Microglia depletion and repopulation do not alter the effects of cranial irradiation on hippocampal neurogenesis

Cranial radiotherapy can cause lifelong cognitive complications in childhood brain tumor survivors, and reduced hippocampal neurogenesis is hypothesized to contribute to this. Following irradiation (IR), microglia clear dead neural progenitors and give rise to a neuroinflammatory microenvironment, which promotes a switch in surviving progenitors from neuronal to glial differentiation. Recently, depletion and repopulation of microglia were shown to promote neurogenesis and ameliorate cognitive deficits in various brain injury models. In this study, we utilized the Cx3cr1 CreERt2-YFP/ + Rosa26 DTA / + transgenic mouse model to deplete microglia in the juvenile mouse brain before subjecting them to whole-brain IR and investigated the short-and long-term effects on hippocampal neurogenesis. Within the initial 24 h after IR, the absence of microglia led to an accumulation of dead cells in the subgranular zone, and 50-fold higher levels of the chemokine C-C motif ligand 2 (CCL2) in sham brains and 7-fold higher levels after IR. The absence of microglia, and the subsequent repopulation within 10 days, did neither affect the loss of proliferating or doublecortin-positive cells, nor the reduced growth of the granule cell layer. Our results argue against a role for a pro-inflammatory microenvironment in the dysregulation of hippocampal neurogenesis and suggest that the observed reduction of neurogenesis was solely due to IR.


Introduction
Radiation therapy (RT) is a cornerstone in the treatment of pediatric high-grade brain tumors (de Rojas et al., 2019).However, it is accompanied by a high risk of neurocognitive sequelae, manifesting as difficulties in information processing, learning, and memory acquisition (Mulhern and Palmer, 2003;Ullrich and Embry, 2012).These progressive late complications are especially debilitating for childhood cancer survivors, as they result in impaired intellectual development, academic attainment and professional outcomes, thus decreasing their overall independence and quality of life (Ullrich and Embry, 2012).Reduced hippocampal neurogenesis is hypothesized to, partly, explain the RTassociated cognitive late complications (Boström et al., 2013;Monje et al., 2007, Monje et al., 2002).Neural stem and progenitor cells (NSPCs) are spatially confined to the subgranular zone (SGZ) in the dentate gyrus (DG) of the hippocampal formation, the compound structure responsible for memory formation (Seri et al., 2001;van Praag et al., 2002).These cells proliferate and, as a result, are prone to irradiation (IR)-induced apoptosis due to DNA damage and the formation of free radicals, a situation particularly exacerbated at younger ages, because of the higher numbers of NSPCs in the juvenile brain (Blomstrand et al., 2014;Fukuda et al., 2005, Fukuda et al., 2004).
Microglia, the immune cells in the brain, exhibit a ramified morphology and continuously survey their surroundings (Nimmerjahn et al., 2005).Following irradiation, microglia undergo a robust and acute reaction, coinciding with the loss of NSPCs and upregulation of genes related to inflammation and phagocytosis (Arcuri et al., 2017;Osman et al., 2020).The inflammatory microenvironment further reduces neurogenesis by promoting a switch from neuronal to glial differentiation (Monje et al., 2002).In recent years, pharmacological and genetic tools have been developed to deplete microglial populations in order to elucidate their roles in pathogenesis and repair (Shi et al., 2022).For example, in mouse models of Alzheimer's and Parkinson's disease, microglia depletion was shown to reduce cognitive impairment, neuronal damage, astrocytic activation, production of pro-inflammatory factors and oxidative stress (Dagher et al., 2015;Mancuso et al., 2019;Zhang et al., 2021).In the context of neurogenesis, a recent study demonstrated that repopulating microglia could attenuate cognitive decline in a traumatic brain injury model by stimulating neurogenesis after microglia depletion using the CSF-1R antagonist PLX5622 (Willis et al., 2020).
In this study, we utilized a transgenic mouse model to genetically deplete microglia from the brains of juvenile mice (Zhou et al., 2022) before subjecting them to cranial IR, aiming to avert the adverse effects of IR-induced neuroinflammation on neurogenesis (Monje et al., 2002).
Our results indicate that the absence of microglia in the brain prior to IR had no effect on neurogenesis, nor did the subsequent microglia repopulation of the hippocampal parenchyma.

