Impaired oligodendrogenesis in the white matter of aged mice following diffuse traumatic brain injury

Senescence is a negative prognostic factor for outcome and recovery following traumatic brain injury (TBI). TBI‐induced white matter injury may be partially due to oligodendrocyte demise. We hypothesized that the regenerative capacity of oligodendrocyte precursor cells (OPCs) declines with age. To test this hypothesis, the regenerative capability of OPCs in young [(10 weeks ±2 (SD)] and aged [(62 weeks ±10 (SD)] mice was studied in mice subjected to central fluid percussion injury (cFPI), a TBI model causing widespread white matter injury. Proliferating OPCs were assessed by immunohistochemistry for the proliferating cell nuclear antigen (PCNA) marker and labeled by 5‐ethynyl‐2′‐deoxyuridine (EdU) administered daily through intraperitoneal injections (50 mg/kg) from day 2 to day 6 after cFPI. Proliferating OPCs were quantified in the corpus callosum and external capsule on day 2 and 7 post‐injury (dpi). The number of PCNA/Olig2‐positive and EdU/Olig2‐positive cells were increased at 2dpi (p < .01) and 7dpi (p < .01), respectively, in young mice subjected to cFPI, changes not observed in aged mice. Proliferating Olig2+/Nestin+ cells were less common (p < .05) in the white matter of brain‐injured aged mice, without difference in proliferating Olig2+/PDGFRα+ cells, indicating a diminished proliferation of progenitors with different spatial origin. Following TBI, co‐staining for EdU/CC1/Olig2 revealed a reduced number of newly generated mature oligodendrocytes in the white matter of aged mice when compared to the young, brain‐injured mice (p < .05). We observed an age‐related decline of oligodendrogenesis following experimental TBI that may contribute to the worse outcome of elderly patients following TBI.


| INTRODUCTION
Traumatic brain injury (TBI) is a major cause of death and disability (Capizzi et al., 2020;Maas et al., 2022).With increased life expectancy, an increasing incidence of elderly patients hospitalized after TBI is observed each year, and unfortunately high age is associated with a poor prognosis (Dams-O'Connor et al., 2013;Thompson et al., 2006;van der Vlegel et al., 2022).Axonal injury, observed across all TBI severities, is associated with cognitive impairment, fatigue and personality disorders (Blennow et al., 2012;McAllister, 2011;Sandsmark, 2016).In addition, wide-spread signs of axonal injury, clinically named diffuse axonal injury (DAI), are commonly observed in severe TBI.When present in the deep white matter tracts, there is an association with prolonged time of reduced level of consciousness, and some DAI patients persist with a permanent disorder of consciousness.These DAI lesions are associated with cognitive impairment, fatigue and other persistent symptoms leading to a poor quality of life (Adams et al., 1989;Kinnunen et al., 2011;Vieira et al., 2016).This type of TBI affects not only axons, but also myelin and the cellular constituents of the white matter tracts (Johnson et al., 2013;Mahoney et al., 2022;Perez et al., 2014).
Importantly, reactive proliferation of glial precursors cells takes place after TBI (Bye et al., 2011;Chirumamilla et al., 2002;Clark et al., 2021;Sullivan et al., 2013;Susarla et al., 2014).These cells retain a high level of proliferative activity throughout adulthood and are potentially capable of differentiating into oligodendrocytes (Dimou et al., 2008;Rivers et al., 2008).Oligodendrocyte precursor cells (OPCs) may thus replace lost OLs and be involved in regeneration of disrupted myelin sheaths.However, these processes appear insufficient in promoting recovery following TBI.
In senescence, the brain loses its self-renewal and regenerative capacity (Baker & Petersen, 2018).In addition, alterations have been observed in the white matter of the aging human brain (Merenstein et al., 2021;Wharton et al., 2015).Clinical and experimental studies have found exacerbated brain damage as well as an impaired ability of the aged brain for self-repair following TBI (Krukowski et al., 2018;Kumar et al., 2013;Marquez de la Plata et al., 2008;Mosenthal et al., 2004;Sun et al., 2019).Furthermore, the function of oligodendrocytes and OPCs might be worse with aging, contributing to impaired white matter integrity and cognitive decline (Bendlin et al., 2010;Sams, 2021).We, therefore, hypothesized that TBI in the elderly may exacerbate white matter deterioration through an incapability of OPCs to regenerate post-injury.

| Experimental design and animals
Male C57Bl/6 (Taconic, Denmark) mice were used and divided into two groups based on age.Mice aged 8-12 weeks (21-32 g) were included in the young groups while mice aged 62 weeks ±10 (SD) (31-44 g) were included in the aged groups.Mice were housed with free access to food and water for a minimum period of 7 days prior to sham injury or TBI.
For the EdU study, the same randomization method was used and mice, regardless of injury status, received intraperitoneal (i.p.) EdU injections (50 mg/kg) from day 2 post-injury for 5 consecutive days until they were sacrificed at day 7 post-injury.Fifteen young mice (sham: n = 7, cFPI: n = 8) and 17 aged mice (sham: n = 6, cFPI: n = 11) were used.

| Central fluid percussion injury model
The surgical protocol for the central fluid percussion injury, an experimental model of widespread axonal injury, has been previously described in detail (Ekmark-Lewén et al., 2013).In brief, anesthesia was induced in a ventilated Plexiglas chamber containing air with 4% isoflurane.The mice were then moved to a stereotaxic frame where anesthesia was delivered through a nosecone (isoflurane 1.2% and N 2 O/O 2 70/30%).Local anesthesia (bupivacaine; AstraZeneca, Stockholm, Sweden) was applied in the scalp, after which a skin incision was made to expose the skull.A 3.0 mm diameter craniotomy was drilled over the midline, keeping the dura mater and the superior sagittal sinus intact.
Thereafter, a plastic cap was attached over the craniotomy using tissue adhesive and secured by dental cement.The content of the plastic cap was filled with isotonic saline and the mouse was connected to the fluid percussion device (VCU Biomedical Engineering Facility, Richmond, VA).
The pendulum arm was released creating a fluid pulse transmitted towards the exposed dura matter and the cranial cavity.Injury-induced apnea was recorded.After the injury, the cap was removed, the bone flap was re-positioned, and the scalp skin sutured.Animals were put in a cage provided with a heating pad until they recovered from the anesthesia, and subsequently returned to their initial home cage.For sham injury, the same steps were followed, including anesthesia and craniotomy, but the pendulum was not released.

