BTLA-mediated regulation of neuroimmune responses enhances recovery after intracerebral hemorrhage

Journal: 2026;23():e69  |  DOI: 10.1186/s12974-026-03705-6  |  PMID: 41578304  |  PMCID:

Authors: Yuxing Wu; Yanping Zhang; Peiji Fu; Maosen Huang; Shuai Chen; Yi Li; Tao Wu; Yousef Rastegar-Kashkooli; Nannan Cheng; Taiqi Zhao; Yunlong Wang; Xiaojun Wang; Jian Wang; Junmin Wang; Chao Jiang

Abstract

Modulating harmful neuroimmune responses is a promising therapeutic approach for hemorrhagic stroke, a condition that still lacks effective treatment. The immune checkpoint B and T Lymphocyte Attenuator (BTLA) helps suppress immune activation; however, its role in intracerebral hemorrhage (ICH) remains unclear. This study explores whether a BTLA-activating antibody can reduce neuroinflammation, mitigate brain injury, improve recovery after ICH, and elucidate the underlying mechanisms.

An ICH model was generated in male C57BL/6 mice by stereotactic injection of collagenase VII-S into the left striatum. The mice received intraperitoneal administration of an agonistic anti-BTLA antibody to serve as a BTLA agonist. Therapeutic effects were evaluated using a multimodal approach that included flow cytometry, Western blotting, immunofluorescence staining, histological examination, and behavioral tests. Additionally, microglial depletion was performed by feeding the CSF1R inhibitor PLX5622.

Our findings demonstrate that a single dose of an agonistic anti-BTLA antibody, administered 30 min post-ICH, significantly reduced the infiltration of CD45^high, CD3⁺, and CD3⁺CD4⁺ immunocytes, as well as the activation of CD3⁺CD4⁺ and CD3⁺CD8⁺ immunocytes, in the hemorrhagic brain following acute ICH. Simultaneously, it reduced neutrophil infiltration into the hemorrhagic brain and suppressed the activation of peripheral CD3⁺CD8⁺ immune cells. It also alleviated molecular and cellular neuroinflammation in the hemorrhagic brain during the early phase after ICH. These effects in the brain of adult male mice ultimately reduced both acute-phase brain injury volume and long-term residual lesions, while facilitating neurological recovery. However, microglial depletion abolished the anti-inflammatory effects of the agonistic anti-BTLA antibody, indicating that its action is contingent on microglia-mediated immunomodulation.

The agonistic anti-BTLA antibody significantly attenuates neuroinflammation and reduces brain injury following ICH, accompanied by enhanced neurological recovery. This protective effect appears to be mediated through microglia-dependent mechanisms. Our findings highlight BTLA may be a novel and promising immunomodulatory target for the treatment of ICH.

The online version contains supplementary material available at 10.1186/s12974-026-03705-6.

Introduction

Intracerebral hemorrhage (ICH) represents one of the most devastating forms of stroke, accounting for nearly half of the global stroke burden [^[1], ^[2]]. It is associated with high mortality and disability, often resulting in long-term motor and cognitive deficits. Despite its clinical severity, effective therapeutic interventions for ICH remain limited [^[3]]. The pathophysiology of ICH includes both primary and secondary injuries, primary injury results from the mass effect of the hematoma, while the breakdown products of blood trigger secondary injury. This secondary process involves complex mechanisms such as neuroinflammation, oxidative stress, and cytotoxicity [^[4]]. Among these mechanisms, dysregulated neuroimmune and inflammatory responses play a central role in exacerbating neuronal damage and worsening neurological outcomes [^[5]]. Therefore, controlling neuroinflammation after ICH has become a necessary therapeutic approach to reduce secondary brain injury [^[6]].

Following the ICH, an immune-inflammatory cascade is rapidly triggered [^[7]]. This response involves microglial activation, the infiltration of peripheral immune cells such as neutrophils and T lymphocytes, and the excessive release of proinflammatory cytokines, including interleukin-1β (IL-1β), tumor necrosis factor-α (TNF-α), and interferon-γ (IFN-γ) [^[3], ^[8]]. Studies in animal models show that microglial activation peaks between 3 and 7 days after ICH, alongside pronounced astrogliosis and neutrophil recruitment [^[9], ^[10]]. T cell-mediated adaptive immune responses are crucial in ICH, as overactivation of CD4⁺ and CD8⁺ T cells disrupts immune balance and exacerbates secondary injury [^[11], ^[12]]. Moreover, the early T cell activation marker CD69 is markedly upregulated after ICH [^[13]], while inflammatory mediators like Regulated upon Activation, Normal T cell Expressed and Secreted (RANTES) and High Mobility Group Box 1 (HMGB1) remain elevated, further driving the neuroinflammatory response [^[14]–^[16]]. Together, these findings indicate that targeting T cell overactivation and the associated cytokine release may offer a promising therapeutic approach to improve outcomes after ICH.

B/T-lymphocyte attenuator (BTLA), an immune checkpoint molecule structurally homologous to cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) and programmed Cell Death Protein 1 (PD-1), primarily functions as a negative regulator of T-cell activation [^[16], ^[17]]. Accumulating evidence demonstrates BTLA’s critical immunomodulatory role in both systemic and localized inflammatory diseases. For instance, BTLA-knockout mice exhibit exacerbated inflammatory responses in models of endotoxic shock and experimental glomerulonephritis [^[18]]. Agonistic antibodies (e.g., clone 6A6) work by activating specific biological pathways, effectively mimicking the action of natural signaling molecules [^[19]–^[21]]. These findings collectively suggest that targeted activation of BTLA signaling may represent a novel therapeutic strategy for inflammatory disorders [^[17]].

