caused similar extents of DNA damage in GNPs and NEPs.We next examined the difference in the DDR of NEPs and GNPs after irradiation.For this purpose, we harvested NEPs and GNPs 24 h after 2 Gy irradiation and examined the phosphorylation of γ-H2AX, a marker of DDR.The results showed that the phosphorylation of γ-H2AX-positive cells in irradiated NEPs was less than that in GNPs (Figure 2F,G).Moreover, mRNA microarray was performed to determine the DDR gene expression profile in nonirradiated NEPs and GNPs.We observed markedly downregulated expression of DDR genes in NEPs compared with GNPs (Figure 2H), suggesting the decreased capacity of DNA repair of NEPs, which probably allow them to tolerate DNA damage by irradiation.
ebellar EGL by immunohistochemistry. Briefly, cerebella of Math1-GFP mice or Nestin-CFP mice were irradiated at P4 and collected at P8 for immunohistochemistry. Cerebella from Math1-GFP mice or Nestin-CFP mice at P8 without irradiation were also harvested as F I G U R E 2 Nestin-expressing progenitors (NEPs) received similar DNA damage as granule neuron precursors (GNPs) to irradiation but had a downregulated damage response and repair (DDR).GNPs and NEPs were isolated from the cerebella of nonirradiated Math1-GFP and Nestin-CFP mice, respectively, and were irradiated in vitro with a 2 Gy dose (A-G) or not (H).(A-D) Comet assay was performed at 6 or 24 h after irradiation.(E) Statistical quantification of the relative comet tail length in panels (A-D).(F, G) Phosphorylated γH2AX immunostaining was carried out 24 h post irradiation, and DAPI was used to counterstain the nuclei.(H) mRNA microarray was performed and analyzed to compare the DDR gene expression profiles in nonirradiated NEPs and GNPs.The heatmap shows the downregulation of DDR genes in NEPs compared with GNPs.
mice expressing inducible Cre recombinase in Nestin+ cells with R26R-GFP mice in which cells permanently express GFP after Cre recombination.Nestin-CreER T2 /R26R-GFP mice were orally treated with tamoxifen with or without irradiation at P4.The cerebella were harvested at P21 to analyze the fate of NEPs following irradiation.As shown in Figure 3F-J, in the IGL of the cerebellum, approximately 40% of mature granule neurons (NeuN+) were positive for GFP.As a comparison, in the Nestin-CreER T2 /R26R-GFP cerebellum without irradiation, GFP+ cells accounted for less than 20% of mature granule neurons, suggesting that normally a small proportion of granule neurons originate from NEPs, whereas F I G U R E 3 Nestin-expressing progenitors (NEPs) regenerated to repopulate cerebellar neuronal populations following irradiation.(A-D) Math1-GFP (A, B) and Nestin-CFP mice (C, D) were irradiated (B, D) with 4 Gy at P4 or not (A, C), and the cerebellum sections were collected at P8 followed by immunostaining for GFP/CFP and the proliferation marker Ki67.(E) Wild-type (WT) mice were irradiated or not at P4, and the brains or cerebella were collected for picture (upper and middle panels) and hematoxylin and eosin stained mid sagittal sections (lower panel) at P7 and P21.(F-I) Nestin-CreER T2 /R26R-GFP mice were irradiated (H, I) or not (F, G) at P4 and were treated with tamoxifen daily until the mice reached P21.Then, the cerebellum was dissected and sectioned for immunostaining for GFP and NeuN, DAPI counterstained the nuclei (G, I: zoom in of F, H). (J) Quantification of GFP/NeuN double-positive cell percentages among total NeuN-positive cells in panels (F-I).n = 4 mice.**p < 0.01.
pressed by inhibition of the Shh pathway.For this purpose, 24 h after irradiation, Nestin-CFP mice were treated with daily oral gavage of vismodegib, an established antagonist of the Shh signaling pathway component, Smoothened.Then, the cerebella were harvested at P8 to examine NEP proliferation by immunohistochemistry.As shown in Figure 4B,C, the proliferation of NEPs was markedly repressed by vismodegib treatment.These data indicate that Shh signaling is required for NEP regeneration after irradiation.