| Animals, surgical results and tissue acquisition
Our previous study has demonstrated a cFPI-related mortality rate of 13% (Flygt et al., 2018).The cFPI-related acute mortality in the present study was 14% in the young and 21% in the old mice.In young brain-injured animals post-injury apnea was 29 ± 10 (SD) sec (range 20-50 s) while in aged brain-injured animals, post-injury apnea was 35 ± 11 (SD) sec (range 20-58 s).One sham-aged animal developed an unexpected neurological deficit post-injury and a cFPI-aged mouse had a bleeding complication and was excluded.The pressure pulse created by the cFPI device was 1.98 ± 0.06 (SD) psi in young mice and 1.93 ± 0.13 (SD) psi in aged mice.
Mice were sacrificed by intraperitoneal injection of sodium pentobarbital (Pentobarbitalnatrium 400 mg/mL, VET ATL, Apoteket, Uppsala, Sweden; 200 mg/kg) at 2 or 7 dpi and subsequently transcardially perfused using 4% paraformaldehyde (PFA) (Merck, Darmstadt, Germany).Brains were removed and post-fixed overnight at 4 C in 4% PFA.Thereafter, brains were placed in 25% sucrose solution for 72 h and, subsequently, sectioned on a freezing microtome with the use of dry ice.Sections were stored in anti-freeze cryoprotectant solution (30% ethyleneglycol & 30% glycerol) at À20 C for further use.

| Cell quantification
For the quantification of DAPI/PCNA+, PCNA/Olig2+, EdU+ and EdU/Olig2+ cells, images were acquired with an Olympus BX51 light microscope (Olympus, Tokyo, Japan) using the Extended Focus Imaging (EFI) function and imported to ImageJ (Billerica, MA, USA) for manual cell quantification.Images were merged into composite files and minor linear adjustment of brightness was applied for minimization of background staining to the point where the intensity histogram of each image was not disrupted and maintained its entire original spectrum value (Cromey, 2010).Three images from the corpus callosum, one medial and two lateral, were obtained with a 20Â magnification objective and cell counts were averaged.In addition, two images were taken from the external capsule bilaterally, and cell counts were averaged.Cell counts of the same region were averaged between different sections.Three sections per animal were used for each analysis.
For the analysis, the size of each z-stack was approximately 26 μm with a step size of 1.7 μm on a 512 Â 512 resolution.For the quantification of Iba1+/EdU+, PCNA+/Nestin+ and PCNA+/Nestin+/Olig2+ cells, the size of each z-stack was set to 15 μm with a step size of 1.5 μm on a 1024 Â 1024 resolution.Two sections per animal were used for each analysis.For representative images, the z-stack size was kept at 26 or 15 μm, respectively, with a step size of 0.9 μm on a 1024 Â 1024 resolution.

| EdU labelling and detection
In order to study 5-ethynyl-2 0 -deoxyuridine (EdU) incorporation, 50 mg/kg of EdU (Thermo Scientific, MA, USA) was injected i.p. in young and aged mice following either sham surgery or cFPI.Administration of EdU started on day 2 post-injury and continued for five consecutive days, injecting one dose of EdU per day.Cellular EdU was visualized using a Click-iT ® EdU Cell Proliferation Assay (Invitrogen, Thermo Scientific, MA, USA) according to the manufacturer's recommendations.Antigen retrieval for EdU was accomplished through heat-induced method as described in the immunohistochemical section.Streptavidin-conjugated fluorophores (Streptavidin Alexa 488, Streptavidin Alexa 555) were used for the detection of EdU.

| Automatic cell quantification and Olig2 cell classification
Olig2+ cells were automatically quantified and classified based on their immunofluorescent intensity into low-expressing (Olig2 Low ) and high-expressing (Olig2 High ) cells using QuPath 0.3.2.For this analysis, only one type of fluorophore was used for the visualization of Olig2+ cells.Composite 20Â images of the blue (DAPI) and red (Olig2) channel were created using ImageJ.Images underwent background subtraction by utilizing the "Subtract Background" function in ImageJ.
Rolling ball radius for the function was set to 70 pixels.Thereafter, fluorescent images were imported into the QuPath software.The region of interest was selected using a polygon free-hand manual selection."Cell Detection" command was applied with the default software parameters and with threshold set to 5 on the Blue Channel in order to detect all nuclei within the region of interest.
Identified objects within the selected area, corresponding to detected nuclei, were automatically measured for their Red Channel Mean fluorescence intensity.This indicates the average pixel intensity that the software detects within each nucleus.
The generated intensity values histogram termed "Red Mean" includes the means of all individual nuclei included within a polygon selection.In order to control for intensity and background staining variation, the threshold point was based on the intensity values histogram "Red Mean" of each included region of interest.
As such, identified cells, classified based on the "Set Cell Intensity Classifications" command, were distinguished based on two threshold values.For high-expressing cells (Olig2 High ), threshold was set to Mean + two SDs of the overall measured intensity distribution.Cells with nuclei intensity higher than the mean value of the intensity histogram were classified as the total Olig2+ expressing cells.The difference between the two aforementioned cell populations was classified as low-expressing Olig2+ cells (Olig2 Low ).