Like infiltrating T lymphocytes, microglia, the resident innate immune cells in the brain, may also express BTLA [^[22]]. Additionally, crosstalk between microglia and T lymphocytes occurs in the hemorrhagic brain [^[23], ^[24]]. In this study, we investigated the inhibitory effect of an agonistic anti-BTLA antibody, acting as a BTLA agonist, on T lymphocyte infiltration and neuroinflammation in the hemorrhagic brain. We confirmed BTLA expression in microglia within the hemorrhagic brain. We assessed the therapeutic potential of an agonistic anti-BTLA antibody for ICH, focusing on the role of microglia and their interaction with infiltrating T lymphocytes in the injured brain. This is the first investigation into the impact of BTLA on T-cell infiltration, neuroinflammation, and functional outcomes following ICH.

Materials and methods

Animals

A total of 164 young male C57BL/6 mice aged 10–12 weeks (body weight 22–26 g) were obtained from the Experimental Animal Center of Zhengzhou University. All mice were housed in specific pathogen-free (SPF) facilities at the Fifth Affiliated Hospital of Zhengzhou University under single-cage conditions, with ad libitum access to food and water. The animal room environment was maintained at 21–25 °C, with a relative humidity of 40%-60% and a 12-hour light/12-hour dark cycle. The study protocol was approved by the Animal Ethics Committee of the First Affiliated Hospital of Henan University CM (YFYDW2023042/B), and all experimental procedures were conducted in strict accordance with the ARRIVE guidelines [^[25]].

Mouse model of intracerebral hemorrhage

The ICH model was established using stereotactic collagenase injection as previously described [^[8], ^[26]]. Anesthesia was induced with 3.0% isoflurane and maintained at 1.0% via a nasal cone using a gas mixture of 80% nitrogen and 20% oxygen. Core body temperature was continuously monitored and maintained at 36.5–37.5 °C using a thermoregulation system. The mice were positioned prone in a digital stereotaxic frame (RWD Life Science, China), and the scalp was disinfected with iodophor before a 1-cm midline incision was made to expose the bregma and lambda sutures. According to the mouse brain atlas (Paxinos and Franklin, 2019), the left caudate nucleus was targeted (coordinates: 0.6 mm anterior to bregma, 2.0 mm lateral, 3.2 mm depth). After drilling a burr hole (RWD, Model 78001, 0.5-mm drill bit), 0.5 µL of collagenase VII-S solution (Sigma-Aldrich, Cat# C2399; 0.075 U in sterile saline) was infused at 0.1 µL/min using a 1-µL Hamilton microsyringe driven by a pump (RWD, Model 788130) mounted on the stereotaxic instrument.

Sham mice received an identical volume of sterile saline injection at the targeted location. The needle was retained for 10 min after the injection before being slowly withdrawn (at approximately a rate of 1 mm/min) [^[27]]. The burr hole was then sealed using bone wax (SANYOU, China), and the skin incision was closed with cyanoacrylate glue (Loctite, Germany). Postoperatively, mice were placed on a 37 °C heating pad until they had fully recovered, then returned to their home cages with ad libitum access to food and water. Any animals that died prematurely (e.g., due to anesthesia overdose or intracranial infection) were excluded from analysis. Only mice that survived and met the predefined criteria for successful model establishment were included in the final datasets [^[8]].

Treatment protocols and experimental groups

To investigate the role of BTLA signaling in post-ICH immune cell infiltration, Th1-type immune responses, brain damage, and long-term neurological function, wild-type C57BL/6 mice were randomly divided into four groups using computer-generated randomization [^[28]]: (1) Sham ICH + isotype control, (2) Sham ICH + anti-BTLA agonist antibody, (3) ICH + isotype control, (4) ICH + anti-BTLA agonistic antibody. A single intraperitoneal dose of either the isotype control antibody (BE0091, BioXcell) or the anti-BTLA agonistic antibody (BE0132, BioXcell) was administered at a dose of 15.4–18.2 mg/kg (400 µg in 70.5 µl per mouse) 30 min post-surgery [^[20], ^[21], ^[29], ^[30]]. As reported earlier, systemic administration of a 400 µg anti-BTLA agonistic antibody per mouse induces immunological responses distinct from those observed in BTLA-deficient mice under inflammatory conditions in vivo [^[20], ^[31]].

To clarify the role of microglia in BTLA-mediated neuroprotection, two additional groups were established: (1) ICH + normal diet + anti-BTLA agonistic antibody, (2) ICH + special diet containing CSF1R inhibitor PLX5622 + anti-BTLA agonistic antibody. Mice in the normal diet group had free access to the AIN-76 A standard chow without the CSF1R inhibitor PLX5622 (MCE, HY-AF0001). Mice in the second group received a special diet containing the CSF1R inhibitor PLX5622 (PLX5622 formulated in AIN-76 A standard chow at 1200 ppm, MCE, HY-114153 C) for 14 days before ICH induction [^[32], ^[33]]. As previously shown, the anti-BTLA agonist antibody (15.4 to 18.2 mg/kg, 400 µg in 70.5 µl per mouse, BioXcell, BE0132) was also injected intraperitoneally 30 min after surgery [^[20], ^[21], ^[29]]. Each mouse in the special diet group consumed approximately 4.8 mg of PLX5622 daily, based on an average intake of 4 g per day. Evidence shows that this dose of PLX5622, given for 14 days, can eliminate 90% of microglia in mouse brains and maintain this effect for 7 days [^[34]–^[36]].

