Blocking TRPV1 in brain immune cells may reduce brain damage after bleeding by limiting cell cleanup processes

Male C57BL/6 mice (8–10 weeks old, 25–30 g) were obtained from Vital River Laboratory Animal Co. Ltd. All mice were housed under SPF conditions with a controlled 12‐h light/dark cycle, constant temperature and humidity, and free access to food and water. TRPV1^flox/flox; Cx3cr1^Cre conditional knockout (cKO) mice were generated by crossing TRPV1^flox/flox mice with Cx3cr1^Cre mice (both from Shanghai Model Organisms Center Inc.). TRPV1^flox/flox littermates served as controls. Genotyping was performed by PCR using the following primers: TRPV1 flox (F: ATGGGGTGGGTGATGCTATGTGAC; R: AGCCATGGGTGCTGCTAAAT) and Cre (F: CAAGCGCTGTTGGTGAGAGA; R: GGATTCTCCTCGACGTCACC). All procedures were approved by the Animal Care Committee of Laboratory Animal Center Fudan University (Approval ID: 202509048Z) and followed ARRIVE guidelines.

The experimental design was depicted in Figure. S1

The ICH model was established by stereotaxic injection of collagenase IV (0.0375 U in 0.5 μL saline) into the right striatum (coordinates: 0.5 mm anterior, 2.2 mm lateral to bregma, 3.5 mm depth) [26] under pentobarbital sodium anesthesia (0.3%, 0.1–0.2 mL/10 g). Sham‐operated mice underwent needle insertion without collagenase administration.

Mice were randomly assigned to different groups and received first injection 1 h after ICH induction, followed by once‐daily injections for three consecutive days. The doses of CAP (5 mg/kg) and CPZ (1 mg/kg) were selected based on previous studies [27, 28].

Short‐term neurological function was assessed using the modified Neurological Severity Score (mNSS), forelimb placement test, and corner turning test [29]. Medium‐ and long‐term functions were evaluated with the open field test (Day 14) and the Morris water maze (Day 28).

Hematoma volume and hemoglobin content were quantified 3 days post‐ICH using established methods [30]. The relative Hb levels are expressed as values normalized to the sham group (set as 1).

Brain water content was measured by the wet/dry weight method. Brain tissues were separated into ipsilateral and contralateral hemispheres and cerebellum. Each sample was weighed immediately to obtain wet weight, then dried at 100°C for 72 h and reweighed for dry weight. The brain water content was calculated as [(wet weight − dry weight)/wet weight] × 100%.

The lesion volume was evaluated via an 11.7 T small‐animal MRI apparatus (Bruker Corp.). The following setup parameters were used to acquire T2‐weighted images: repetition time, 2500 ms; image matrix, 256 × 256; echo time, 26 ms; in‐plane resolution, 0.1 × 0.1 mm²; field of view, 25.6 × 25.6 mm²; and slice thickness, 0.4 mm.

The mice were transcardially perfused with 2.5% glutaraldehyde. Subsequently, striatal samples (1 mm³) were immersion‐fixed in the same solution at 4°C for 12 h. After fixation, tissues were treated with 1% osmium tetroxide, dehydrated in ethanol, and embedded in epoxy resin. Ultrathin sections (70–90 nm) were prepared using a Leica EM UC7 ultramicrotome, contrasted with uranyl acetate and lead citrate, and visualized using a HITACHI HT7800 TEM. For quantification of mitophagosomes and damaged mitochondria, all TEM images were captured and analyzed by two independent investigators who were blinded to the experimental groups. And the average counts were used for statistical analysis.

Total protein was extracted from tissues or cells using a commercial kit (SD‐001/SN‐002, Invent Biotechnologies). Equal amounts of protein were separated by SDS‐PAGE and transferred to nitrocellulose membranes (Millipore). After blocking with 5% nonfat milk for 1 h, the membranes were incubated overnight at 4°C with primary antibodies (see Table), followed by incubation with species‐matched secondary antibodies (Cell Signaling Technology, 1:5000) for 2 h at room temperature. Protein bands were visualized using an enhanced chemiluminescence (ECL) substrate. S1

A terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining kit (Roche, 11684795910) was used to identify in situ cell apoptosis in the peri‐hematoma area.

