Heterozygous ^{Ascl1 − CreRT2} mice [Ascl1^∧tm1(Cre/ERT2)Jejo/J; stock #012882] (Kim et al., 2007) and homozygous Ai14 mice [B6.Cg-Gt(ROSA)26Sortm14(CAG-tdTomato)Hze/J; stock #007914] (Madisen et al., 2010) obtained from The Jackson Laboratory (Bar Harbor, ME) were crossed to generate Ascl1-CreERT2; CAG-FloxStopTom mice. Mice were maintained on a C57Bl/6J background. Mice were provided with food and water ad libitum at the University of Kentucky Medical Center animal vivarium and were housed up to five mice per cage at a constant temperature (23 ± 2°C) with a 14/10 h light/dark cycle. All animal procedures were approved by the University of Kentucky's Institutional Animal Care and Use Committee under IACUC protocol 2019-3293.

Young adult (8–10 weeks of age) male and female mice were used except where noted. Mice were matched for age, rather than weight, since adult hippocampal neurogenesis is strongly regulated by age (Yang et al., 2015).

For long-term neurogenesis studies, male and female Ascl1-CreERT2; CAG-FloxStopTom mice were assigned to CCI (n = 10 male, 9 female) or naive (n = 8 male, 11 female) groups. Surgeries were performed as previously described (Madathil et al., 2013). Briefly, the head of a mouse anesthetized using 3% isoflurane was secured in a stereotaxic frame. Anesthesia was maintained using 2.5% isoflurane delivered through a nose cone. A midline scalp incision was made and a 5 mm diameter craniectomy was performed over the left parietal cortex, lateral to the sagittal suture and midway between Bregma and Lambda. The exposed cortex was contused using a pneumatically controlled impacting device (TBI-0310, Precision Systems and Instrumentation, Lexington, KY) which drives a piston with a 3 mm diameter rounded tip at a rate of 3.5 m/s to a depth of 1.0 mm to produce a moderate injury. A moderate injury is defined here as cortical tissue loss in the absence of hippocampal ablation, leaving both pyramidal and GC layers grossly intact. After CCI injury, a circular disk made from dental cement was glued over the craniotomy to protect the brain surface and the scalp was closed. Mice received a 1 mL injection of saline subcutaneously for hydration and were placed on a heating pad to maintain body temperature until they recovered from anesthesia. Mice were then returned to their home cages. Naive mice did not undergo anesthesia or any surgical procedures.

Mice were injected intraperitoneally with 120 mg/kg BW tamoxifen (Sigma, St. Louis, MO) in corn oil on two consecutive days to induce Cre recombination. Naive mice were injected at 8 weeks of age. Injured mice were injected on days 2 and 3 after CCI to induce recombination in Ascl1-expressing cells born acutely after injury.

For examination of acute proliferation, a separate cohort of 9–11 week old male double transgenic reporter mice (n = 6 Ascl1-CreERT2; CAG-FloxStopTom) and tdTomato mice (n = 8 CAG-FloxStopTom) received CCI injury followed by tamoxifen injections as above. Mice were then injected intraperitoneally with 50 mg/kg of 5-ethynyl-2′-deoxyuridine (EdU; cat. A10044, Thermo Fisher Scientific, Waltham, MA) on days 3, 4, and 5 or days 6, 7, and 8 after injury. Mice were euthanized on day 10.

Animals were euthanized at 10 days or 6 weeks after CCI or tamoxifen injection (naïve) by sodium pentobarbital (150 mg/kg i.p. Fatal-plus solution, Vortech Pharmaceuticals, LTD) and transcardially perfused with heparinized saline followed by 10% buffered formalin. Brains were removed from skulls 24 h after post-fixation in 10% formalin, further post-fixed for 24 h, and then cryoprotected using 30% sucrose solution. Brains were snap frozen in cold 2-methylbutane (< -25°C) and sectioned coronally at 40 μm thickness. Where possible, analyses were performed by individuals blinded to relevant independent variables such as sex, injury condition, genotype and time point of the samples. In many cases, it was not possible to be truly blinded to injury condition or hemisphere (ipsilateral/contralateral) due to obvious differences in the hippocampal formation ipsilateral to the CCI.

For analysis of cell proliferation, three sections containing dorsal hippocampus were selected at 400 μm intervals for each animal. Free-floating sections were reacted using a Click-iT^tm EdU Cell Proliferation Kit (cat. C10337, Thermo Fisher Scientific). TrueBlack (cat. 23007, Biotium, Fremont, CA) was then applied according to the manufacturer's instructions to quench autofluorescence. Slides were coverslipped with DAPI Fluoromount-G (Southern Biotech, Birmingham, AL) and stored at 4°C.