F I G U R E 4 | 11 of 14 HU
Regeneration of Nestin-expressing progenitors (NEPs) relied on the Shh signaling pathway.(A) Cerebella of Nestin-CFP mice were irradiated with a 4 Gy dose at P4 and were collected at P6 for NEP isolation.NEPs derived from naïve P6 mice served as controls.Then, mRNA was extracted from NEPs to perform microarray.KEGG functional enrichment analysis showed the top 15 activated signaling pathways in NEPs from irradiated cerebellum compared with nonirradiated controls.(B, C) Nestin-CFP mice were irradiated with 4 Gy at P4, and vismodegib (50 mg/kg) or vehicle was administered to the irradiated mice by oral gavage 1 day later for three consecutive days.Then, cerebellum sections were collected and immunostained for GFP and Ki67.n = 4 mice in panel (B, C). et al.injection (FigureS7A-C).Then, to further confirm whether IFNγ contributed to irradiation-induced Shh ligand augmentation, irradiated cerebella 12 h post irradiation were also collected to perform slice culture experiments.During the 24-h culture, an IFNγ neutralizing antibody was added to the culture medium at a 10 μg/mL dose, and purified rat IgG served as a control.As shown in Figure6I, Shh levels in control IgG-treated slice culture samples derived from irradiated cerebella were augmented significantly compared with those derived from nonirradiated cerebella, whereas treatment with IFNγ neutralizing antibody dampened the augmentation.These results suggest that IFNγ contributed to the increased Shh ligand production in the irradiated cerebellum.Furthermore, to determine whether F I G U R E 5 Augmented Shh ligand was provided by Purkinje cells following irradiation.Wild-type (WT) (A, B, F-L) and Shh-Cre-GFP (C-E) mice were irradiated on the cerebellum with 4 Gy or not at P4. (A) Two days post irradiation, cerebella were harvested, and the tissue lysates were prepared for Shh-N protein evaluation by Western blotting.GAPDH served as the protein sample loading control.(B) Statistical quantification of the density ratio of Shh-N/GAPDH in panel (A).(C-E) Shh-Cre-GFP mice were irradiated (D, E) or not (C), and cerebellum sections were obtained 2 days later and immunostained for GFP or together with the Purkinje cell marker, Calbindin.DAPI was used to stain the nuclei.(F, G) Mid sagittal cerebellum sections from irradiated (F) and nonirradiated (G) mice were immunostained with Calbindin.(H) Quantification of Calbindin-positive cell numbers in each lobule of the irradiated and nonirradiated cerebella in panel (F, G). (I, J) The cerebella of mice with or without irradiation were collected 48 h post irradiation, EdU was administered to mice 6 h before harvesting the cerebella, and the cerebellum sections were immunostained for Calbindin and EdU.(K, L) Twenty-four hours after irradiation, cerebellum sections were obtained and immunostained for the Purkinje cell precursor marker Lhx1, DAPI counterstained the nuclei.n = 4 mice.*p < 0.05; **p < 0.01.F I G U R E 6 Increased IFNγ stimulated Shh ligand expression following irradiation and contributed to Nestin-expressing progenitor (NEP) proliferation.Wild-type (WT) (A-D, G-I) and Nestin-CFP (E, F, J, K) mice were irradiated with 4 Gy or not at P4. (A) Twenty-four or 72 h after irradiation, cerebellum tissue was harvested for RNA isolation.The mRNA levels of IFNγ were determined by qPCR.(B, C) Twenty-four hours after irradiation, cerebellum tissue was collected and homogenized, the protein levels of IFNγ in the cerebellum were evaluated by ELISA (B), and the phosphorylated STAT1 (p-STAT1) and total STAT1 protein levels were assessed by Western blotting (C).(D) Statistical quantification of the density ratio of phosphorylated STAT1/ total STAT1 in panel (C).(E, F) Twenty-four hours after irradiation, irradiated (F) and control (E) cerebellum sections were prepared and immunostained for CFP and the microglial marker Iba-1, and DAPI counterstained the nuclei.(G) Microglia were sorted by FACS from irradiated or control mice 24 h after irradiation for RNA extraction, and the mRNA levels of the indicated cytokines were determined by qPCR.*p < 0.05; **p < 0.01.(H) Cerebella of naïve WT mice at P4 were collected and sliced in agarose gel and then subjected to slice culture on a semipermeable membrane with or without 200 U/mL recombinant IFNγ for 24 h.Then, the tissue was collected for Shh protein evaluation by Western blotting.(I) WT mice were irradiated at P4, 12 h after irradiation, cerebella were harvested for slice culture, and during slice culture, 10 μg/mL neutralizing anti-IFNγ antibody or control rat IgG was administered for 24 h, slices from nonirradiated cerebellum treated with vehicle served as a control.Then, the slice tissue was collected to evaluate the Shh protein levels by Western blotting, GAPDH served as the sample loading control.(J, K) Twelve hours after Nestin-CFP mice were irradiated; slice culture was performed and cultured for 48 h.During the culture, 10 μg/mL neutralizing anti-IFNγ antibody or control rat IgG was administered to the culture.Then, the slice sections were prepared and immunostained for Ki67 and CFP.n = 4 mice.IFNγ could promote NEP proliferation in the irradiated cerebellum, the cerebella of Nestin-CFP mice were irradiated and collected for slice culture 12 h later, and the slice culture was treated with IFNγ neutralizing antibody or control IgG for 48 h.Then, the slices were immunostained for Ki67 and CFP to detect NEP proliferation.The results showed that NEP proliferated during slice culture, while IFNγ neutralizing antibody treatment inhibited their proliferation (Figure 6J,K).Similarly, administration of IFNγ by intracerebellum injection into naïve Nestin-CFP mice also increased NEP proliferation in vivo (Figure S7D-F).Moreover, INFγ null/null mice were irradiated at P4, and the cerebellum sections were prepared 4 days later to detect the EGL regeneration.Although NEPs could not be clearly identified by immunofluorescence staining for Nestin in INFγ null/null mice since the abundant glial cells highly express Nestin with cytoplasm and cell fiber distribution, we did see that the thickness of the EGL as well as the proliferating cell number in the EGL was much lower in irradiated INFγ null/null mice compared with WT controls (Figure S8).All the results above suggest that the increased IFNγ production following cerebellum irradiation enhanced Shh ligand levels and contributed to NEP proliferation.
detailed mechanistic studies are needed to provide strategies for the clinical development of interventions that can mitigate radiotherapyinduced sequela.In our current study, using a radiotherapy-induced cerebellar injury model in neonatal mice, we demonstrated the resistance and regeneration of NEPs in irradiated cerebellum and investigated the regulatory mechanisms of their regeneration.
Shh-Cre-GFP mice were kindly provided by Dr. Jian-quan Chen Soochow University.INFγ null/null mice were kindly provided by Dr. Zeng-jie Yang at Fox Chase Cancer Center.Wild-type (WT) C57BL/6 mice were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd.All animals were maintained in the SPF animal facility of Soochow University and Chongqing Army Medical University.