Homozygous Cx3cr1
CreETt2-YFP and Rosa26 DTA mice were a gift from Robert Harris, Karolinska Institutet, also available in the Jackson Laboratory (Bar Harbor, ME, USA) with stock numbers 021,160 and 009669, respectively.Mice heterozygous for both transgenes, meaning Cx3cr1 CreERt2-YFp/+ Rosa26 DTA/+ (where " + " indicates wild-type), were chosen for the experimental procedures, while Cx3cr1 CreERt2-YFP/ + Rosa26 +/+ were used as controls.For the effects of tamoxifen wild-type female and male C57Bl6/J mice (Charles River, Sulzfeld, Germany) were utilized.The mouse colonies were bred and maintained in the KM-B Animal Facility of Comparative Medicine at Karolinska Institutet.All animals with the correct genotype generated from our breeding colony were used for experiments, males and females.Even though the study was not designed to detect sex differences, efforts were made to distribute the males and females evenly between the groups.In total, we used 135 mice for the study, 71 of which were males and 64 were females.All animals had access to food and water ad libitum and were housed socially (2-6 mice per cage) on a 12-hour light/dark cycle in individually ventilated cages.All experiments in this study were approved by and performed following the Swedish National Board for Laboratory Animals' guidelines and the European Community Council Directive (86/609/EEC) under the ethical permits N163/15, N127/16, and 13676-2020.

Tamoxifen preparation and injection
Tamoxifen (TAM, T5648-1G, Sigma-Aldrich, Merck, Darmstadt, Germany) was diluted in corn oil (C8267, Sigma-Aldrich) to a concentration of 12.5 mg/ml.The animals were injected intraperitoneally with 10 µl/g every 24 h on postnatal days (P) 18, 19 and 20.The animals were weighed before each injection and then every 3 days for health monitoring.

Irradiation
Twenty-one-day-old mice were initially anesthetized with 5 % isoflurane in an induction chamber in a mixture of air and oxygen (1:1), then transferred to the X-ray irradiator (CIX3 X-ray cabinet, XStrahl, UK) and placed in a prone position.The anesthesia was maintained with 1.5 % isoflurane during the irradiation procedure.Mice were cranially (whole-brain) irradiated with a 1.5 cm in diameter circular field.The collimator was an XStrahl Perspex tip applicator.A single dose of 8 Gy was delivered with a dose rate of 1.332 Gy/min (dosimetry uncertainty ca 2 %) at 300 kV and 10 mA.Focus skin distance (FSD) to the animal head was 32 cm.External filtration, giving a half-value layer (HVL) of 4.0738 mm Cu, was applied by adding a Thoraeus filter (1.0 mm Sn, 0.25 mm Cu, 1.50 mm Al).Sham controls (SH) were subjected to the same duration of anesthesia in the absence of IR.Animals were allowed to recover from the anesthesia and return to their cages.

Sample preparation for histology
Animals were deeply anesthetized with 100 mg/kg sodium pentobarbital and transcardially perfused with ice-cold phosphate-buffered saline (PBS).Left hemispheres were fixed in 4 % paraformaldehyde (PFA, Histolab Products AB, Sweden) for 2 days, followed by cryoprotection in 30 % sucrose (Sigma-Aldrich) prepared in 0.1 M phosphate buffer.After they had sunk, the brains were cut into 25 µm sagittal sections using a sliding microtome (Leica SM2010R, Leica Microsystems GmbH, Wetzlar, Germany).The sections were stored in a tissue cryoprotecting solution containing 25 % glycerol and 25 % ethylene glycol in 0.1 M phosphate buffer.

Cell quantification and hippocampal subregion measurements
The H&E stained sections were acquired with an AxioImager M2 microscope and the immunofluorescent images were acquired with a Carl Zeiss LSM 700 laser scanning confocal microscope (both Carl Zeiss, Oberkochen, Germany).Pyknotic cells, DCX + and Ki67 + cells were manually quantified in the SGZ/GCL of the DG of the hippocampus.For the Ki67 + cells, the length of the SGZ for each section was calculated using the Segmented Line tool in ImageJ (version 1.54b, US National Institutes of Health, Bethesda, Maryland, USA, https://imagej.nih.gov/ij/) and the density of Ki67 + cells was calculated as cells/mm.The granule cell layer (GCL) volume was calculated according to the Cavalieri principle as previously described (Xie et al., 2016).