Leukocyte isolation

Mice were anesthetized with 3.0% isoflurane inhalation at 36 h, 3 days, and 7 days after ICH induction. Approximately 0.5–0.8 mL of peripheral blood was then collected from each mouse via retro-orbital bleeding and placed into EP tubes pre-coated with 50 U/mL heparin sodium at each time point (n = 5–6 mice per group). Red blood cells were lysed using RBC lysis buffer (Solarbio, R1010) and incubated on ice for 15 min. The reaction was then stopped by adding 10 mL ice-cold PBS (1×). Samples were centrifuged at 500 g for 10 min at 4 °C, and the resulting pellet was resuspended in PBS and passed through a 70 μm cell strainer (Biosharp, BCS-70-CS) to obtain leukocytes for subsequent analysis [^[37], ^[38]].

Brain tissue single-cell suspension preparation

To minimize animal usage, mice that had already undergone blood collection were immediately perfused with ice-cold PBS to clear circulating blood cells [^[39]]. Brains were then removed and kept in ice-cold 1× PBS at 36 h, 3 days, and 7 days after ICH (n = 5–6 mice per group). Single-cell suspensions were prepared using the Adult Mouse Brain Dissociation Kit (K3020502, RWD Life Science, China) according to established protocols [^[37], ^[38], ^[40]]. Each hemorrhagic hemisphere was minced and placed into a tissue processing tube (SCT-100, RWD Life Science) containing Enzyme A/B mixture. Mechanical/enzymatic dissociation was carried out using a Single-Cell Suspension Processor (DSC-400, RWD Life Science) with the M_ABrain_Heater_1 program. The resulting suspensions were filtered through 70 μm strainers (BCS-70-CS, Biosharp) to remove debris and residual erythrocytes. After washing and dispersing the cells, they were centrifuged at 400 g for 10 min (4 °C) and resuspended in PBS (1×) for flow cytometry.

Flow cytometry analysis

At 36 h, 3 days, and 7 days post-ICH, flow cytometry was conducted to assess T lymphocyte infiltration and functional changes in the hemorrhagic brains and peripheral blood of each mouse (n = 5–6 mice per group). Single-cell suspensions from injured brain tissues and peripheral blood were incubated for 30 min at 4 °C in the dark with the following fluorochrome-conjugated antibodies: APC-Cy7 anti-mouse CD45 (103116, BioLegend), APC anti-mouse CD3 (100236, BioLegend), FITC anti-mouse CD4 (100510, BioLegend), Pacific Blue anti-mouse CD8a (100725, BioLegend), and PE anti-mouse CD69 (104507, BioLegend). Dead cells were excluded using the Zombie Aqua™ Fixable Viability Kit (423101, BioLegend). Samples were analyzed on a BD FACSCelesta flow cytometer, with compensation and gating set using fluorescence-minus-one (FMO) and isotype controls [^[8], ^[37]].

Tissue processing and short- and long-term brain injury evaluation

To examine neuropathological changes at various stages after ICH, mice were anesthetized with 3.0% isoflurane and transcardially perfused with saline, followed by 4.0% paraformaldehyde at day 3 (acute phase) and day 28 (chronic phase) post-ICH [^[39]]. Harvested brains were post-fixed in 4.0% PFA for 24 h and then cryoprotected in 20% and 30% sucrose solutions made in PBS buffer. Tissues embedded in OCT were coronally sectioned using a cryostat (CM1950, Leica, German) at a thickness of 50 μm with 360 μm intervals for LFB/CV staining, and at 30 μm for serial sections, which were stored at -20 °C.

Myelin and neuronal integrity were assessed using Luxol Fast Blue (LFB; Sigma, S3382) and Cresyl Violet (CV; Sigma, C5042) dual staining [^[37]]. Areas lacking LFB/CV staining were used to identify damaged tissue. Following standard protocol, 50-µm-thick brain sections were mounted on glass slides, air-dried at 37 °C for 12 h, and incubated overnight at room temperature in a pre-warmed (60 °C) 0.1% LFB working solution. After two ultrapure water rinses, the slides were differentiated in 0.05% lithium carbonate for 5 min and rinsed again. The sections were then counterstained with 1% CV solution (pH 4.0) for 30 min at room temperature, rinsed, dehydrated through a graded ethanol series (70%, 80%, 96%, and 100%), cleared in xylene, and cover-slipped in Entellan.

Lesion volumes (mm³) were calculated on day 3 (n = 10 mice per group) and day 28 (n = 12 mice per group) after ICH by multiplying the lesion area by the section interval. Brain swelling was measured on day 3, and brain atrophy was assessed on day 28 using SigmaScan Pro software. The percentage of brain swelling was calculated as: [(ipsilateral volume - contralateral volume) / contralateral volume] × 100%, reflecting the hematoma’s mass effect and the extent of associated edema. Brain atrophy, resulting from cellular damage, was expressed as a percentage of the contralateral volume using the formula: [(contralateral volume - ipsilateral volume) / contralateral volume] × 100% [^[41]].

Immunofluorescence staining

To assess cellular neuroinflammation 3 days after ICH, 30-µm coronal brain sections (stored at -20 °C) from six mice per group were immunostained for ionized calcium-binding adapter molecule 1 (Iba-1), glial fibrillary acidic protein (GFAP), and myeloperoxidase (MPO) [^[8]]. To verify BTLA expression in microglia and evaluate microglial depletion efficiency, additional sections from three mice per group were co-stained for BTLA and Iba-1 or stained separately for Iba-1 alone at the same time point after ICH.