After euthanasia, mice were perfused transcardially with PBS followed by 4% paraformaldehyde (PFA). The harvested brains were postfixed in 4% PFA for 24 h, cryoprotected in 30% sucrose, and sectioned coronally (30 μm) using a Leica CM1950 microtome. Sections were permeabilized with 1% Triton X‐100 (20 min) and blocked with 5% BSA (1 h), after which they were incubated with primary antibodies (Table) overnight at 4°C. Following PBS washes, sections were incubated with corresponding secondary antibodies (Table) for 2 h at room temperature, and finally mounted with an antifade medium (Beyotime Biotechnology, P0131) containing DAPI. S2 S2

Three days after ICH, degenerative neurons were identified using a Fluoro‐Jade C (FJC) ready‐to‐dilute staining kit (Millipore). Briefly, sections were dried at 37°C for 2–3 days, followed by sequential incubation in solutions A, B, and C. They were then heated at 50°C for 10 min and finally cleared in xylene for 2 min.

After PBS rinses, brain sections were incubated in Nissl's stain reagent (Servicebio) for 3 min, briefly rinsed with distilled water, air‐dried at 65°C, and coverslipped with neutral resin.

BV2 and HT22 cell lines (Pricella Biotechnology) were cultured in high‐glucose DMEM (Gibco) containing 10% FBS (Gibco) and 1% penicillin/streptomycin at 37°C in a 5% CO₂ incubator. All experiments utilized cells between passages 10–30. For the in vitro ICH model, cells were exposed to 10 μM hemoglobin (Hb; Sigma) for 24 h. When applicable, the AMPK activator A769662 (30 μM) or inhibitor BAY‐3827 (1 μM) was added to the culture medium.

The mice were anesthetized and intracardially perfused with cold PBS. Then brains were removed, placed in ice‐cold HBSS, and dissected to remove the cerebellum, meninges, blood vessels, and choroid plexus. The tissue was then minced, digested with 0.25% trypsin at 37°C for 5 min. After filtering through a 70‐μm strainer and centrifugation, the pellet was resuspended in 37% isotonic Percoll and separated on a discontinuous Percoll gradient (70%/37%/30%) by centrifugation at 300 g for 30 min at 18°C (low acceleration/brake). Microglia were collected from the 37%/70% interface, washed with PBS, and cultured in complete DMEM on PDL‐coated plates.

siRNAs targeting human TRPV1 (siTRPV1) and scrambled siRNA were synthesized by GenePharma Biotech (Shanghai, China). TRPV1 siRNAs (Table) were transfected into BV2 cells via Lipofectamine 3000 transfection reagent (Invitrogen). S3

HT22 cells were stained with JC‐1 working solution (Beyotime Biotechnology, C2003S) at 37°C for 20 min to assess mitochondrial membrane potential, following the kit protocol.

BV2 cells were washed twice with HBSS and then loaded with 2 μM Rhod‐2/AM (Ca²⁺ probe; 40776ES72, Yeasen) for 30 min. Subsequently, the cells were washed three times with HBSS and further incubated in HBSS for 30 min at 37°C.

The mitochondria and cytoplasm components in BV2 cells were extracted using the mitochondrial isolation kit (Beyotime, C3601), and then the subsequent experiments were conducted. The extraction method should follow the instructions provided by the manufacturer.

Blood was collected from the mice and red blood cells were separated by gradient centrifugation. Follow the manufacturer's instructions to incubate red blood cells with the PKH26 reagent (Beyotime, C2071S). Then, add the red blood cells to the BV2 cell culture medium and co‐culture them for 12 h.