| Quantitative real-time PCR
The workflow of Quantitative Real-time PCR has been previously described (Michalettos et al., 2021).In brief, tissue samples were dissected from snap-frozen brains (using isopentane at À70 C) at 2 and 7 days post-injury.A 2 mm coronal section, spanning approximately the middle of bregma and lambda, was obtained and thereafter the superficial cortex was removed.Two horizontal cuts were used to obtain the corpus callosum.The dissected tissue constituted a combination of white and gray matter including a partition of the cortex and the hippocampus (See Graphical Illustration, Figure 10).
Extraction of RNA from all brain tissue samples was done with RNeasy ® Mini Kit (Qiagen, Hilden, Germany) with the use of dithiothreitol (DTT).Extracts were analyzed for their RNA concentration and purity with NanoDrop (Thermo Scientific, Wilmington, USA).cDNA synthesis was done using iScript cDNA Synthesis Kit (Bio-Rad, Hercules, USA) in a MJ Mini Personal Thermal Cycler (Bio-Rad, Hercules, USA) according to the instructions of the manufacturer's protocol.
Quantitative PCR was carried out using the CFX Connect™ Real-Time System (Bio-Rad, Hercules, USA) with SsoAdvanced Universal SYBR ® Green Supermix (Bio-Rad, Hercules, USA) using a reaction volume of 10 μL.Gene-specific primers were used to quantify transcript levels of genes of interest (Supplementary Tables [Data S1]).Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) expression was used as reference control.All reactions were run under the following PCR parameters: a denaturation step of 95 C for 30 sec and an annealing/extension step for 30 sec for a total of 35 cycles.Annealing temperatures and primer concentration were individually optimized for each primer pair (Supplementary Table 1).
For the estimation of reaction efficiencies, serial dilutions of an initial pooled sample of all the samples were used.Reaction efficiencies ranged from 96% to 108%.Raw fluorescent unit (RFU) values were converted to relative levels of cDNA according to the reaction amplification formula (Pfaffl, 2001).A common fluorescent threshold was applied to all plate reactions using the same plate material (MultiPlate ® PCR Plates™, White, Bio-Rad).Gene expression was calculated as the ratio of target gene transcript levels to GAPDH transcript levels and are expressed as means ± standard error of the mean (S.E.M.).

| Statistical analysis
Statistical analysis and graphs were made with R 4.2.2 and GraphPad Prism 9.4.1.To assess for normal distribution, cell counts data were imported in R and visualized through histograms and post-ANOVA QQ-plots.Auxiliary Shapiro-Wilk normality tests were conducted in GraphPad Prism in order to ensure normality of data.Both cell counts and QPCR data were normally distributed and statistical comparisons between experimental groups were made using a Two-Way ANOVA (age and injury effect) for each time point, followed by Tukey-multiple comparison tests.Normally distributed data are presented as mean ± standard error of the mean (SEM).
The choice of sample size for each experimental group was made based on previous findings (Flygt et al., 2017).In specific, we conducted power analysis according to the data generated from a previous EdU-labelling study by utilizing the EdU+/Olig2+ cell count data which were derived from the study of the corpus callosum and the external capsule.Due to the explorative nature of this study, we did not possess preliminary data for the PCNA quantification end-point analyses, thus, power analysis was conducted only for the study design of EdU administration.We had previously observed approximately a 90% and a 120% increase of Olig2+/EdU+ cells in the corpus callosum and the external capsule, respectively, of young mice compared to their sham controls.We hypothesized a 50% reduction in double co-labeled EdU+/Olig2+ cells in the case of aged cFPI-mice compared to young cFPI-mice.We assumed equal variance of the cFPIyoung and the cFPI-aged group as well between the sham-young and the sham-aged group.Additionally, we hypothesized a 50% decrease of EdU+/Olig2 in sham-aged mice compared to the sham-young group.A detailed overview of the size of effects and the corresponding achieved power of each comparison is depicted in Supplementary Table 2. Selected power was set to 0.8.Power analysis was conducted using the G*Power 3 software (Faul et al., 2007).Based on the power calculations, a groups size of 8 was deemed sufficient to detect the hypothesized differences.
Due to lack of knowledge on the response of elderly animals, we chose a slightly higher sample size number (n = 11) for the cFPIgroup.

| Reduced number of proliferating Olig2+ cells in white matter tracts of naïve aged mice
To evaluate proliferation in the corpus callosum and the external capsule of young and aged mice, we analyzed the number of PCNApositive cells.Our first step was to compare the number of proliferating cells between naïve young and aged mice.Naïve aged mice had fewer proliferating cells in both the corpus callosum and the external capsule when compared to naïve young mice (Figure 1a,b).PCNA immunoreactivity was particularly enriched in the subventricular zone (SVZ) and in the granule cell layer of the hilus in the dentate gyrus (DG) of the hippocampus used as internal control for successful PCNA staining (data not shown) (Ino & Chiba, 2000).
In order to specifically identify proliferating OPCs, we analyzed cells co-labeled for PCNA and Olig2, a transcriptional factor expressed throughout the oligodendrocyte lineage.Similar to the results of PCNA+ cells, naïve aged mice had fewer PCNA+/Olig2+ cells in both white matter regions of interest compared to naïve young mice.This result indicates that the aged naïve brain exhibits reduced proliferation of Olig2+ progenitors within the white matter.Inboxes represent a higher image magnification of regions where PCNA staining was observed.Student's t-test was used to statistically test for differences between the two experimental groups.

| cFPI elicits a proliferative response in the corpus callosum and external capsule of both young and aged mice
In order to study cell proliferation post-injury, we applied two analyses.We used quantification of PCNA+ cells at day 2 (2dpi) and day 7 (7dpi) post-injury as well as quantification of EdU+ cells at 7 dpi following daily EdU injections from 2dpi to 6dpi (Figure 2a).At 2dpi, in both groups, there was an increased number of PCNA+ cells in braininjured cFPI mice compared to their age-matched sham-injured controls (Figure 2b).This difference reached significance only in the external capsule of young brain-injured mice (Figure 2c).The number of PCNA+ cells of young cFPI mice in both regions of interest was significantly higher compared to that found in aged cFPI mice (Figure 2c).By day 7 post-injury (7dpi), there was no difference in the number of PCNA+ cells between young sham and young cFPI mice in either the corpus callosum or the external capsule (Figure 2d).In contrast, in the aged mice, we detected a significantly higher number of PCNA+ cells compared to aged sham-injured mice (Figure 2d).Aged sham-injured mice had a lower number of PCNA+ cells in the external capsule and the corpus callosum compared to young sham-injured mice at day 7, respectively (Figure 2d).groups compared to their age-matched sham-injured controls (Supplementary Figure 1A-C).This increase reached statistical significance in the corpus callosum of the aged mice groups (Supplementary Figure 1C).