Three evenly spaced perihematomal sections per mouse were washed with PBS and blocked with 3% BSA (RT, 60 min). Sections were then incubated overnight at 4 °C with primary antibodies: rabbit anti-GFAP (astrocytes, 1:200, Proteintech 16825-1-AP), rabbit anti-Iba-1 (microglia/macrophages, 1:1000, Dako 019-19741), anti-MPO (neutrophils, 1:150, Abcam ab9535), goat anti-Iba-1 (microglia/macrophages, 1:1000, Dako 011-27991), and rabbit anti-BTLA (1:100, Abcam ab212089), with negative controls omitting the primary antibodies. After washing with PBST, sections were incubated (RT, 70 min) with fluorescent secondary antibodies: Alexa Fluor 488-goat anti-rabbit IgG (1:1000, Invitrogen A-11034), Alexa Fluor 594-goat anti-rabbit IgG (1:1000, Invitrogen R37117), and Alexa Fluor 488-donkey anti-goat IgG (1:1000, Invitrogen A-11055). Following additional PBST washes, sections designated for assessing cellular neuroinflammation and microglial depletion efficiency were mounted on adhesive slides (Citotest, 188105 W) using an antifade mounting medium without DAPI (Solarbio S2100). For BTLA/Iba-1 co-staining, sections were similarly mounted on adhesive slides (Citotest, 188105 W) with an antifade medium containing DAPI (Servicebio, G1407). After sealing, four perihematomal fields per section were then imaged using a Nikon Ni-U microscope (20×). Positive cell density (cells/mm²) was quantified via ImageJ 2.0 (NIH).

Microglial/macrophage activation status was evaluated using Iba-1 immunoreactivity and cell-morphology analysis [^[42]]. Activated microglia/macrophages displayed intense Iba-1 immunoreactivity and exhibited spherical, amoeboid, or rod-shaped morphology, measuring over 7.5 μm in at least one dimension with short, thick processes. In contrast, resting microglia/macrophages showed smaller cell bodies (under 7.5 μm in diameter), faint Iba-1 staining, and elongated, fine processes. Additionally, reactive astrocytes were identified by increased GFAP immunoreactivity, characterized by thickened, elongated processes in perihematomal regions [^[41]]. Using ImageJ, we manually set appropriate thresholds to analyze microglial/macrophage activation in hemorrhagic brain tissue and evaluate microglial clearance efficiency in normal mouse brains. We also assessed the activation status of astrocytes and the infiltration of neutrophils.

Fluoro-Jade B staining

Neuronal degeneration at 72 h post-ICH was evaluated using Fluoro-Jade B (FJB; Biosensis, TR-150-FJB) staining (n = 6 mice per group) [^[13], ^[43]]. Three 30-µm coronal perihematomal sections from the same anatomical coordinates per mouse were mounted on 2% gelatin-coated slides, air-dried at 50 °C for at least 30 min, and processed for FJB staining. Briefly, slides were sequentially immersed in: 1% sodium hydroxide in 80% alcohol (2 min), 70% alcohol (2 min), distilled water (2 min), 0.06% potassium permanganate with shaking (10 min), and distilled water again (2 min). They were then incubated in FJB working solution (0.0004% in 0.1% acetic acid) for 20 min, followed by three 1-min rinses in distilled water. After drying vertically on a paper towel and on a 50 °C slide warmer (5–10 min), slides were cleared in xylene and coverslipped with DPX.

Sections mounted on the slides were imaged using a Nikon Ni-U fluorescence microscope (excitation, 450–490 nm; 20× objective), with four fields captured per Sect. (12 fields/mouse). FJB-positive neurons exhibited characteristic bright green fluorescence, accompanied by degenerative morphological changes (e.g., shrunken somata and fragmented dendrites). FJB-positive cell density (cells/mm²) was quantified using ImageJ 2.0 [^[37]].

Brain water content

Cerebral edema was evaluated at day 3 post-ICH (peak vascular permeability phase; n = 6 mice per group) [^[8], ^[13]]. Mice under deep isoflurane anesthesia were sacrificed, and brains were rapidly dissected into three regions: the ipsilateral striatum, the contralateral striatum, and the cerebellum. Samples were immediately weighed (wet weight, WW) using a precision balance (0.1 mg accuracy), then dried at 100 °C for 24 h to obtain dry weight (DW). Water content was calculated as: [(WW − DW)/WW] × 100%.

Western blotting

To evaluate the expression levels of proinflammatory factors in the perihematomal region at 36 h post-ICH, Western blot analysis was performed (n = 5 mice per group) [^[8], ^[13]]. Mice were anesthetized with 3.0% isoflurane and perfused transcardially using saline to remove circulating blood cells and decrease background signal from plasma proteins. The ipsilateral hemorrhagic hemisphere was rapidly dissected and flash-frozen in liquid nitrogen. Tissue samples were homogenized on ice for 30 min in RIPA lysis buffer containing 1% PMSF protease inhibitor (Solarbio) and then centrifuged at 14,000 × g for 30 min at 4 °C. The supernatant was collected, and protein concentration was measured using an Enhanced BCA Protein Assay Kit (Beyotime, P0010S). Equal amounts of protein (30 µg per lane) were denatured at 99 °C for 10 min and separated on 15% SDS-PAGE gels before being transferred onto PVDF membranes (0.22 μm pore size).

After transfer, membranes were blocked with 5% non-fat milk for 2 h at room temperature and then incubated overnight at 4 °C with primary antibodies: anti-HMGB1 (1:800, Abcam, ab18256), anti-IL-1β (1:600, Abcam, ab200478), anti-RANTES (1:1000, Cell Signaling Technology, 36467), anti-IFN-γ (1:6000, Abcam, ab133566), and β-actin (1:20,000, Proteintech, 66009-1-Ig) as an internal control. After washing with TBST, membranes were incubated for 1 h at room temperature with HRP-conjugated secondary antibodies (anti-rabbit IgG, 1:6000, Proteintech, SA00001-2; anti-mouse IgG, 1:6000, Proteintech, SA00001-1). Protein bands were visualized using an ECL chemiluminescence kit (Affinity, KF001) and imaged on a Bio-Rad ChemiDoc MP system. Band intensities were quantified with ImageJ software and expressed as the ratio of the target protein to β-actin [^[8], ^[13]].