Chemokine C-C motif ligand (CCL) 2 enzyme-linked immunosorbent assay (ELISA)
To quantify the total CCL2 protein content, the right brain hemisphere of each animal was sonicated in 1 ml of an ice-cold Tris-HCl (50 mM, pH 7.3) solution containing 5 mM EDTA.The homogenate samples were centrifuged at 10 000 × g for 10 min at 4 • C. The protein concentration was determined using the BCA protein quantification assay (Thermo Fisher, 23225) and the level of CCL2 was measured using the CCL2/MCP-1 Quantikine ELISA kit (R&D systems, MJE00).The

(caption on next page)
results were normalized to the total protein concentration determined by the BCA assay.

Statistical analysis
Statistical analysis was performed using GraphPad Prism (version 9.3, GraphPad, Inc., San Diego, CA, USA).Data from the histological and ELISA analyses are presented as mean ± s.d..A two-way ANOVA followed by a Tukey's post hoc test was used to compare the total numbers of DCX + and Ki67 + cells, respectively, between the control Cx3cr1 CreERt2- YFP/+ Rosa26 DTA/-and the microglia-depleted Cx3cr1 CreERt2-YFP/ + Rosa26 DTA/+ (SH) and irradiated (IR) animals for each time point, and for the supplementary figure comparing VEH and TAM on wild-type C57BL/6 mice.Statistical significance was assumed when p ≤ 0.05.