| Reduced proliferation of Olig2+ progenitors in aged mice following central fluid percussion brain injury
We, then, proceeded to analyze cells co-labeled for PCNA and Olig2, a transcriptional factor associated with the oligodendrocyte lineage.
At 2 dpi, young cFPI mice had a higher number of PCNA+/Olig2+ cells in the corpus callosum and external capsule compared to young sham-injured mice (Figure 3a-c).This difference was not observed at 7 dpi (Figure 3d).In the aged group, there were no changes in the number of PCNA+/Olig2+ cells of cFPI mice in either white matter regions compared to sham-injured controls, neither at 2 nor 7 dpi (Figure 3c,d).
EdU labelling of proliferating Olig2+ cells in both the corpus callosum and the external capsule over a period of 5 days between 2 dpi and 7 dpi confirmed a TBI-induced increase of this cell population in young mice at day 7 post-injury (Figure 4a-d).
In contrast, no significant increase was observed when comparing aged brain-injured and aged sham-injured mice (Figure 4c).In both the corpus callosum and the external capsule, young sham-injured mice had a higher number of Olig2+/EdU+ cells when compared to aged sham-injured mice (Figure 4c).Thus, both PCNA and EdU approaches support the hypothesis that proliferation of mitotic Olig2+ cells is impaired in the aging mouse brain as well as following TBI using a model of widespread axonal injury.We further investigated potential changes in the overall population of white matter oligodendroglia characterized by the expression of Olig2.High levels of Olig2 are expressed in OPCs while astrocytes and mature oligodendrocytes have low Olig2 expression (Buffo et al., 2005;Wang et al., 2021).To distinguish these cell populations, we classified Olig2+ cells as high-expressing cells (Olig2 High ) and low-expressing cells (Olig2 Low ) (Figure 5a,b).We detected a significantly higher number of Olig2 High cells in young cFPI mice compared to aged cFPI mice in the corpus callosum and the external capsule on 2 dpi (Figure 5c).These differences were not observed on 7 dpi.In addition, no difference in Olig2 Low cells was observed among the experimental groups both on 2 and 7 dpi.
Overall, these results reflect the accumulation of proliferating Olig2+ cells seen in young mice at day 2 post-injury and no substantial change in the numbers of Olig2-expressing mature oligodendroglia during the first week following TBI.

| Proliferation of PDGFRα precursors is not impaired in aged mice following cFPI
We further characterized the population of proliferating PCNA+ cells in the white matter tracts of mice following cFPI.Using the plateletderived growth factor receptor-α (PDGFRα) marker, we quantified co-labeled PCNA+/PDGFRα+ cells in the white matter tracts at 2dpi (Figure 6a,b).In the corpus callosum of sham-injured mice, the number of PCNA+/PDGFRα+ cells of young mice was significantly higher when compared to aged animals (Figure 6c).There was no F I G U R E 4 A TBI-induced increase of co-labeled EdU+/Olig2+ cells is observed in the white matter of young but not of aged mice.TBI-induced increase of these cells in the young mice (Figure 6c).In contrast, we observed a significant increase of PCNA+/PDGFRα+ cells in the corpus callosum of aged mice, increased to levels similar of that in the young ones (Figure 6c).A small but significant difference was observed in the corpus callosum of young and aged cFPI-injured mice (Figure 6c).No difference was detected in the external capsule between aged brain-injured and aged sham controls (Figure 6c).
Since PDGFRα is considered a marker of early glial progenitors that can potentially give rise to mature OLs, we hypothesized that proliferation of these cells could have occurred prior to the evaluated time-point of 2 dpi.We, thus, quantified PDGFRα+ cells at 2 dpi in both white matter regions of interest.
No difference was observed among all experimental groups (Supplementary Figure 2).We proceeded to quantify EdU+ cells, which expressed the PDGFRα marker following the 5-day labelling EdU time-period.Similar to the PCNA+/PDGFRα+ results, we did not detect a TBI-related increase of EdU+/PDGFRα+ cells in young brain-injured mice at 7 dpi following EdU injections (Figure 7a,b).In contrast, we observed a significant TBI-induced increase of EdU+/PDGFRα+ cells in the corpus callosum of aged mice (Figure 7b).No differences were observed among experimental groups in the external capsule.These results suggest that proliferation of early PDGFRα-expressing OPCs is not impaired in the white matter of aged mice following TBI.

| Decreased number of proliferating Olig2+ cells expressing nestin although not PDGFRα in aged mice post-injury
Despite a reduced number of proliferating Olig2+ cells in the white matter regions of aged mice following cFPI, the number of proliferating PDGFRα+ cells was not affected by age.We co-labeled Olig2 and PDGFRα in young and aged mice and found that the majority of PDGFRα+ cells expressed Olig2+ in both white matter regions of interest (Figure 8a).In both the corpus callosum and the external capsule, the average number of PCNA+/Olig2+ (Figure 3c) and PCNA+/PDGFRα+ (Figure 6c) cells were similar in mice of the aged group but not in mice of the young group.We, therefore, hypothesized that a population of proliferating Olig2+/PDGFRα-negative cells exists.To phenotype this population, we performed triple immunofluorescent PCNA/PDGFRα/Olig2 staining.On day two, we detected a subset of PCNA+/PDGFRα-/Olig2+ cells in both white matter regions of interest and especially in the external capsule of young brain-injured mice (Figure 8b).In brain-injured aged mice, a few PCNA+/PDGFRα-/Olig2+ cells were detected, restricted mostly to the external capsule (Figure 8b).
The localization of these cells in the proximity of the ventricle of the external capsule prompted us to investigate the expression of Nestin in this cell type.In both young and aged mice and especially in cFPI mice, a high proportion of PCNA+ cells were positive for Nestin (Figure 9a).Nestin+ and PCNA+/Nestin+ cells showed a mix of astrocytic and microglial morphology (Figure 9a).
In the corpus callosum of young cFPI mice, there was a significantly higher number of PCNA+/Nestin+ cells compared to the aged cFPI mice.In the external capsule, both young sham-injured and young cFPI mice exhibited a higher number of PCNA+/Nes-tin+ cells when compared to aged sham-injured and cFPI mice (Figure 9d).By combining triple immunofluorescent PCNA/Nestin/Olig2 staining, we could detect a subset of PCNA+/Olig2+ cells which expressed Nestin (Figure 9b,c).These cells were detected in both white matter regions of interest (Figure 9b,c).PCNA+/Nestin +/Olig2+ cells exhibited a bipolar and less branched morphology than PCNA+/PDGFRα+/Olig2+ cells (Figure 9b,c).Quantification of PCNA+/Nestin+/Olig2+ cells revealed a diminished number of this cell type in the white matter of aged cFPI mice compared to young cFPI mice (Figure 9d).Furthermore, a higher number was detected in the external capsule of young sham-injured mice compared to aged sham-injured mice.(Figure 9d).In addition, the expression of PDGFRα did not co-localize with the expression of Nestin (data not shown).In summary, these results point towards two distinct populations of mitotic Olig2+ progenitors in the white matter following TBI, and a decreased number of proliferating Nes-tin+/Olig2+ cells in aged cFPI mice.