For each sample, both the β-actin and target proteins were processed under the same conditions, and the corresponding bands were used for statistical analysis. To display full-membrane images, strips cut from the same membrane were reassembled and re-exposed. The bands in these reassembled full-membrane images are provided solely as supplementary data and were not used for statistical analysis.

Neurological function evaluation

Neurofunctional recovery was assessed using the modified neurologic deficit score (NDS) and corner turn test (CTT) (n = 12 mice per group) [^[44]]. NDS evaluations were conducted on postoperative days 1, 3, 7, 14, 21, and 28 and included six parameters: body symmetry, gait coordination, circling behavior, forelimb symmetry, climbing ability, and forced turning. Each parameter was scored from 0 to 4, for a total score of 0 to 24. Mice with scores < 10 or > 18 on day 1 were excluded. For the CTT, mice were placed on a platform angled at 30 degrees, and the direction of turns was recorded over 10 trials with at least 1-minute intervals to calculate the percentage of left turns. All tests were performed between 9:00 and 11:00 by two blind investigators under standardized conditions, including a quiet, evenly lit environment and cleaning the platform between trials [^[13], ^[41]].

White matter injury and myelin loss

The extent of white matter damage at 28 days post-ICH was quantified using Luxol Fast Blue (LFB) staining (n = 6 mice per group) [^[13]]. Thirty-micrometer coronal brain sections were processed with standard LFB staining to visualize myelin. For each mouse, three evenly spaced Sect. (360 μm intervals) were selected, and four fields from both the external and internal capsule regions were imaged per section, totaling 12 fields per mouse, using a light microscope with a 20× objective. The percentage of LFB-positive area was measured with ImageJ software and calculated as (area of intact myelin [dark blue staining] / total white matter area) × 100%. All images were captured with fixed settings (exposure time 200 ms, gain: 1.0). Within each image’s defined region of interest (ROI), the LFB-positive area was quantified through manual thresholding to produce a binary segmentation [^[45]]. Although all images originated from subcortical regions, relevant nuclei areas were manually outlined and excluded in each image to determine the total white matter area, which formed the final ROI for analysis.

Statistical analysis

Sample size was determined using a power of 0.9 and an alpha level of 0.05, based on estimates derived from our prior experimental data [^[13]]. Statistical analysis was performed using GraphPad Prism 9.0 (GraphPad Software, LLC) and SPSS 13.0 (SPSS Inc., Chicago, IL). Data distribution characteristics were first assessed using the Kolmogorov-Smirnov normality test and the Levene’s test for homogeneity of variance. For normally distributed data with equal variance, unpaired Student’s t-tests were used for two-group comparisons, while Welch’s t-tests were applied for unequal variances. Non-normally distributed data were analyzed using the Mann-Whitney U test. For multiple group comparisons, one-way ANOVA with Bonferroni post hoc testing was used for normally distributed data, and the Kruskal-Wallis test followed by Dunn’s post hoc test was used for nonparametric data. Longitudinal neurofunctional assessments (including NDS and CTT) were evaluated using generalized estimating equations (GEE, nonparametric), followed by two-tailed paired t-tests or Mann-Whitney U tests as appropriate. Survival rates were compared using chi-square tests. Results are presented as mean ± standard deviation (SD) for normally distributed data or median (interquartile range) for non-normally distributed data. All tests were two-tailed, and P < 0.05 was considered statistically significant. The statistics reporting was provided in the Supplemental Material (Supplemental Table 1).

Result

Anti-BTLA agonist antibody treatment inhibits T lymphocyte infiltration and activation in the hemorrhagic brain of mice with ICH

T lymphocytes begin to infiltrate the perihematomal region within 24 h of ICH, peaking between days 3 and 5 [^[8]]. Flow cytometric analysis was used to examine brain-infiltrating immunocyte subsets at 36 h, 3 days, and 7 days post-ICH (Fig. A). The results showed that intraperitoneal administration of an anti-BTLA agonist antibody significantly reduced the proportions of CD45^high to live cells, CD3⁺ to CD45^high cells, and CD3⁺CD4⁺ to CD3⁺ T cells in the hemorrhagic brain on day 3 after ICH (all P < 0.05; Fig. B and C). However, the treatment had little effect on the percentage of CD3⁺CD8⁺ to CD3⁺ T cells at any of the three time points (All P > 0.05; Fig. C).

Fig. 1The effects of anti-BTLA agonist antibody on T lymphocyte infiltration and activation in the hemorrhagic brain of mice with acute ICH. A Representative gating strategies of T lymphocytes isolated from the hemorrhagic brain showing the effects of anti-BTLA agonist antibody in ICH. B Percentage changes of CD45high and CD3+ cells infiltrating the hemorrhagic brain of mice treated with isotype control or anti-BTLA agonist at 36 h, 3 d, and 7 d after ICH. C Proportional changes of CD3+CD4+ and CD3+CD8+ cells to CD3+ cells in the hemorrhagic brain of mice treated with isotype control or anti-BTLA agonist at the three time points mentioned above. D Analysis of the ratios of CD4+CD69+ and CD8+CD69+ cells to CD3+ cells quantified with flow cytometry in the hemorrhagic brain at each time point. *P < 0.05, **P < 0.01. All data are presented as mean±SD or median (interquartile range). FSC indicates Forward Scatter, and SSC, Side Scatter

To investigate the immunomodulatory effects of anti-BTLA on T lymphocyte activation, we assessed lymphocyte activation status using CD69, an early T cell activation marker [^[46]]. Flow cytometric analysis demonstrated that anti-BTLA agonist antibody treatment significantly reduced the proportion of CD4⁺CD69⁺ T cells to CD3⁺ T cells when compared with the isotype control at 36 h (t = 3.969, P = 0.0033; Fig. D) and 3 days (Z = -12.7, P = 0.0159; Fig. D) after ICH. Furthermore, anti-BTLA agonist antibody treatment reduced the percentage of CD8⁺CD69⁺T cells to CD3⁺T cells in the hemorrhagic brain on day 3 after ICH (t = 3.601, P = 0.007; Fig. D).