For contusion volume analysis, brain sections selected at 400 μm intervals spanning the rostral-caudal extent of the contusion were mounted on slides and stained for Nissl substance. A digital slide scanner (Zeiss Axio Sxan Z.1, Oberkochen, Germany) was used to image Nissl-stained sections at 20 × magnification to create a single high-resolution digital image. Halo software (version 2.3; Indica Labs, Albuquerque, NM) was used to analyze the images. For each section, the ipsilateral neocortical area containing surviving neurons, verified by examination of cell morphology, and the entire contralateral neocortical area were traced manually. Contusion area was calculated as the difference between the contralateral and spared ipsilateral neocortical areas. The contusion volume was then calculated using Cavalieri's principle.

Free-floating immunohistochemistry was performed to co-label tdTom+ neurons with the mature neuronal marker anti-NeuN (monoclonal mouse, 1:400, Novus NBP1-92693, St. Charles, MO) or the immature neuron marker anti-DCX (polyclonal rabbit, 1:2500, Abcam ab18723). Coronal sections were initially rinsed in TBS then blocked in 5% normal goat serum and 0.1% Triton-X-100 in TBS for 30 min at room temperature. Primary antibody was applied overnight at 4°C. The following day tissue was rinsed with TBS and incubated with Alexa Fluor 488 conjugated secondary antibody (1:500, Invitrogen, Carlsbad, CA) for 1 h at room temperature, and then rinsed with TBS. Sections were mounted on gelatin-coated slides which were then coverslipped with DAPI Fluoromount-G and stored at 4°C.

Statistical analysis was conducted in either Prism or SAS. Counts of EdU+ cells, body weights and contusion volumes were analyzed with an unpaired two-tailed t-test with pooled variance. Counts of tdTom+ neurons were normalized to the length of the DG (mm). Normalized cell counts were log transformed due to right skewness and a mixed model was fit with a random effect to account for correlated observations within injured mice (ipsilateral and contralateral) and with fixed effects for sex, condition (CCI ipsilateral, CCI contralateral, and naïve), and the two-way interaction. Degrees of freedom were adjusted using the Kenward-Rogers adjustment. Sholl analysis data were analyzed using a 2-way repeated measures ANOVA with Fishers LSD for post-hoc comparisons. Length of primary dendrite and total dendrite length were analyzed using a 1-way ANOVA with Fishers LSD for post-hoc comparisons. MFB datasets were log transformed due to right skewness prior to conducting parametric 1-way ANOVAs. Parametric analysis of MFB parameters was corroborated using non-parametric Kruskal-Wallis ANOVAs; results were congruent except where noted in the Results. For 1-way ANOVA analyses only selected post-hoc tests were performed, as determined a priori; these included comparing sex within each injury condition and comparing injury condition within each sex. For contusion volume, measurements could not be completed on one female with CCI due to tissue tearing and folding. A subset of female naïve mice (n = 7–8) was randomly selected for Sholl tracing and MFB analyses. One male CCI mouse was excluded from the Sholl analysis and MFB volume measurements because it had no tdTom+ neurons in the DG or MFBs in the CA3 region of interest. Data are reported as mean and SEM. Graphs depict means and individual data points, with error bars representing SEM.

To genetically label NPCs with a fluorescent dye, tamoxifen-inducible Ascl1-CreERT2; CAG-FloxStopTom mice were used. Ascl1 is a proneural transcription factor that is transiently expressed in a subset of NPCs in the peripheral and central nervous system (Guillemot and Joyner, 1993; Wapinski et al., 2013). In the hippocampal DG of adult mice, the majority of cells expressing Ascl1 are Type-2A early NPCs (Kim et al., 2007). Ascl1 expression is downregulated as the cell commits to a neuronal fate, designated as Type-2B (Figure 1A). A subset of Type-1 activated NSCs express Ascl1 (Andersen et al., 2014; Kim et al., 2007) and can be identified by co-labeling with GFAP.

Ten days after tamoxifen induction in naïve mice, the majority of tdTom+ cells in the DG phenotype as neuroblasts or immature neurons as defined by their expression of the microtubule-associated protein doublecortin (DCX) (Figure 1B). As adult-born neurons mature, they elaborate dendritic arbors, downregulate expression of DCX and express the mature neuronal marker NeuN. Accordingly, at 6 weeks after inducing Cre-recombination in Ascl1 expressing NPCs, tdTom+ cells were no longer DCX+ (Figure 1C) but co-labeled with NeuN (Figure 1D). Retention of fluorescent dye in cells as they developed into fully mature GCs allowed visualization and assessment of the cells' morphological development of dendrites as well as their axonal projections to the CA3 region.