| Reduced regeneration of CC1+ mature oligodendrocytes in the white matter of aged mice following cFPI
Following the 5-day period of daily EdU labelling injections, our next step was to determine whether we could detect mature oligodendrocytes at 7 dpi which were positive for EdU (Figure 10a).In both young and aged cFPI-injured mice, EdU+/CC1+/Olig2+ cells could be detected in both white matter regions of interest (Figure 10c).Quantification of EdU+/CC1+/Olig2+ revealed a TBI-related increase of this cell type in the corpus callosum of both young and aged cFPIinjured mice compared to their age-matched controls (Figure 10b).
However, the number of EdU+/CC1+/Olig2+ cells in the corpus callosum of aged cFPI-injured mice was significantly lower than that of young cFPI-injured mice (Figure 10b).A TBI-related increase could also be observed in the external capsule of young mice, while no increase was evident in aged mice.Additionally, young sham-injured mice had more EdU+/CC1+/Olig2+ cells in the external capsule compared to aged sham-injured mice (Figure 10b).These results indicate an impaired capability of the aging brain to regenerate mature oligodendrocytes, due to age alone as well as following TBI.
In order to further address the maturation process of newly generated OPCs following injury, we investigated the expression levels of Nkx2.2, Ascl1, Myrf and Tcf712 on 2 and 7 dpi (Figure 11).These transcription factors (TFs) are associated with the oligodendrocyte lineage.
Both at 2 and 7 dpi, no TBI-related changes were observed in young mice when compared to their age-matched controls (Figure 11).In contrast, on 2 dpi, aged cFPI-mice had increased transcript levels of Ascl1 and Myrf compared to young cFPI-mice (Figure 11).This up-regulation of Myrf transcript levels in aged cFPI animals was also significant compared to aged sham-injured mice (Figure 11).At 7 dpi, a down-regulation of transcript levels of all four genes of interest was observed in aged cFPI-mice when compared to young cFPI-mice (Figure 11).Additionally, a significantly decreased amount of Ascl1 and Nkx2.2 transcripts was evident in aged sham-mice compared to young sham-mice (Figure 11).Both on 2 and 7 dpi, a consistent downregulation of Tcf712 transcripts was observed in aged sham-injured mice compared to young sham-injured (Figure 11).These results argue for an unaltered expression of oligodendrocyte-related TFs in the young brain post-injury.In contrast, an altered expression pattern of these TFs was observed in the aged brain following TBI.

| Increased number of Iba1+/EdU+ cells in the corpus callosum of injured aged mice following cFPI
In contrast to the numbers of Olig2+/EdU and PDGFRα+/ΕdU+ cell counts, the total number of EdU+ cells post injury was considerably higher and cannot be defined as proliferating OPCs alone (Supplementary Figure 1).Based on their morphology and distribution, we hypothesized that such cells could be potentially linked to inflammation.Co-labelling of EdU with Ionized calcium binding adaptor molecule 1 (Iba1), a microglia/macrophage marker, showed a considerable amount of Iba1+ cells that had incorporated EdU, both in young and aged mice (Figure 12).The number of Iba1+/EdU+ cells was significantly higher following TBI in the corpus callosum of aged mice when compared to both their age-matched sham controls and young cFPI mice (Figure 12d).No differences were observed in the external capsule (Figure 12d).proliferation or due to cell death, following TBI.These results argue for an exacerbated inflammatory response in brain-injured, aged mice.
In addition, we did not observe GFAP+/EdU+ cells in the selected white matter tracts, indicating that white matter astrocytes had not proliferated during the 5-day EdU labelling period (Supplementary Figure 3).