Our findings indicate that early treatment with an anti-BTLA agonist antibody significantly suppresses CD4⁺ T cell infiltration and reduces activation of both CD4⁺ and CD8⁺ T lymphocytes in the hemorrhagic brain during the acute phase of ICH. Previous research has shown that T lymphocytes contribute to secondary injury in the early phase after ICH but may promote repair and regeneration at later stages. Therefore, the translational potential of BTLA-targeted agonism as a time-sensitive immunomodulatory approach for ICH warrants further investigation.

Anti-BTLA agonist antibody treatment attenuates CD8⁺ T lymphocyte activation in the peripheral blood of mice with ICH

Clinical studies have shown that the distribution of peripheral blood T lymphocyte subsets is closely associated with disease severity in ICH patients [^[47]]. To further examine BTLA expression dynamics in peripheral immune responses, we performed flow cytometric analysis of circulating T lymphocytes at 36 h, 3 days, and 7 days after ICH (Fig. A). The results revealed that treatment with the anti-BTLA agonist antibody did not significantly change the proportions of CD45⁺ to live cells, CD3⁺ to CD45⁺ cells, and CD3⁺CD4⁺ and CD3⁺CD8⁺ to CD3⁺ T lymphocytes in peripheral blood compared with the isotype control at any of the time points (all P > 0.05; Fig. B and C).

Fig. 2The effects of anti-BTLA agonist antibody on T lymphocyte infiltration and activation in the peripheral blood of mice with acute ICH. A Representative flow cytometry sorting strategies for CD45+ cells and T lymphocyte subsets isolated from the peripheral blood following acute ICH. B Percentage alterations of CD45+ and CD3+ cells in the peripheral blood of mice treated with isotype control or anti-BTLA agonist at 36 h, 3 d, and 7 d after ICH. C Quantification of the percentages of CD3+CD4+ and CD3+CD8+ cells to CD3+ cells in the peripheral blood of mice treated with isotype control or anti-BTLA agonist at the above three time points. D Comparison of the proportions of CD4+CD69+ and CD8+CD69+ cells to CD3+ cells quantified with flow cytometry in the peripheral blood at corresponding time points. *P < 0.05, **P < 0.01. All data are presented as mean±SD or median (interquartile range). FSC indicates Forward Scatter, and SSC, Side Scatter

Meanwhile, the anti-BTLA agonist antibody treatment did not change the proportion of CD4⁺CD69⁺ T lymphocytes to CD3⁺ T lymphocytes at any of the three time points (all P > 0.05; Fig. D). However, it significantly reduced the proportion of CD8⁺CD69⁺ T lymphocytes to CD3⁺ T lymphocytes on day 3 (t = 2.962, P = 0.0181; Fig. D) and 7 (t = 3.551, P = 0.0062; Fig. D) after ICH.

These findings indicate that systemic administration of the anti-BTLA agonist antibody does not alter the distribution of T lymphocyte subsets but significantly suppresses CD8⁺ T lymphocyte activation over a prolonged period following acute ICH. Given the increased risk of infectious complications associated with stroke-induced immunosuppression [^[13], ^[48]]. Further studies should explore whether anti-BTLA therapy influences this susceptibility.

Anti-BTLA agonist antibody treatment reduces cerebral lesion volume and edema without mitigating overall brain swelling after ICH

Because of hematoma breakdown, cerebral edema in mice peaked on the third day after ICH [^[8]]. To assess early brain injury at this time point, we used Luxol Fast Blue (LFB)/Cresyl Violet (CV) staining to measure lesion volume, brain swelling, and brain water content. Pathological examination on day 3 post-ICH showed marked demyelination and neuronal injury in both the isotype control and anti-BTLA-treated groups. Lesion areas were identified as regions of myelin loss and lack of staining on LFB/CV-stained sections (Fig. A).

Fig. 3The influence of anti-BTLA agonist antibody treatment on acute-phase brain injury following ICH. A Representative LFB/CV-stained sections demonstrating pathological changes in brain tissue on day 3 post-ICH. The dashed black line demarcates the lesion area (scale bar = 1 mm). B and C Quantitative analysis of brain lesion volume and brain swelling in mice that received isotype or anti-BTLA agonist 3 d after ICH. D Brain water content in the ipsilateral and contralateral striata and the cerebellum of mice with or without anti-BTLA agonist administration 3 d after sham operation or ICH was analyzed statistically. *P < 0.05, **P < 0.01. All data are presented as mean±SD

Quantitative analysis revealed that the anti-BTLA agonist antibody significantly reduced the lesion volume (t = 3.169, P = 0.0053; Fig. B) but not brain swelling (t = 0.987, P = 0.3755; Fig. C) compared with the isotype control. Possibly due to differences in sensitivity profiles, the anti-BTLA agonist antibody also significantly reduced brain water content in the ipsilateral striatum, the region affected by the hemorrhage, compared with isotype-treated mice (P = 0.0027; Fig. D). Notably, neither ICH nor the anti-BTLA agonist antibody significantly affected water content in the contralateral striatum (P = 0.05673) or cerebellum (P = 0.2414). The inhibitory effects of anti-BTLA agonist antibody treatment on T lymphocyte infiltration and activation may partially contribute to these outcomes.