Statistical analyses were performed using GraphPad Prism 9. All data were presented as mean ± SD. Comparisons between two groups were made using Student's t‐test, while one‐way or two‐way ANOVA followed by Tukey's post hoc test was applied for multi‐group comparisons. Kruskal–Wallis test followed by Dunn's post hoc test was used to analyze the mNSS score. A p‐value of less than 0.05 was considered statistically significant.

While TRPV1 is implicated in brain ischemia–reperfusion injury [31], its role in intracerebral hemorrhage (ICH) requires further investigation. We assessed TRPV1 protein levels around the hemorrhage site over time. Western blot analysis showed that TRPV1 expression was significantly upregulated on Day 3 post‐ICH (p < 0.05, Figure 1b) and remained elevated for 1 week. Immunofluorescence co‐staining revealed that TRPV1 was predominantly expressed in IBA1‐positive microglia, and to a lesser extent in NeuN‐positive neurons, with minimal expression in GFAP‐positive astrocytes or OLIG2‐positive oligodendrocytes (Figure 1c,d). Given the pivotal role of microglia in hematoma clearance and their accumulation around hemorrhagic areas, these findings suggest a potential role for TRPV1 in ICH pathogenesis, particularly within microglia.

The effects of CAP on target cells depend on dose [32] and the degree of TRPV1 activation relative to disease severity. To determine optimal dosing, ICH mice received intraperitoneal injections of CAP (0.1, 0.5, or 1 mg/kg) or CPZ (0.5, 1.0, or 2 mg/kg). All CAP doses significantly improved the mNSS 1 week after ICH (p < 0.05, Figure S2), with 0.5 mg/kg showing efficacy comparable to 1 mg/kg; higher doses were avoided due to potential toxicity. While 0.5 mg/kg CPZ was ineffective (p > 0.05, Figure S2), both 1.0 and 2 mg/kg CPZ were beneficial (p < 0.05, Figure S2), with no difference between them. Consequently, 0.5 mg/kg CAP and 1 mg/kg CPZ were selected for subsequent studies. Although systemic TRPV1 modulation can affect the broader immune response [33], we chose intraperitoneal injection for its clinical translational feasibility.

Treatment with CPZ significantly reduced hematoma volume on Day 3 post‐ICH (p < 0.05, Figure 2a,b) and decreased ipsilateral brain water content (p < 0.05, Figure 2d). In contrast, CAP did not affect hematoma volume (p > 0.05, Figure 2a,b) but worsened cerebral edema (p < 0.05, Figure 2d). Hemoglobin levels on Day 3 were consistent with the hematoma volume data (Figure 2c). No significant differences were observed in the contralateral hemisphere or cerebellum among groups (p > 0.05, Figure 2d).

Widespread blood–brain barrier (BBB) disruption due to peri‐hematomal vascular inflammation is a known consequence of ICH [34]. Consistent with this, CPZ treatment significantly upregulated the tight junction proteins ZO‐1 (p < 0.05, Figure S3a,b) and claudin‐5 (p < 0.05, Figure S3c,d) compared to the vehicle group at 3 days post‐ICH. This suggests that CPZ may ameliorate cerebral edema by restoring BBB integrity.

We also monitored hematoma volume in vivo using an 11.7 T high‐resolution MRI system. Initial hematoma volumes were comparable across all groups (p > 0.05, Figure 2e,f). Subsequent imaging revealed that CPZ treatment significantly reduced hematoma volume (p < 0.05, Figure 2e,f), whereas CAP had no significant effect compared to the vehicle group (p > 0.05, Figure 2e,f).

We next asked whether TRPV1 blockade confers neuroprotection after ICH. TUNEL and FJC staining revealed a marked increase in apoptotic and degenerating neurons after ICH (p < 0.05, Figure 2g,h). CPZ administration reduced the number of TUNEL‐positive neurons by approximately 50% (p < 0.05, Figure 2g,h), a finding consistent with reports in TRPV1 knockout mice [11]. Conversely, CAP treatment aggravated neuronal apoptosis (p < 0.05, Figure 2g,h). Consistent with the TUNEL data, CPZ also significantly decreased the number of degenerating FJC‐positive neurons (p < 0.05, Figure 2i,j). These data demonstrate that TRPV1 inhibition effectively mitigates ICH‐induced neuronal damage.

