The animal studies were reviewed and approved by the Institutional Animal Care and Use Committee, and the studies were conducted per the Guide for the Care and Use of Laboratory Animals and the ARRIVE guidelines for reporting animal experiments. Mice were housed in a pathogen-free setting at a 12-h light/12-h dark cycle. Food and water were provided ad libitum.

To generate microglia-specific conditional HDAC3 knockout mice, HDAC3 floxed mice, with LoxP sites on either side of HDAC3 exon 7 [HDAC3^fl/fl; originally generated by Dr. Scott W. Hiebert lab (Bhaskara et al., 2008) and acquired from Dr. Meghan E McGee-Lawrence’s lab at Augusta University] were bred with Cx3cr1CreER or Cx3cr1 mice (Stock No: 021160; The Jackson Laboratory, United States). The genotypes of the offspring were confirmed by PCR, using genomic DNA extracted from mice tails with a kit from Qiagen (Hilden, Germany). The 3-week-old mice (both the conditional knock out and the experimental control) were administered Tamoxifen (TAM, 75 mg/kg i.p. per day; four alternate days). To verify genetic recombination in brain tissue, DNA was isolated from the mouse brain tissue using a Thermo Scientific GeneJET DNA purification kit, and PCR was carried out on the purified DNA using HDAC3 primers. The primer sequences used for PCR are provided in Supplementary Table 1.

For studies, mice were randomly assigned. The power analysis (PASS software, NCSS LLC) was conducted using previous data that implied n = 6 for RT-PCR to obtain a power of 80%. The study employed 172 mice in total. Of those, nine mice were excluded. The exclusion criteria included the absence of a hematoma or animal death. The data from male and female subjects were separately analyzed to determine if there is a sex-dependent effect on the functional role of microglial HDAC3 after ICH.

Male and female mice (8–12 weeks old) were anesthetized with an i.p. injection of ketamine and xylazine (100 mg/kg and 10 mg/kg, respectively). The mice were then secured on a stereotaxic head frame (Stoelting, WI, United States), and a small animal temperature controller maintained the body temperature at 37 ± 0.5 °C. For a cranial opening, a high-speed dental drill (Dremel, United States) was used, and the burr hole was made 2.2 mm lateral to the bregma. A 26-G Hamilton syringe containing 0.04 U of type IV bacterial collagenase (Sigma, St. Louis, MO, United States) in 0.5-μL saline was inserted 3.0 mm into the left striatum through the burr hole to induce spontaneous brain hemorrhage (Sukumari-Ramesh et al., 2012a, 2012b). For sham or experimental control, animals underwent the surgical procedure as detailed above, but 0.5 μL saline was injected into the brain striatum. Upon removal of the needle, the burr hole was sealed with bone wax, and the incision was closed. Mice were kept at 37 °C until recovery.

The neurobehavioral outcome was determined using a 24-point scale (Wu et al., 2012; King et al., 2011). The scale comprises six individual behavioral tests: climbing, circling, compulsory circling, whisker response, bilateral grasp, and beam walking. Each test was scored from 0 (performs with no impairment) to 4 (severe impairment). The sum of the scores on all six tests, or the composite neurological deficit score, is a measure of neuroglial deficits (sensory-motor deficits). The neurobehavioral analysis was carried out independently by an investigator blinded to the experimental groups.

Mice were anesthetized deeply with isoflurane and perfused transcardially with ice-cold phosphate-buffered saline, pH 7.4 (PBS). Three millimeters brain sections containing the hematomal and perihematomal brain regions post-ICH were collected using a brain matrix and flash-frozen in liquid nitrogen. Brain tissues were then homogenized in 800 μL Trizol reagent (Invitrogen, Catalogue number: 15596018). The homogenate was then incubated for 5 min at room temperature before adding 160 μL chloroform. Samples were then centrifuged at 12,000 g for 15 min at 4 °C, and the aqueous phase was collected into fresh tubes. Four hundred microliters isopropanol was added to the aqueous phase, and the solution was incubated for 10 min at room temperature before centrifugation at 12,000 g for 10 min at 4 °C. The supernatant was decanted, and the remaining pellet was washed in 75% ethanol. Samples were then centrifuged at 7,500 g for 5 min. The supernatant was decanted, and the RNA pellets were air-dried for 10 min. Pellets were then resuspended in nuclease-free water and incubated in a heat block at 70 °C for 15 min. Quantitative RT-PCR was performed on a Bio-Rad C1000 Touch Thermal Cycler using a GoScript reverse transcriptase kit (Promega, Madison, WI) and SsoAdvanced Universal SYBR Green Supermix (BioRad, Hercules, CA). The RNA expression of Nos2, S100A8, S100A9, IL-1β, IL-6, TNF-α, Arg-1, and CD206 was evaluated. Data were normalized to GAPDH, a housekeeping gene, and expressed as fold change compared with the control. The primer sequences used for the quantitative RT-PCR are provided as part of Supplementary Table 1.