Because the Ascl1-CreERT2 mouse line used here has only one functioning Ascl1 allele (Kim et al., 2007) and Ascl1 is required for NSC/NPC activation, we compared cellular proliferation in Ascl1^+/+ mice (CAG-FloxStopTom mice) and Ascl^+/− mice (Ascl1-CreERT2; CAG-FloxStopTom). Following EdU injections on days 3, 4, and 5 postinjury to capture cells proliferating early after tamoxifen induction, no difference was noted in numbers of EdU+ cells in the SGZ and inner third of the GC layer of the ipsilateral hippocampus at 10 days following injury (Figure 1E). Similarly, proliferation on days 6, 7, and 8 after injury was equivalent in the two groups of mice (Figure 1F). This suggests that the Ascl1 heterozygosity of the double transgenic reporter mice did not influence cellular proliferation within the neurogenic niche across the first week after CCI.

CCI is a widely used model of contusive brain injury resulting in a progressive cortical neuron death within a hemorrhagic contusion leading to the formation of a cavity (Williams et al., 2022). We sought to determine whether age-matched male and female mice might have different degrees of cortical damage, which could potentially confound interpretation of neurogenic responses as the extent of hippocampal neurogenesis is dependent on injury severity (Wang et al., 2016). At 8–10 weeks of age, female mice weighed significantly less (20.7 ± 1.0 g) than male mice (25.8 ± 1.9 g) at the time of injury (p < 0.0001). Nonetheless, 3 days after injury males and females exhibited similar weight gain (males 0.8 ± 0.8 g, females 0.8 ± 1.1 g). Pathological features in injured males (Figure 2A) and females (Figure 2B) were typical of moderate severity CCI. There was extensive neocortical tissue loss and cavitation ipsilateral to impact while the underlying hippocampal formation remained grossly intact, with no ablation of the pyramidal or GC layers. Using the Cavalieri method, volumetric estimates of the contusion within the neocortex were performed. Age-matched male and female mice had equivalent cortical contusion volumes 6 weeks after CCI injury (Figure 2C).

To examine how sex influences numbers of GCs born acutely after injury that survive to maturity, the number of tdTom+ neurons in the DG GC layer was quantified in male and female naïve mice and CCI-injured mice at 6 weeks after injury. In naïve mice, numerous tdTom+ neurons were localized in the inner one third of the GC layer (Figures 3A, D). Labeled adult-born neurons had round cell somas with well-developed dendritic arbors extending through the molecular layer. In CCI-injured mice, posttrauma-born neurons were less abundant in the ipsilateral DG in both males and females. Their somas were localized within the GC layer in both the contralateral and ipsilateral DG, although neurons that had migrated to the outer two-thirds of the GC layer were more frequently observed than in naïve mice (Figures 3B, C, E, F). This deeper migration was most notable in the contralateral DG of female mice, which also showed more abundant tdTom+ neurons (Figure 3E). Dendrites of posttrauma-born neurons could be visualized extending well into the molecular layer.

Quantification of numbers of tdTom+ neurons revealed a main effect of sex [F_{(1, 33.8)} = 21.87; p < 0.0001] with females demonstrating increased neuron numbers. There was a main effect of the condition [F_{(2, 27.7)} = 87.09; p < 0.0001] and an interaction of sex and condition [F_{(2, 27.7)} = 7.42; p = 0.0026]. Post-hoc analyses revealed that both male and female CCI-injured mice had significantly fewer tdTom+ neurons in the ipsilateral hippocampus (Figures 3C, F) compared either to naïve controls (Figures 3A, D) (males p < 0.0001, females p < 0.0067) or to cell counts in the contralateral hippocampus (Figures 3B, E) (males p < 0.0001, females p < 0.0001), consistent with a trauma-induced decrease in neurogenesis in the ipsilateral DG. Males showed a more pronounced deficit in posttraumatic neurogenesis, with significantly fewer tdTom+ cells in the ipsilateral hippocampus compared to females (p < 0.0001; Figure 3G). Neurogenesis in the contralateral hippocampus of CCI-injured mice was not different from that in naïve mice for either males or females. However, the number of tdTom+ neurons in the contralateral hippocampus of CCI-injured females was significantly greater than the number in the contralateral hippocampus of CCI-injured males (p = 0.0013; Figure 3G), pointing to sexually dimorphic neurogenic responses in the contralateral as well as the ipsilateral hippocampus.