To evaluate long‐term neuronal damage, we performed Nissl staining on hippocampal tissues 4 weeks after ICH. CPZ‐treated mice showed significantly greater neuronal survival in the DG, CA1, and CA3 regions relative to the vehicle group (p < 0.05, Figure 2k,l). Conversely, CAP treatment resulted in significant neuronal loss restricted to the DG region (p < 0.05, Figure 2k,l), whereas other subfields were unaffected (p > 0.05, Figure 2k,l). These findings demonstrate that TRPV1 inhibition effectively mitigates neuronal loss following ICH.

To determine whether the neuroprotective and hematoma‐resolving effects of CPZ translate to improved functional outcomes, we subjected mice to a series of behavioral tests. The CAP‐treated group was excluded from these and subsequent experiments due to its consistent detrimental effects on hematoma resorption, neuronal survival (Figure 2), and early neurological function (Figure S2).

CPZ treatment significantly improved short‐term neurological function after ICH across multiple behavioral assays. Compared to the vehicle group, CPZ administration reduced mNSS scores on Days 5 and 7 (p < 0.05, Figure 3a), increased the success rate in the foot‐fault test on Day 7 (p < 0.05, Figure 3b), and improved cornering test scores on Days 5 and 7 (p < 0.05, Figure 3c). Notably, cornering test performance in the CPZ group recovered to a level comparable with the sham group (p > 0.05, Figure 3c).

To evaluate long‐term neurological function, the open field test was performed 14 days post‐ICH. CPZ‐treated mice exhibited significant increases in total travel distance and central zone distance compared to the vehicle group (p < 0.05, Figure 3d–f). Furthermore, in the Morris water maze test beginning on Day 28, CPZ treatment improved spatial learning and memory. During acquisition, CPZ‐treated mice located the hidden platform faster on Days 4 and 5 (p < 0.05, Figure 3g,h), despite persistent deficits in the ICH model relative to sham controls (p < 0.05, Figure 3h). In the probe trial, they also spent more time in the target quadrant and had a shorter platform location distance than vehicle‐treated mice (p < 0.05, Figure 3i,j), indicating that TRPV1 blockade ameliorates ICH‐induced cognitive deficits.

Since microglia are key responders in ICH [35] and predominantly upregulate TRPV1, we investigated whether TRPV1 blockade protects neurons by modulating their function. CPZ treatment did not change the total number of IBA1⁺ microglia at Day 3 post‐ICH (p > 0.05, Figure 4a,b) but significantly promoted their phenotypic shift: it increased the proportion of IBA1⁺Arg1⁺ regulatory microglia (p < 0.05, Figure 4a,c) while reducing IBA1⁺iNOS⁺ pro‐inflammatory microglia (p < 0.05, Figure 4d,e). These results indicate that CPZ reprograms microglia from a pro‐inflammatory to a regulatory state.

To establish the microglial dependency of CPZ's effects, we depleted microglia using the CSF1R inhibitor PLX5622 (Figure 4f), which effectively reduced the microglial population (p < 0.05, Figure 4g,h). Depletion exacerbated ICH‐induced neuronal apoptosis (p < 0.05, Figure 4i,j) and abolished the neuroprotective effect of CPZ, as apoptosis levels showed no difference between CPZ‐ and vehicle‐treated groups (p > 0.05, Figure 4i,j). These results demonstrate that microglia are essential for CPZ's anti‐apoptotic effect, indicating microglia‐dependent mediation.

To definitively establish microglial TRPV1 as the key mediator in ICH, we generated a microglia‐specific TRPV1 cKO model. Through genotyping and immunohistochemical staining, we validated effective TRPV1 knockout in microglia (Figure 4k). TRPV1 cKO mice exhibited reduced neuronal apoptosis (p < 0.05, Figure 4l,m) and improved neurological function across multiple behavioral tests on Day 7 post‐ICH (p < 0.05, Figure 4n–p). This neuroprotective phenotype mirrored the effects of CPZ treatment, confirming that TRPV1 exacerbates ICH injury primarily through its actions in microglia.