| DISCUSSION
In this study, we evaluated the proliferative response of glial progenitors associated with the oligodendrocyte lineage in young and aged mice using an experimental TBI model causing widespread axonal injury.For this reason, we focused our investigation on proliferation exclusively in white matter brain regions.Our results suggest that oligodendrogenesis is impaired in the aging brain following TBI.Oligodendrogenesis can be defined as a two-step process, involving the initial stage of progenitor cell proliferation and later the maturation into potentially myelinating and functional oligodendrocytes.Our data provide evidence that proliferation of oligodendrocyte progenitor cells (OPCs) is impaired in the aging mouse brain following TBI which might be associated with decreased or delayed regeneration of mature oligodendrocytes.
Contrary to the embryonic establishment of distinct OPC and mature oligodendrocyte populations (Kondo & Raff, 2000;Ligon et al., 2006;Marques et al., 2016;Pozniak et al., 2010; Wang an impairment in the aged rodent brain to recruit functional oligodendrocytes to the site of lesion (Doucette et al., 2010;Neumann et al., 2019;Sim et al., 2002).Similarly, in a cerebral hypoperfusion stroke model affecting the white matter, oligodendrogenesis was impaired both at the level of OPC proliferation and on the level of regeneration of mature oligodendrocytes (Miyamoto et al., 2013).
Transcriptomic and electrophysiological approaches have revealed that OPCs exhibit alterations of ion channels throughout aging, affecting their differentiation potential (Spitzer et al., 2019).Finally, epigenetic control of the OL differentiation process has also been shown to be affected by age (Shen et al., 2008).
Using cFPI in mice, proliferation of mitotic Olig2+ cells was observed in several white matter regions at one-week post-injury (Flygt et al., 2017), results confirmed in the present study.Olig2 is a transcription factor that is expressed throughout the whole OL maturation process and drives the progression of the oligodendrocytic lineage (Dimou et al., 2008;Kuhn et al., 2019;Wegener et al., 2015).
In addition, an increased number of EdU-labeled Olig2+ cells was observed in the white matter using a focal TBI model (the controlled cortical impact; CCI) (Dent et al., 2015).Olig2 and Sox10 synergistically act in order to promote progression of the oligodendrocyte lineage, and both are constitutively expressed throughout all the stages of oligodendrocyte development (Liu et al., 2007;Sock & Wegner, 2021).Olig2 expression may thus highly overlap with the expression of Sox10.Our results support the notion that the origin of these mitotic Olig2+ cells in the injured white matter can be distinguished into two separate cell populations, a PDGFRα+ and a Nestin + progenitor population, the first representing proliferating residing OPCs within the white matter and the latter indicating migrating progenitors derived from different gliogenic brain regions.
PDGFRα+ glia has been suggested to constitute a population of OPCs which are able to give rise to mature oligodendrocytes in the adult CNS (Nishiyama et al., 2014;Rivers et al., 2008).In contrast to other studies (Bye et al., 2011;Flygt et al., 2017;Sullivan et al., 2013;Susarla et al., 2014;von Streitberg et al., 2021), we did not observe an increased number of proliferating PDGFRα+ cells at 2 dpi and 7 dpi in young mice.This could be explained by the use of different experimental TBI models as well as the investigation of different brain regions.On the other hand, aged mice exhibited an increase in proliferating PDGFRα+ cells following cFPI.Differentiation of PDGFRα+ OPCs along the progression of the oligodendrocyte lineage involves the downregulation of the PDGFRα marker (Chittajallu et al., 2005;Rivers et al., 2008).This suggests that proliferative signals by PDGFRα + cells might have occurred prior to 2 dpi in young mice.Additionally, proliferating OPCs that incorporated EdU might have down-regulated PDGFRα by 7 dpi.The fact that the significant increase of PDGFRα +/EdU+ seen in aged brain-injured mice was not observed in young brain-injured mice supports the idea of a delayed downregulation of PDGFRα marker in proliferating OPCs in aged mice.It also indicates that oligodendrocyte maturation might occur earlier after TBI in the young when compared to aged mice.
Studies on experimental models of multiple sclerosis and adult oligodendrogenesis have revealed that Nestin+ cells, derived from either the subventricular zone (SVZ) or the meningeal perivascular space, contribute to the regeneration of oligodendroglia (Đặng et al., 2019;Kang et al., 2019;Nait-Oumesmar et al., 1999;Xing et al., 2014).This indicates that Nestin+ cells, alternatively considered as neural stem cells (NSCs) within neurogenic regions of the adult rodent brain, are capable of transiting to an OPC fate and partaking in the formation of mature oligodendrocytes.Furthermore, Nestin expression and NSC proliferation are decreased in the SVZ layer of 12-months old mice when compared to 1-month old mice (Wang et al., 2016).These findings support our present results, indicating  et al., 2007;Sun et al., 2012).As such, the observed subset of PCNA+/Nestin+/Olig2+ cells may indicate that NSCs within the white matter transit to an OPC fate through the expression of Olig2 (Pieczonka et al., 2023).
The number of proliferating PDGFRα+ cells post-injury was similar in young and aged animals.However, the number of cells in this population was smaller than the overall number of proliferating Olig2+ cells in young brain-injured mice.This indicates that even though proliferating PDGFRα+ cells appear in the aged mouse brain, their proliferative capacity is limited.In vivo studies have reported both a reduction of Olig2+ cells and Olig2+ mRNA levels in the white matter of aging mice (Doucette et al., 2010;Kaya et al., 2022).
Likewise, several studies have shown that neither the number of PDGFRα+ cells in the white matter, nor the expression of the PDGFRα marker is affected by age (Doucette et al., 2010;Kaya et al., 2022;Miyamoto et al., 2013;Sim et al., 2002).Even though we observed Nestin+ cells to be highly proliferative, as indicated by the expression of PCNA, these cells did not express PDGFRα suggesting that they may respond to different growth signals.These signals might, in turn, be affected by aging.
Oligodendrocyte differentiation is a highly intricate process involving the interplay of several transcription factors and regulators such as Olig2, Sox10, Myrf, Tcf712, Nkx2.2 and Ascl1 (Guo & Wang, 2023;Nicolay et al., 2007;Sock & Wegner, 2021).The results of our transcript analysis point towards an unaltered expression of oligodendrocyte-related transcription factors in young mice compared to aged mice, potentially indicating disturbances in the oligodendroglial regulatory network of the aging brain.From a developmental aspect, Ascl1 seems to exert its function on the proliferation of OPCs rather than on their differentiation (Kelenis et al., 2018).Conversely, Nkx2.2, Myrf and Tcf712 seem to promote progression of the differentiation process (Guo & Wang, 2023;Vogl et al., 2013).The up regulation of Ascl1 on 2 dpi seen in aged mice coincides with the increased number of proliferating PDGRFα+ cells, potentially linked to prolonged proliferation of these cells as well as a delay or incapability to down-regulate PDGFRα due to increased Ascl1 activity.
Furthermore, transcript up-regulation of Myrf, an autoproteolytically cleaved differentiating factor (Huang et al., 2018), might act as a compensatory mechanism due to reduced functionality of the protein.
Additionally, in situ hybridization data on the corpus callosum of young and aged monkeys have revealed reduced expression of Nkx2.2 in aging OPCs (Dimovasili et al., 2023).On 7 dpi, our gene expression data show a general down-regulation of differentiationpromoting genes coinciding with reduced number of newly generated EdU+ oligodendrocytes.Additional experiments are needed to firmly establish whether this impaired regeneration is caused by diminished OPC proliferation or delayed OL maturation.
Regarding the high accumulation of EdU+ cells in the white matter tracts post-injury, we demonstrate that a significant portion of these cells are Iba1+ cells.As documented in several studies (Kato et al., 2003;Susarla et al., 2014;Swirski et al., 2014), a proportion of the PCNA+ cells at 2dpi are microglia/macrophages.A recent study using flow cytometry on aged mice after a controlled cortical impact (CCI) injury found, similar to in our present study, increased microglial proliferation and number of invading leukocytes (Ritzel et al., 2019).
Increased microglial proliferation may reflect an exacerbated immunological response observed in injured aged mice.Specifically, increased PCNA+ cells in the injured cFPI mice at 7 dpi might indicate prolonged proliferation of Iba1+ cells.Furthermore, since Nestin is detected in microglia under other neuroinflammatory conditions (Krishnasamy et al., 2017), a subset of PCNA+/Nestin+ cells seen in our present study might represent microglial cells.This study is not without limitations.In the majority of studies, including this one, in which male rodents are widely used, sex-related differences regarding the functionality of OPCs following TBI have not been elucidated.In vitro and in vivo studies of neonatal rat brains have revealed differential response of OPCs and OLs to sex hormones as well as transcriptomic differences of OPCs derived from male and female rats (Swamydas et al., 2009;Yasuda et al., 2020).As such, investigation into potentially different responses in male and female animals post-TBI is needed.Additionally, since the design of this study is limited to the investigation of two time-points, 2dpi and 7dpi, alterations at long-term time points need to be evaluated.Furthermore, numerous markers are expressed during oligodendrogenesis, not all of which were analyzed in our study.The use of a wider range of markers in future studies could allow us to further characterize the identity of the proliferating cells but to also track their maturation process during later time-points post-injury.
Together, this study provides evidence of a reduced proliferative response of OPCs in the aged mouse brain following TBI.This may contribute to a reduced ability for replacement of lost mature oligodendrocytes, and for myelin repair, which may influence functional recovery post-TBI.Enhancing OPC proliferation might be a promising pharmacological treatment target.In future studies, remyelination and maturation of OPCs should be further explored in order to define how OPC proliferation following TBI is associated with functional remyelination.