Deeply anesthetized mice were perfused transcardially with ice-cold PBS and brains were harvested from the animals and incubated overnight in 4% paraformaldehyde at 4 °C. Brains were then transferred to 30% sucrose at 4 °C until permeated. They were then cryopreserved with optimal cutting temperature (OCT) compound and sectioned at 20 μm using a cryostat. Briefly, sections were mounted on glass slides and treated with 10% normal donkey serum in PBS containing 0.4% Triton X-100 for 2 h. Sections were then incubated overnight at 4 °C with primary antibody [Iba1 (ionized calcium binding adaptor molecule), 1:100, goat polyclonal, Abcam, MA, United States; NeuN, 1:100, rabbit monoclonal, Cell Signaling Technologies, MA, United States; GFP, 1:100, mouse monoclonal, ThermoFisher, MA, United States]. Sections were washed 3 times with a 0.1% Triton X-100 in PBS solution and incubated with Alexa Fluor-tagged secondary antibody (Alexa Fluor 488 donkey anti-goat IgG, 1:1,000, Invitrogen, MA, United States; Alexa Fluor 594 donkey anti-goat IgG, 1:1,000, Invitrogen, MA, United States; Alexa Fluor 594 donkey anti-rabbit IgG, 1:1,000, Invitrogen, MA, United States; Alexa Fluor 488 donkey anti-mouse IgG, 1:1,000 Invitrogen, MA, United States) at room temperature in the dark for 1 h. The antibody catalogue numbers are provided in Supplementary Table 2. After washing, the sections were cover-slipped with a DAPI-containing mounting media (DAPI-Fluoromount-G; SouthernBiotech, AL, United States). Immunofluorescence was imaged using a Zeiss 780 inverted confocal laser microscope. As previously reported (Bonsack et al., 2016), we analyzed three nonconsecutive sections per animal and three fields around the hematoma. The number of immunoreactive cells per mouse was quantified using ImageJ software (NIH, United States), and averaged as positive cells per 0.1 mm² in the peri-hematomal brain region (Bonsack et al., 2016).

Brain sections (20 μm) were re-hydrated and incubated with 0.06% KMnO4 for 15 min. Sections were then washed with distilled water, incubated for 30 min in a 0.001% Fluoro-Jade B solution, dried for 30 min at 37 °C, and cover-slipped using DPX mounting media. The sections were visualized with an excitation of 488 nm using a Zeiss 780 inverted confocal laser microscope. We used six sections per animal and examined different areas around the hematoma. Employing Image J software, the number of Fluro-Jade B-positive cells was quantified as detailed previously (Bonsack et al., 2017).

The data analysis was performed using GraphPad Software, GraphPad Prism, or R software (version 4.5.1). The results are expressed as mean ± SD. The original/log-transformed data from each group were analyzed using the Shapiro–Wilk test to determine normal distribution. An unpaired t-test was used to analyze two groups that passed the normality test, and the Mann–Whitney test was used to analyze non-normal distributions. A one-way ANOVA followed by Tukey’s multiple comparisons test was used for four group comparisons. For neurobehavioral analysis, the Wilcoxon rank sum test (a nonparametric test) was performed for the two-group comparison at specific time points, and rank-based ANOVA (nonparametric test) was also used for longitudinal data analysis. A p-value <0.05 is considered statistically significant.

To determine the functional role of microglial HDAC3 after ICH, we first developed microglia-specific HDAC3 conditional knockout (cKO) mice. For that, Cx3cr1CreER mice were bred with HDAC3^f/f (HDAC3^flox/flox) to develop HDAC3^f/+: Cx3cr1, which was then bred with HDAC3^f/f to generate the conditional knockout or cKO (HDAC3^f/f: Cx3cr1). The breeding strategy enabled the generation of HDAC3^f/f (experimental control) and HDAC3^f/f: Cx3cr1 (cKO) in the same litter (Figure 1A). The two groups (experimental control and cKO) were then administered Tamoxifen as detailed in the Methods section. TAM injections were followed by a minimum 4-week waiting period before stereotactic surgery, allowing the renewal of the Cx3cr1-positive monocytes/macrophages with wild-type cells, whereas microglia maintain the gene deletion (Gao et al., 2017). Additionally, the genetic recombination assay confirmed the recombination in the brain tissue of the cKO mice (Figure 1B). Also, Cx3cr1CreER offers a unique experimental system that explicitly targets microglia with high efficiency (Goldmann et al., 2013) and expresses an enhanced yellow fluorescent protein (YFP) under the control of the Cx3cr1 promoter (Ye et al., 2019). Our studies demonstrate GFP (green fluorescent protein, a mutant of YFP that can be recognized by GFP antibody) expression predominantly in Iba1-positive cells in the mouse brain of the cKO (Figure 1C). The data suggest that Cre recombinase, the enzyme responsible for the genetic recombination, localizes predominantly to microglia. Also, GFP expression was absent in NeuN-positive neurons in the cKO mice (Figure 1C). Together, the data validate the conditional knockout model.