To further confirm that CPZ acts specifically through microglial TRPV1, we treated microglia‐specific TRPV1 cKO mice with CPZ after ICH. While a subset of neurons retained TRPV1 expression (Figure S4a), CPZ failed to confer additional protection beyond microglial TRPV1 deletion alone: neuronal apoptosis (Figure S4b,c) and mitophagy markers (LC3B‐VDAC and Parkin‐VDAC colocalization, Figure S4d–g) were comparable between vehicle‐ and CPZ‐treated cKO groups (p > 0.05). These results provide direct genetic evidence that microglial TRPV1 is indispensable for CPZ's protective effects.

We also performed in vitro validation using a transwell system to determine CPZ's microglia‐specificity (Figure S5a). Conditioned medium from Hb‐stimulated BV2 microglia reduced mitochondrial membrane potential in HT22 neurons (JC‐1 staining—an early marker of apoptosis, Figure S5b,c, p < 0.05) and increased neuronal apoptosis (cleaved caspase‐3, Figure S5d,e, p < 0.05). These effects were significantly attenuated when BV2 cells were pretreated with CPZ, confirming indirect neuroprotection through microglial modulation. CPZ pretreatment also reduced IL‐1β and TNF‐α levels in BV2‐conditioned medium (Figure S5f,g, p < 0.05).

Given that mitochondrial damage can lead to the release of mitochondrial DNA (mtDNA), which acts as a damage‐associated molecular pattern (DAMP) to amplify inflammation [36, 37, 38], we next examined cytosolic mtDNA levels. As shown in Figure S5h, Hb stimulation triggered mtDNA release into the cytoplasm of BV2 microglia (Picogreen in cytoplasm), suggesting that mtDNA leakage may contribute to the pro‐inflammatory microglial phenotype. However, no significant changes in the number of mitophagosomes (p > 0.05, Figure S5i,j) or in the colocalization of Parkin with MitoTracker (p > 0.05, Figure S5k,l) were observed in HT22 neurons. These findings indicate that CPZ protects neurons indirectly by suppressing microglial mtDNA release and pro‐inflammatory cytokines rather than directly modulating neuronal autophagy, consistent with our in vivo data showing CPZ shifts microglia toward a regulatory phenotype (Figure 4a–e).

We next investigated the downstream mechanisms of microglial TRPV1 activation. TRPV1 is closely associated with autophagy [16, 39], an intracellular phagocytic activity. Although autophagy is significantly elevated after ICH, its specific role and underlying mechanisms remain unclear [40]. Western blot analysis showed that ICH significantly upregulated the autophagy‐related proteins Parkin, PINK1, and LC3B‐II (p < 0.05, Figure 5a,b), indicating enhanced autophagy. CPZ treatment reduced the levels of Parkin and LC3B‐II (p < 0.05, Figure 5a,b), suggesting suppressed autophagic activity. Furthermore, CPZ partially reversed the ICH‐induced decrease in the mitochondrial marker COX IV (p < 0.05, Figure 5a,b), suggesting that TRPV1 activation contributes to mitochondrial loss via excessive autophagy.

Given the critical role of mitochondrial function in microglial phagocytosis during ICH [14], we specifically investigated mitophagy. Immunofluorescence analysis revealed that ICH significantly enhanced the colocalization of LC3B with VDAC (p < 0.05, Figure 5c,d), indicating enhanced mitophagic flux. This increase was significantly attenuated by CPZ treatment (p < 0.05, Figure 5c,d). Similarly, CPZ also reduced the ICH‐induced colocalization (Manders coefficient) between the mitophagy receptor Parkin and VDAC (p < 0.05, Figure 5e,f). These data demonstrate that TRPV1 blockade suppresses ICH‐induced mitophagy activation. Direct visualization by TEM confirmed these findings: vehicle‐treated group exhibited a substantial increase in mitophagosomes (p < 0.05, Figure 5g–i, red arrows) and damaged mitochondria (yellow arrows), which was significantly reduced by CPZ treatment.