Anti-BTLA agonist antibody treatment inhibits cellular and molecular neuroinflammation around the hematoma after ICH

The crosstalk between infiltrating immune cells and resident brain cells around the hematoma is well established [^[24]]. In the collagenase-induced ICH model, microglial activation peaked at 3–7 days, neutrophil infiltration was most significant at day 3, and astrocyte activation became prominent after day 3. Using immunofluorescence, we found that treatment with an anti-BTLA agonist antibody markedly reduced microglia/macrophage (IBA-1-positive) activation in the perihematomal regions compared with the isotype control on day 3 after ICH (t = 3.795, P = 0.0035; Fig. A and C). Furthermore, the treatment significantly inhibited neutrophil infiltration (MPO-positive cells; Z = 2, P = 0.0087; Fig. A and C) at this time point. Although anti-BTLA agonist antibody treatment showed a tendency to reduce astrocyte activity (GFAP-positive cells, t = 1.681, P = 0.1236) and neurodegeneration (FJB-positive cells, t = 1.394, P = 0.1936) in the perihematomal regions on day 3 after ICH, these effects did not reach statistical significance (Fig. A and C).

Fig. 4The influence of anti-BTLA agonist antibody treatment on cellular and molecular neuroinflammation following acute ICH. A Schematic illustrating selected regions from three comparable sections per mouse for quantitative analysis of Iba-1+, GFAP+, MPO+, and FJB+ cells in the perihematomal regions. B Representative immunofluorescence images of Iba-1 (microglia/macrophages), GFAP (astrocytes), and MPO (neutrophils), along with FJB staining for degenerating neurons in perihematomal regions at day 3 post-ICH. Insets show magnified views of positive cells (scale bar = 50 μm). C Scatter plots quantifying Iba-1+, GFAP+, MPO+, and FJB+ cell densities. D Representative Western blot bands of IL-1β, IFN-γ, RANTES, and HMGB1 expression in tissue surrounding the hematoma at 36 h post-ICH. β-actin was used as the loading control. E Quantification of IL-1β, IFN-γ, RANTES, and HMGB1 expression in the hemorrhagic brain 36 h after ICH in sham-operated and ICH mice treated with isotype control or anti-BTLA agonist. Bands used for statistical analysis were those exposed individually, not the re-exposed bands from the reassembled membranes provided in the supplementary materials. *P < 0.05, **P < 0.01. All data are presented as mean±SD or median (interquartile range)

Infiltrating and resident immune cells contribute to brain injury by producing pro-inflammatory cytokines after ICH [^[49]]. Using western blotting, we evaluated the expression of IL-1β, IFN-γ, RANTES, and HMGB1 in the perihematomal regions 36 h after ICH. The results revealed that their expression increased significantly after ICH (All P < 0.05; Fig. D and E). However, anti-BTLA agonist antibody treatment markedly reduced the expression of IL-1β, IFN-γ, and RANTES at the analyzed time point (All P < 0.05; Fig. D and E). Although the treatment exhibited a trend toward reduced HMGB1 expression, this reduction did not reach statistical significance (P = 0.147; Fig. D and E).

These findings suggest that anti-BTLA agonist antibody treatment may alleviate brain injury by inhibiting neuroinflammation after ICH. Although we did not observe a positive effect of anti-BTLA agonist antibody on neurodegeneration 3 days after ICH, we assessed its impact on long-term brain injury using histological staining to determine whether it exerted a delayed therapeutic effect.

Anti-BTLA agonist antibody treatment reduces residual lesion volume and improves long-term functional recovery in ICH

Functional recovery following ICH typically occurs over the long-term period [^[50]]. To evaluate the effect of anti-BTLA agonist antibody on long-term brain injury, we examined histological changes in LFB/CV-stained brain sections at day 28 post-ICH. The results demonstrated a significant reduction in residual lesion volume in the anti-BTLA-treated group compared with the isotype control group (t = 2.831, P = 0.0097; Fig. A and B), indicating sustained neuroprotective effects. However, no statistically significant differences were observed between groups in either brain atrophy degree (t = 1.856, P = 0.0769; Fig. A and B) or white matter injury severity (t = 1.241, P = 0.2278; Fig. C and E).

Fig. 5The influence of anti-BTLA agonist antibody treatment on long-term brain injury severity post-ICH. A Representative LFB/CV-stained images at 28 days post-operation (black dashed line indicates lesion area; red arrows mark residual lesions; scale bar = 1 mm). B Analysis of residual lesion volume and brain atrophy in ICH mice that received either an isotype control or an anti-BTLA agonist. C A schematic diagram of the fields selected for quantifying myelin loss. D Representative images in the perihematomal regions stained with LFB/CV 28 d after ICH in mice that received isotype control or anti-BTLA agonist (scale bar = 1 mm). E Quantitative analysis of white matter damage 28 d after ICH in mice. **P < 0.01. All data are presented as mean±SD

To systematically evaluate the long-term effects of anti-BTLA agonist antibody on neurological function post-ICH, we conducted a 28-day longitudinal evaluation using neurological deficit scores (NDS) and the corner turn test (CTT) (Fig. A and B). Treatment with the Anti-BTLA agonist antibody significantly improved leftward turning on days 14 and 28 post-ICH (both P < 0.05; Fig. A). It also markedly enhanced the total NDS scores on days 3, 7, 14, 21, and 28 (All P < 0.05; Fig. B). Moreover, the antibody treatment produced notable improvements in specific behavioral domains, including body symmetry on day 3 and gait on day 7, as well as climbing ability on days 3 and 28 and forelimb symmetry on day 21 post-ICH (All P < 0.05; Fig. C).