Results and discussion
Aiming to uncover the potential role of microglia on hippocampal neurogenesis in the context of radiation therapy, the Cx3cr1 CreERt2-YFP/ + Rosa26 DTA/+ transgenic mouse model was utilized, due to its ability to selectively deplete microglia (Lund et al., 2018).Briefly, Cx3cr1 CreERt2- YFP/CreERt2-YFP mice were bred with Rosa26 DTA/DTA mice, which possess a flox-neo/STOP-flox-diphtheria toxin subunit alpha (DTA) gene cassette knocked into the Rosa26 locus.This yielded a model where microglia can be depleted by the administration of TAM, which induces intracellular release of the DTA specifically in microglia (Lund et al., 2018) (Fig. 1a).Cx3cr1 CreERt2-YFP/+ Rosa26 DTA/+ mice received 3 TAM injections with 24-hour intervals, beginning on P18, resulting in 90 % microglia depletion 24 h after the 3rd injection (Fig. 1b), as described previously (Zhou et al., 2022).Microglia depletion was maintained for 7 days after the third TAM injection, but by day 10, microglia had fully repopulated the hippocampus (Fig. 1c).
Radiation is known to dose-dependently deplete the NSPC niche, leading to cell death in the SGZ of the DG within hours (Fukuda et al., 2004).Since microglia are the primary phagocytic cells in the CNS (Neumann et al., 2009), we assessed the clearance of dead cells in the microglia-depleted mice.After whole-brain IR of Cx3cr1 CreERt2-YFP/ + Rosa26 DTA/+ mice on P21, brains were collected 6 or 24 h later (Fig. 1d).Hematoxylin and eosin staining showed an equivalent accumulation of pyknotic cells in the SGZ of the hippocampi of both control and microglia-depleted (MG dep) groups 6 h after irradiation, 2.40 ± 0.13 × 10e 4 and 2.32 ± 0.14x × 10e 4 , respectively (Fig. 1e and f).Twenty-four hours after IR, however, we noticed a significant delay in the removal of the dead cells in the SGZ of the MG dep mice compared to the control group (1.65 ± 0.17 × 10e 4 vs. 0.42 ± 0.12 × 10e 4 cells, respectively, p ≤ 0.0001) (Fig. 1f).Both groups displayed lower average numbers of dead cells 24 h compared to 6 h after IR (Fig. 1f), indicating ongoing clearance of pyknotic neural progenitors in both groups, but at a slower rate in the microglia-depleted brains.
During the course of an infection or other inflammatory events, CCL2 is responsible for controlling the entry of cells into the affected tissue, including the central nervous system (Shi and Pamer, 2011).Previous work (Kalm et al. (2009); Han et al. ( 2016)) has shown that CCL2 is robustly upregulated in microglia in the hippocampus after irradiation.
We then explored the impact of the absence of microglia at the time of irradiation and their subsequent repopulation on neurogenesis.We sacrificed animals at 3 time points: P27, corresponding to the first signs of repopulation; P30, when repopulation of microglia was completed; and P62, for the long-term effects (Fig. 2a).Surprisingly, our results showed that neither microglia depletion prior to cranial irradiation nor the repopulation could prevent the loss of immature neurons in the DG of the hippocampus, as indicated by doublecortin (DCX)-positive cells (Zanni et al., 2021) (Fig. 2b and 2c).Radiation was the only factor affecting the number of DCX + cells (Fig. 2c).It has been suggested that TAM may affect neurogenesis (Smith et al., 2022), but we could not find any differences in the numbers of DCX+cells 7 days after the last TAM injection in wild-type mice, neither in males nor in females (Supplementary Fig. 1).Similarly, the density of proliferating cells (Ki67 + ) in the SGZ (Fig. 2d) was not affected by microglia depletion and subsequent repopulation at any timepoint (Fig. 2e).The proliferating cells in the SGZ, containing the NSPC niche (Seri et al., 2001), are predominantly neural progenitors and, to a lesser extent, radial glia-like stem cells (Aimone et al., 2014;Denoth-Lippuner and Jessberger, 2021).This finding was unexpected for several reasons.Firstly, ionizing radiation is known to drive microglia, as the resident immune cells of the brain, and other glial cells to sustain a chronic neuroinflammatory environment, which in turn perturbs neurogenesis (Constanzo et al., 2020;Lumniczky et al., 2017;Monje et al., 2002).Secondly, it has been shown that activated microglia also phagocytose damaged and stressed live immature neurons, in a process termed phagoptosis by Brown and Neher (Brown and Neher, 2014;Brown and Vilalta, 2015;Butler et al., 2021).Lastly, similar approaches have claimed that the depletion of microglia reduces the production of inflammatory factors (Laudenberg et al., 2024;Szalay et al., 2016).Our conflicting findings, however, are in agreement with studies that emphasize the importance of microglia in the modulation of hippocampal neurogenesis and show that their depletion ultimately may lead to decreased neuronal survival (Alonso  C) receptor 1 (Cx3cr1) gene was replaced by a CreERT2 coding sequence, followed by an enhanced yellow fluorescent protein (EYFP).The introduction of tamoxifen (TAM) induced expression of the Cre recombinase, but only in microglia, resulting in expression of the diphtheria toxin (DTA), which led to depletion of the cells expressing DTA.b) Representative images of the experimental design for TAM dosage optimization.TAM was administered