We next validated the role of TRPV1 in ICH‐associated autophagy using an in vitro model. Hb treatment significantly upregulated LC3B, PINK1, and Parkin expression (p < 0.05, Figure 6a,b), recapitulating the in vivo findings, and these increases were suppressed by CPZ. Furthermore, cellular fractionation revealed that Hb‐induced cytochrome C (Cyto C) release from mitochondria into the cytoplasm was also mitigated by CPZ (p < 0.05, Figure 6c,d). These findings confirm that CPZ stabilizes mitochondrial structure by inhibiting TRPV1‐driven autophagy in microglia.

TRPV channel activation is known to trigger AMPK signaling [17], a key regulator of autophagy that responds to calcium flux [41]. Accordingly, Hb stimulation increased AMPK phosphorylation, an effect attenuated by CPZ (p < 0.05, Figure 6a,b). Pharmacological inhibition of AMPK (BAY‐3827) downregulated PINK1 and Parkin, similar to CPZ, while AMPK activation (A769662) partially reversed CPZ's effects (p < 0.05, Figure 6e,f). Since TRPV1 activation elevates intracellular Ca²⁺, we measured calcium levels and confirmed that CPZ reduced the Hb‐induced Ca²⁺ influx (p < 0.05; Figure 6g,h). Together, these results demonstrate that TRPV1 activation promotes autophagy primarily via the Ca²⁺‐AMPK‐PINK1 pathway.

We also analyzed the cellular localization of LC3B and Parkin. Hb stimulation significantly increased the colocalization of LC3B with VDAC, which was attenuated by CPZ treatment (p < 0.05, Figure 6i,j). Similarly, the colocalization index of Parkin with VDAC demonstrated the same pattern (p < 0.05, Figure 6k,l), confirming that TRPV1 blockade suppresses Hb‐induced mitophagy.

To determine whether CPZ selectively inhibits mitophagy or broadly suppresses autophagy, we performed subcellular fractionation to analyze autophagic activity in mitochondrial versus cytosolic compartments. Following Hb stimulation, CPZ significantly reduced PINK1 and Parkin levels in both mitochondrial and cytosolic fractions (Figure S6a–d, p < 0.05), indicating that CPZ suppresses global autophagy flux rather than selectively targeting mitophagy.

Despite this autophagy suppression, we next asked whether phagocytic function—critical for hematoma clearance—was compromised. Using PKH26‐labeled red blood cell phagocytosis assays, we found that CPZ‐treated BV2 cells exhibited significantly enhanced uptake compared to vehicle‐treated cells, as evidenced by the increased percentage of phagocytic cells and higher number of engulfed RBCs per cell (Figure S6e–g, p < 0.05). These results demonstrate that while CPZ suppresses global autophagy, it paradoxically preserves—and even enhances—phagocytic capacity, likely by preventing the mitochondrial dysfunction caused by excessive autophagy (Figures 5 and 6).

We validated the concordant responses of primary microglia and BV2 cells in the ICH model. Primary microglia, isolated by gradient centrifugation [42] and identified by morphology and IBA1 staining (Figure 6m,n,p), exhibited a CPZ‐induced protective phenotype. This was evidenced by an increase in Arg1⁺ cells and a decrease in iNOS⁺ cells (p < 0.05, Figure 6n–q), consistent with in vivo data (Figure 4a,d). CPZ also inhibited autophagosome formation (p < 0.05, Figure 6r–s), further corroborating our findings in a primary system.

To further corroborate our findings, we knocked down TRPV1 expression in BV2 cells via siRNAs (p < 0.05, Figure S6h,i). TRPV1 knockdown effectively reduced the number of Hb‐induced mitophagosomes (p < 0.05, Figure S6j,k), reinforcing that TRPV1 blockade suppresses aberrant autophagy following ICH.

The data that support the findings of this study are available from the corresponding author upon reasonable request.