To evaluate the functional role of HDAC3, the male and female cKO and control mice were subjected to ICH or sham, and neurobehavioral analysis at day 1 and day 3 post-surgery was performed as detailed in the methods section. Neurobehavioral analysis was also performed on animals a day prior to the surgery. Figures 2A illustrates the experimental timeline. Irrespective of sex, no difference in neurobehavior was observed between the experimental control and the cKO group before surgery, indicating that the genetic deletion of HDAC3 in microglia did not confer any inherent neurobehavioral changes (Figures 2B,C). Notably, at day 1 and day 3 post-ICH, a significant reduction in neurological deficits was observed in male and female cKO compared to the respective control, implicating a novel role of microglial HDAC3 in regulating neurological outcomes after ICH (Figures 2B,C).

Next, we evaluated whether HDAC3 conditional deletion modulated microglial activation after ICH. Brain injury-induced microglial activation is characterized by an increase in microglial number and enhanced cellular hypertrophy in the ipsilateral brain region. Profound microglia activation, a prominent feature of neuroinflammation, occurs at day 3 post-ICH (Bonsack et al., 2016), and microglial activation was determined using immunohistochemical analysis using a marker of microglia/macrophages, Iba1. The number of Iba1-positive cells was significantly reduced in the ipsilateral perihematomal brain regions of both male and female cKO at day 3 post-ICH compared to the respective control (Figures 3A–D). Also, Iba-1-positive cells exhibited a less activated phenotype (less hypertrophic) in the male and female cKO compared to the respective control. Together, the data suggest that microglial HDAC3 conditional deletion could reduce ICH-induced neuroinflammatory response, which, in turn, could culminate in improved outcomes.

To assess neurodegeneration, the brain sections from the male and female control and cKO mice at day 3 post-ICH were subjected to Fluoro-Jade B staining. In the male cKO, the number of Fluoro-Jade B-positive cells exhibited a downward trend in comparison to the respective control; however, the reduction did not reach statistical significance (Figures 4A,B). By contrast, in the female cKO mice, there was a statistically significant reduction in the number of Fluoro-Jade B-positive cells (Figures 4C,D), suggesting that HDAC3 deletion in microglia attenuates neurodegeneration after ICH in female subjects. Together, the data suggest that there could be sex-based differences in HDAC3-mediated regulation of neurodegeneration after ICH, warranting further investigation.

To further examine the role of HDAC3 in the neuroinflammatory response after ICH, the RNA expression levels of proinflammatory mediators were assessed at day 3 post-ICH. HDAC3 conditional deletion in both male and female subjects significantly reduced the levels of proinflammatory mediators, such as Nos2, S100A9, TNF-α, and IL-6 after ICH (Figures 5A,C) compared to the respective control. Interestingly, there was no change in the expression of S100A8 and IL-1β (Figures 5A,C), suggesting HDAC3 may differentially regulate proinflammatory mediators after ICH. Furthermore, we found that the anti-inflammatory marker CD206, which is often expressed in a subset of microglia/macrophages, was upregulated in the female cKO compared to the control after ICH (Figure 5D). However, despite an upward trend, there was no significant increase in CD206 expression in male cKO mice compared to the respective controls, implicating a sex-based difference in HDAC3-mediated regulation of CD206 expression after ICH. Intriguingly, in both male and female cKO mice, another anti-inflammatory microglia or macrophage marker, Arginase-1(Arg-1), was significantly upregulated after ICH compared to the respective control (Figures 5B,D). This observation is critical as there was also a concomitant reduction in the protein (Figures 6A–D) and RNA expression of the proinflammatory microglia/macrophage marker iNOS in the cKO group compared to the respective control. Together, the data suggest that microglial HDAC3 could regulate microglial phenotypes, steering them towards a reparative phenotype after ICH.

Given the emerging role of HDAC3 in myeloid phagocytosis (Yang et al., 2024; Yang et al., 2017), which in turn could regulate hematoma size, we next questioned whether microglial HDAC3 regulates hematoma volume post-ICH. To determine this, the male and female cKO and control mice were subjected to ICH, and at day 3 post-ICH, susceptibility-weighted MRI (SW-MRI) was performed (Figures 7A–D). There was no significant change in hematoma volume in either male or female cKO group compared to the respective control. The data suggest that HDAC3 conditional deletion-mediated improvement in neurological outcome after ICH is independent of hematoma size.

To evaluate the efficacy of microglial HDAC3 deletion in long-term functional recovery, the male and female cKO and control animals were subjected to ICH, and neurobehavioral analysis was performed at days 5, 14, and 28 post-ICH. Both male and female cKO mice exhibited improved long-term outcomes compared to the respective control (Figures 8A,B).