in 1, 2 or 3 doses in Cx3cr1 CreERt2-YFp/+ Rosa26 DTA/+ mice, with 24-hour intervals, beginning on postnatal day 18.The animals were sacrificed 24 h after the final tamoxifen injection.c) Confocal images showing the time course of the repopulation of microglia after they were depleted with 3 consecutive tamoxifen doses.On day 7, after 3 TAM injections, the microglial population was still depleted.On day 10, we observed a complete repopulation.d) Experimental scheme.TAM was administered intraperitoneally once daily for 3 consecutive days to control Cx3cr1 CreERt2-YFp/+ Rosa26 DTA/-(Control) mice and Cx3cr1 CreERt2-YFp/+ Rosa26 DTA/+ (MG dep: Microglia depletion) mice.The animals were subjected to whole-brain irradiation (IR) on postnatal day (P) 21.Brains were collected 6 or 24 h after IR. e) Visualization of cells and pyknotic nuclei in the subgranular zone (SGZ) of the mouse hippocampus 6 and 24 h after IR, using hematoxylin and eosin (H&E) staining.The magnified images illustrate the level of pyknosis observed in each experimental group.f) Quantification of the pyknotic nuclei in the same areas 6 and 24 h after IR, respectively.g) Concentration of CCL2 measured by enzyme-linked immunosorbent assay (ELISA) 6 and 24 h after whole-brain IR in the four experimental groups.For all statistics: n = 5-6, mean + S.  Bellido et al., 2023;Diaz-Aparicio et al., 2020;Willis et al., 2020).Previous reports have indicated that IR leads to impaired growth of the juvenile GCL and, consequently, to a smaller final volume of this structure in the hippocampus (Blomstrand et al., 2014;Boström et al., 2013;Zhou et al., 2017).Depletion of microglia did not affect this outcome.The GCL volume was significantly reduced 6 weeks after TAM and irradiation, 0.195 ± 0.009 and 0.208 ± 0.006 mm 3 in the IR control and IR MG dep groups compared to 0.240 ± 0.009 and 0.249 ± 0.006 mm 3 in the SH control and SH MG dep groups, p-value ≤ 0.01, respectively (Fig. 2f and 2 g).Previous studies using the CSF1R inhibitor PLX5622 to deplete microglia in mouse models of cranial radiotherapy, demonstrated positive effects on cognitive performance and less inflammation, but they did not investigate neurogenesis (Acharya et al., 2016;Feng et al., 2016;Krukowski et al., 2018).
In conclusion, our results showed that neither microglia depletion prior to irradiation nor their subsequent repopulation by macrophages/ microglia, presumably without IR damage, rescued neurogenesis.We demonstrated that the lack of microglia delayed the removal of dead neural progenitors and led to a massively increased production of CCL2, thus promoting a pro-inflammatory environment.Furthermore, the age of mice (P21 to P30) represents a period of extensive synaptic pruning and maturation of neuronal networks.Microglia are responsible for synaptic pruning and the absence of these cells during this critical period may interfere with brain development (Paolicelli et al., 2011).Our findings challenge the proposed role for a pro-inflammatory microenvironment in the dysregulation of hippocampal neurogenesis and suggest that the observed reduction of neurogenesis was solely due to IR.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Fig. 1 .
Fig.1.Microglia depletion in the Cx3cr1 CreERt2-YFp/+ Rosa26 DTA/+ mouse model delayed the clearance of irradiation-induced dead cells in the SGZ of the hippocampus.a) Schematic representation of the mouse model.Exon 2 of only one allele of the chemokine (C-X3-C) receptor 1 (Cx3cr1) gene was replaced by a CreERT2 coding sequence, followed by an enhanced yellow fluorescent protein (EYFP).The introduction of tamoxifen (TAM) induced expression of the Cre recombinase, but only in microglia, resulting in expression of the diphtheria toxin (DTA), which led to depletion of the cells expressing DTA.b) Representative images of the experimental design for TAM dosage optimization.TAM was administered in 1, 2 or 3 doses in Cx3cr1 CreERt2-YFp/+ Rosa26 DTA/+ mice, with 24-hour intervals, beginning on postnatal day 18.The animals were sacrificed 24 h after the final tamoxifen injection.c) Confocal images showing the time course of the repopulation of microglia after they were depleted with 3 consecutive tamoxifen doses.On day 7, after 3 TAM injections, the microglial population was still depleted.On day 10, we observed a complete repopulation.d) Experimental scheme.TAM was administered intraperitoneally once daily for 3 consecutive days to control Cx3cr1 CreERt2-YFp/+ Rosa26 DTA/-(Control) mice and Cx3cr1 CreERt2-YFp/+ Rosa26 DTA/+ (MG dep: Microglia depletion) mice.The animals were subjected to whole-brain irradiation (IR) on postnatal day (P) 21.Brains were collected 6 or 24 h after IR. e) Visualization of cells and pyknotic nuclei in the subgranular zone (SGZ) of the mouse hippocampus 6 and 24 h after IR, using hematoxylin and eosin (H&E) staining.The magnified images illustrate the level of pyknosis observed in each experimental group.f) Quantification of the pyknotic nuclei in the same areas 6 and 24 h after IR, respectively.g) Concentration of CCL2 measured by enzyme-linked immunosorbent assay (ELISA) 6 and 24 h after whole-brain IR in the four experimental groups.For all statistics: n = 5-6, mean + S.D., *p-value ≤ 0.05, **p-value ≤ 0.01, ***p-value ≤