F
I G U R E 1 Reduced number of PCNA+ and PCNA+/Olig2+ cells in the white matter tracts of naïve aged mice.(a) Cell counts of immunofluorescence PCNA staining combined with DAPI counterstain at 20Â magnification in the corpus callosum and external capsule of young and aged mice.[young n = 4, aged n = 4] (b) Cell counts of co-labeled PCNA/Olig2 cells in the white matter regions of young and aged mice.(c, d) Representative immunofluorescence images of PCNA/DAPI (c) and PCNA/Olig2/DAPI (d) staining in the corpus callosum and external capsule of young and aged naïve mice.Arrows point to proliferating cells expressing high levels of the PCNA protein (c) and PCNA+/Olig2+ cells in (d).
Following the five-day labelling period with EdU injections, we observed an increase of EdU+ cells in cFPI-injured mice of both age F I G U R E 2 TBI induces an increased number of PCNA+ cells in the white matter of young and aged mice.(a) Experimental design involving two approaches to evaluate proliferation and graphical illustration of a mouse coronal section depicting the regions of interest in which cell quantification was conducted.Blue arrows indicate quantification of PCNA+ cells at day 2 (2dpi) and day 7 (7dpi) post-injury.Red arrows indicate daily labelling of cells with intraperitoneal EdU injections (50 mg/kg) starting from 2dpi until 6dpi while the yellow arrow indicates quantification of EdU+ cells at 7dpi.(b) Representative epifluorescence 20Â magnification images of PCNA staining in young and aged mice at 2 dpi.(c) Quantification of PCNA+/DAPI cells at 2 and 7 dpi post-injury in both age groups.*p < .05,**p < .01,***p < .001and ****p < .0001,respectively.Graphical illustrations were made with BioRender.cFPI, central fluid percussion injury.

F
I G U R E 3 A reduced number of co-labeled PCNA+/Olig2+ cells in the white matter of aged mice at 2 dpi following cFPI.(a) Representative 20Â magnification confocal images of PCNA/Olig2 co-staining from the corpus callosum of young cFPI and aged cFPI mice at 2 dpi.(b) Confocal 40Â magnification composite z-stack image in the external capsule of a young, brain-injured mouse at 2 dpi depicting DAPI/PCNA/Olig2 triplepositive cells.The inbox image contains the region from which the 40Â image was taken in 20Â magnification.Arrows point to co-labeled PCNA/ Olig2-positive cells.(c, d) Quantification of PCNA+/Olig2+ cells in mice at 2 dpi and at 7 dpi, respectively.*p < .05,**p < .01,***p < .001and ****p < .0001,respectively.cFPI-central fluid percussion injury.
(a, b) Representative 20Â magnification images of EdU/Olig2 co-staining from the corpus callosum and the external capsule of sham-injured and cFPI mice of both age groups at 7 dpi.(c) Quantification of EdU+/Olig2+ cells post-injury in both age groups at 7 dpi.(d) Representative confocal 40Â magnification z-stack image in the corpus callosum of a young cFPI mouse including individual images from each fluorescent channel.Arrows point to EdU/Olig2-positive cells.*p < .05,**p < .01,***p < .001and ****p < .0001,respectively.cFPI-central fluid percussion injury.