tdTom+ GCs in naïve mice and in the contralateral DG of injured mice displayed an apical dendrite that extended through the GC layer and initiated branching in the inner molecular layer (Figures 4A, B). However, dendrites of tdTom+ GCs in ipsilateral DG of injured mice often began branching within the GC layer (Figures 4C, D; see also Figure 3). To examine how TBI influences the development of dendritic arbors of GCs born acutely after injury that survive to maturity, tdTom+ neurons in the upper blade of the GC layer were digitally traced and reconstructed for Sholl analysis. Dendritic complexity did not differ between male and female naïve mice [F_{(30, 390)} = 0.8615; p = 0.6796] (Figure 4E). Compared to same sex naïve controls, both male and female CCI-injured mice had significantly altered dendrite complexity [male: F_{(30, 450)} = 4.876; p < 0.0001, female: F_{(30, 420)} = 3.492; p < 0.0001]. Although there was no significant difference in dendrite complexity between male and female CCI-injured mice when comparing across the entire length of the arbor [F_{(30, 480)} = 1.307; p = 0.1308], the Sholl analysis curves suggested unique changes for males and females. In injured males, the Sholl curve showed a more marked downward shift across distances between 70 and 200 μm from the soma (Figure 4E). Branching was significantly less from 80 to 190 μm from the cell soma while in CCI-injured female mice branching was only significantly reduced at 130 μm. Conversely, dendritic branching was increased significantly increased at 230–270 μm from the cell soma in female mice and at 30 μm in male mice.

Adult-born GCs in naïve females had a significantly shorter distance to the first branch point than those in naïve males (p < 0.05; Figure 4F) although the total dendritic length of all dendrite branches was comparable (Figure 4G). In CCI-injured male mice the primary dendrite was significantly shorter compared to naïve males [F_{(3, 29)} = 3.623; p = 0.0246] (p = 0.003). In contrast, CCI-injured females had no change in length of the primary dendrite (Figure 4F). Similarly, posttrauma-born GCs in males developed with dendritic arbors of significantly decreased total length compared to GCs born in naïve males [F_{(3, 29)} = 4.022; p = 0.0165] (p = 0.044, Figure 4G), while branch length was unaffected by CCI in females. Further, GCs born after injury in females had significantly greater total dendritic length than those in males (p = 0.0046).

To determine if TBI alters MFB number or volume in the CA3 region, MFBs were reconstructed in 3D from high resolution confocal images (Figures 5A, A', B, B'). Far fewer tdTomato+ MFBs were visible in the hippocampus of male mice with CCI when compared to all other groups. Naïve mice demonstrated no sex-dependent differences in numbers of MFBs in the CA3 region, MFB surface volume, or the relative ratio of numbers of MFB to numbers of tdTom+ GC (Figures 5C–E). Numbers of MFBs varied significantly across groups [F_{(3, 30)} = 14.91; p < 0.0001]. Post-hoc tests revealed numbers of MFB in the CA3 region of male mice with CCI was significantly reduced to 39 ± 11 from 253 ± 53 in naïve controls (p < 0.0001) (Figure 5C). Numbers of MFBs in female mice with CCI (173 ± 67) were somewhat lower than in naïve controls (210 ± 20). This decrease was not statistically significant by ANOVA after log transformation (p = 0.132) but was significant using nonparametric analysis (p = 0.049). Brain-injured females had significantly greater numbers of MFBs from posttrauma-born neurons than did brain-injured males (p = 0.0002). Only CCI-injured female mice displayed increased MFB surface volume (17.6 ± 2.2 μm³) when compared to sex-matched naïve controls (11.8 ± 0.49 μm³) (p = 0.004) [F_{(3, 30)} = 3.436; p = 0.0293] (Figure 5D). The average MFB surface volume in male mice was unchanged after CCI. To determine if the decrease in MFB numbers in the CA3 region was proportional to the decrease in tdTom+ GCs in the DG after TBI, we normalized the number of MFB to the number of tdTom+ GCs for each mouse. Despite a >50% increase in MFB numbers per GC in CCI-injured females (7.40 ± 1.34) compared to males (4.75 ± 1.36), the numbers of MFB per GC did not vary statistically across groups [F_{(3, 30)} = 1.057; p = 0.3820] (Figure 5E).