Fig. 2 .
Fig.1.Microglia depletion in the Cx3cr1 CreERt2-YFp/+ Rosa26 DTA/+ mouse model delayed the clearance of irradiation-induced dead cells in the SGZ of the hippocampus.a) Schematic representation of the mouse model.Exon 2 of only one allele of the chemokine (C-X3-C) receptor 1 (Cx3cr1) gene was replaced by a CreERT2 coding sequence, followed by an enhanced yellow fluorescent protein (EYFP).The introduction of tamoxifen (TAM) induced expression of the Cre recombinase, but only in microglia, resulting in expression of the diphtheria toxin (DTA), which led to depletion of the cells expressing DTA.b) Representative images of the experimental design for TAM dosage optimization.TAM was administered in 1, 2 or 3 doses in Cx3cr1 CreERt2-YFp/+ Rosa26 DTA/+ mice, with 24-hour intervals, beginning on postnatal day 18.The animals were sacrificed 24 h after the final tamoxifen injection.c) Confocal images showing the time course of the repopulation of microglia after they were depleted with 3 consecutive tamoxifen doses.On day 7, after 3 TAM injections, the microglial population was still depleted.On day 10, we observed a complete repopulation.d) Experimental scheme.TAM was administered intraperitoneally once daily for 3 consecutive days to control Cx3cr1 CreERt2-YFp/+ Rosa26 DTA/-(Control) mice and Cx3cr1 CreERt2-YFp/+ Rosa26 DTA/+ (MG dep: Microglia depletion) mice.The animals were subjected to whole-brain irradiation (IR) on postnatal day (P) 21.Brains were collected 6 or 24 h after IR. e) Visualization of cells and pyknotic nuclei in the subgranular zone (SGZ) of the mouse hippocampus 6 and 24 h after IR, using hematoxylin and eosin (H&E) staining.The magnified images illustrate the level of pyknosis observed in each experimental group.f) Quantification of the pyknotic nuclei in the same areas 6 and 24 h after IR, respectively.g) Concentration of CCL2 measured by enzyme-linked immunosorbent assay (ELISA) 6 and 24 h after whole-brain IR in the four experimental groups.For all statistics: n = 5-6, mean + S.D., *p-value ≤ 0.05, **p-value ≤ 0.01, ***p-value ≤ 0.001, ****p-value ≤ 0.0001 by two-way ANOVA Tukey's multiple comparisons test, alpha = 0.05.Scale bars = 200 μm (b, c) and 100 µm (e).SH: sham animals that were not irradiated.(For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) This work was supported by the Swedish Childhood Cancer Fund (Grant reference numbers: PR2021-0114, MT2021-0016, PROF2020-0001), the Swedish Research Council (ref number: 2022-01175), the Swedish Cancer Foundation (ref number: 20 1297 PjF), the Swedish Brain Foundation (ref number: FO2022-0313), the Stockholm County Council (ALF grants, ref number: FoUI-962615), Radiumhemmets Forskningsfonder, the Frimurare Barnhuset Foundation of Stockholm, and the Märta and Gunnar V. Philipson Foundation.CRediT authorship contribution statement Kai Zhou: Writingreview & editing, Visualization, Validation, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization.Georgios Alkis Zisiadis: Writingreview & editing, Writingoriginal draft, Visualization, Validation, Project administration, Investigation, Formal analysis, Data curation.Monique Havermans: Validation, Investigation.Adamantia Fragkopoulou: Writingoriginal draft, Project administration, Investigation, Formal analysis.Cecilia Dominguez: Visualization, Investigation.Makiko Ohshima: Validation, Investigation.Ahmed M Osman: .Carlos F.D. Rodrigues: Writingreview & editing, Validation, Supervision, Investigation.Klas Blomgren: Writingreview & editing, Supervision, Resources, Project administration, Investigation, Funding acquisition, Data curation, Conceptualization.