F
I G U R E 5 Increased number of Olig2 High cells in the white matter of young mice following TBI and no alterations in the number of Olig2 Low on 2 and 7 dpi.Based on the fluorescence intensity of their nuclei, Olig2+ cells were automatically classified into Olig2 Low and Olig2 High based on the application of a Low and a High intensity threshold, respectively.(a) Workflow regarding the image analysis of Olig2/DAPI staining images for the quantification and classification of Olig2+ cells.(b) A representative Olig2+/DAPI composite epifluorescence image of the corpus callosum throughout different steps of the workflow, including background subtraction, free-hand selection of the region of interest and application of the respective threshold.(c) Quantification of Olig2 Low and Olig2 High in the corpus callosum and the external capsule in young and aged mice on 2 and 7 dpi, respectively.Data are presented as mean ± S.E.M. *p < .05,**p < .01,***p < .001and ****p < .0001,respectively.cFPI-central fluid percussion injury.

F
I G U R E 6 TBI induces an increased number of co-labeled PCNA+/PDGFRα+ cells in the corpus callosum of aged mice but not of young mice at 2 dpi.(a, b) Representative 20Â magnification confocal images of PCNA/PDGFRα co-staining from the corpus callosum of young cFPI and aged cFPI mice at 2dpi, respectively.(c) Quantification of PCNA+/PDGFRα+ cells at 2 dpi in both age groups.(d) Confocal 40Â magnification composite z-stack image in the external capsule of a young and aged injured mouse at 2dpi depicting DAPI/PCNA/PDGFRα triple-positive cells.Arrows point to PCNA/PDGFRα-positive cells.*p < .05,**p < .01,***p < .001and ***p < .0001,respectively.cFPI-central fluid percussion injury.

F
I G U R E 7 A TBI-induced increase of co-labeled EdU+/PDGFRα+ cells in the corpus callosum of aged mice is not seen in young mice at 7 dpi following cFPI.(a) Representative 20Â magnification confocal images of EdU/PDGFRα co-staining from the corpus callosum of shaminjured and cFPI mice of both age groups at 7 dpi.(b) Quantification of EdU+/PDGFRα+ cells post-injury in both age groups at 7dpi.(c, d) Representative 20Â magnification images of EdU/PDGFRα co-staining in the external capsule and confocal 40Â magnification z-stack image of the external capsule in young cFPI and aged cFPI mice.Arrows point to EdU/PDGFRα-positive cells.*p < .05,**p < .01,***p < .001and ***p < .0001,respectively.cFPI-central fluid percussion injury.
This finding indicates that inflammatory cells such as macrophages and microglia undergo proliferation or are involved in phagocytosis of cells which have incorporated EdU, either through F I G U R E 8 Proliferating Olig2+/PDGFRα-cells can be detected in the white matter following cFPI.(a) 20Â magnification images of the two white matter regions of interest, the corpus callosum and the external capsule, showing a substantial overlap between the markers Olig2 and PDGFRα in both young and aged injured mice.(b) Triple-immunofluorescent 20Â and 40Â magnification confocal images depicting PCNA+ cells (open arrows) co-localizing with both markers of interest (Olig2 and PDGFRα) in both age groups.Inserts depict the region to which 40Â magnification was applied.A population of PCNA+ cells (arrows) positive for Olig2 but negative for PDGFRα was observed by 40Â magnification confocal imaging in young and aged cFPI mice.cFPI-central fluid percussion injury.

F
I G U R E 9 Decreased number of PCNA+/Nestin+/Olig2+ progenitors in the white matter of aged mice on 2 dpi following cFPI.(a) Representative 20Â magnification images of PCNA/Nestin/Olig2 co-staining from the corpus callosum of sham-injured and cFPI mice of both age groups at 2 dpi.(b) 40Â magnification images of PCNA/Nestin/Olig2 co-staining in both white matter regions of interest.(c) 63Â magnification images of PCNA+/Nestin+/Olig2+ cells.The upper image depicts an orthogonal view of a sectioning plane approximately half-way through the acquired z-stack.Lower image depicts the maximum projection of the acquired z-stack.(d) Quantification of PCNA+/Nestin+ and PCNA+/Nestin+/Olig2+ cells post-injury in both age groups at 2dpi.Arrows point to triple-positive cells.*p < .05,**p < .01,***p < .001and ***p < .0001,respectively.cFPI-central fluid percussion injury.et al., 2001), oligodendrogenesis has been much less investigated in the adult and aging CNS, especially in the context of TBI.Evidence from different scientific fields converge to suggest reduced oligodendrogenesis in the aging rodent brain.Specifically, chemically induced focal demyelination or cuprizone-induced demyelination studies show

F
I G U R E 1 1 Expression of oligodendrocyte-related transcription factors on 2 and 7 dpi following cFPI.Graphical illustration represents the region of interest which was dissected for RNA extraction.Transcript levels of the transcription factors are depicted for day 2 and day 7 postinjury, respectively.Data are normalized and centered to the expression level of sham-injured young mice and are shown as log 2 fold-change.Groups labeled with [#] exhibit significant difference when compared to the other groups.Groups labeled with [s] exhibit significant difference compared to the sham-injured, young group.[day 7 -cFPI aged vs cFPI young: Ascl1 ( p < .0001),Myrf ( p < .05),Nkx2.2 (p < .05),Tcf712 ( p < .001).Graphical illustration was made with BioRender.*p < .05,**p < .01,***p < .001and ***p < .0001,respectively.cFPI-central fluid percussion injury.that the contribution of immature Nestin+ progenitors, potentially originating from the SVZ, in the regeneration of oligodendroglia in the aging white matter is diminished following TBI.Olig2 and Sox10 are two transcription factors which possess essential roles in oligodendrocyte differentiation and determining cell fate (Liu

F
I G U R E 1 2 Increased number of EdU+/Iba1+ cells in the corpus callosum of aged mice following cFPI.A proportion of the total EdU+ cells was observed to co-localize with Iba1 in cFPI mice of both age groups, in the corpus callosum (a) and the external capsule (b).The graphical inserts (BioRender) within the images pinpoint to the relative position of the depicted region.(c) Quantification of Iba1+/EdU+ cells post-injury in both age groups at 7dpi.*p < .05,**p < .01,***p < .001and ***p < .0001,respectively.cFPI-central fluid percussion injury.