Fig. 6The influence of anti-BTLA agonist antibody treatment on neurologic deficits throughout the 28-day research period in mice with ICH. A and B Evaluation of neurologic deficit with the corner turn test and NDS 1, 3, 7, 14, 21, and 28 days after ICH. C Individual neurologic function assessments across the 28-day research period. D and E Dynamic changes in body temperature and body weight occurred throughout the 28-day research period. *P < 0.05, **P < 0.01. All data are presented as mean±SD or median (interquartile range)

During the experimental period, anti-BTLA agonist antibody treatment did not significantly affect mortality rates in mice (χ² = 0.2614, P = 0.609). The mortality rates were 13.33% (8/60) in the ICH + isotype control group and 16.67% (10/60) in the ICH + anti-BTLA agonistic antibody group. Similarly, there were no significant changes in rectal temperature or body weight over the 28-day study periods (both P > 0.05; Fig. D and E). Overall, these findings suggest that modulating BTLA signaling can reduce acute-phase brain injury and promote long-term functional recovery, highlighting its potential for clinical use.

Microglial depletion abolishes the suppressive effects of anti-BTLA agonist antibody on molecular neuroinflammation in ICH

BTLA is a well-recognized immune checkpoint molecule expressed on immune cells [^[51]]. Similar to other immune checkpoint pathways, such as PD-1 and CTLA-4 [^[52]], our study identified for the first time that BTLA is co-expressed in microglia (Fig. A). To determine whether the anti-BTLA agonist antibody suppresses the immune responses in the hemorrhagic brain by acting on infiltrating T lymphocytes or on microglia/macrophages, we examined its effects in mice depleted of microglia/macrophages. Feeding mice a diet containing the CSF1R inhibitor PLX5622 for 14 days effectively depleted microglia (t = 13.47, P = 0.0002; Fig. B and C).

Fig. 7The influence of microglia depletion on the effects of an anti-BTLA agonist for molecular neuroinflammation in mice with acute ICH. A Double fluorescent staining of Iba-1 (green) and BTLA (red) in the perihematomal region 3 days after ICH. Nuclei were counterstained with DAPI (blue). Merged images are shown (scale bar = 50 μm). B Representative immunofluorescence images showing Iba-1+ cells in the perihematomal region of microglia-depleted mice 3 days post-ICH (scale bar = 50 μm). C Quantitative analysis of Iba-1+ cells in mice that received vehicle or CSF1R inhibitor, which was used for the depletion of microglia on day 3 after ICH. D Representative Western blot bands of IL-1β, IFN-γ, RANTES, and HMGB1 expression in tissue surrounding the hematoma at 36 h post-ICH. E Quantitative analysis of IL-1β, IFN-γ, RANTES, and HMGB1 expression in ICH mice that received vehicle or CSF1R inhibitor. Statistical analysis was performed on individually exposed bands, not on the re-exposed versions presented in the reassembled full-membrane supplementary figures. **P < 0.01. All data are presented as mean±SD

After microglia depletion, the anti-BTLA agonist antibody had no significant effect on the expression of IFN-γ (t = 1.672, P = 0.1255), IL-1β (t = 1.182, P = 0.2646), RANTES (t = 0.3544, P = 0.7304), or HMGB1 (t = 0.7359, P = 0.4787) in the perihematomal region compared with the isotype control at 36 h after ICH (Fig. D and E). Although we did not directly assess whether the antibody crosses the blood-brain barrier, our results suggest that its effects may involve inhibition of microglial activation in the hemorrhagic brain or suppression of the cross-talk between T lymphocytes and microglia.

Discussion

Multiple studies have demonstrated that both innate and adaptive immunity are pivotal in driving pro-inflammatory responses in the hemorrhagic brain following acute ICH [^[12], ^[53], ^[54]]. BTLA, an inhibitory receptor expressed on most lymphocytes, helps to restrain excessive immune activation [^[55], ^[56]]. In this study, we used an anti-BTLA agonist antibody to explore its role in ICH. Our results demonstrate that anti-BTLA treatment primarily acts through microglia to limit T cell infiltration and activation around the hematoma, thereby reducing brain lesion volume and edema, and improving long-term neurological recovery. Additionally, anti-BTLA treatment also suppresses early molecular and cellular neuroinflammation following ICH. Together, these findings suggest that BTLA may serve as a promising target for immunomodulation in the treatment of ICH.

Fundamental research has shown that BTLA is constitutively expressed in naive T cells, transiently upregulated during early activation, and markedly downregulated during prolonged activation phases [^[57]]. This highly dynamic expression pattern suggests that BTLA may function as a “molecular brake” on T cell activation. In mice with acute ICH, treatment with an anti-BTLA agonist antibody reduced the infiltration of CD45^high, CD3⁺, and CD4⁺ T cells, as well as the activation of both CD4⁺ and CD8⁺ T cells, in the hemorrhagic brain. In contrast, its peripheral effects were more limited, showing inhibition only of CD8⁺ T cell activation in the blood during the acute phase. Although these differential effects on CD4⁺ and CD8⁺ T cell infiltration and activation between the brain and periphery may reflect differences in their relative abundances, the precise mechanisms underlying this selective immunomodulation remain to be elucidated.

Although limitations exist, the impact of peripheral immune cell infiltration on the immune-inflammatory response in the hemorrhagic brain has been well established [^[3], ^[24]]. Studies have shown that peripherally infiltrated T lymphocytes interact with the brain’s resident immune cells, microglia, amplifying cellular and molecular immune responses in the hemorrhagic brain and exacerbating hemorrhagic brain injury [^[9], ^[47]]. In addition to inhibiting the infiltration and activation of T lymphocytes, the anti-BTLA agonist antibody treatment also reduced the extent of brain injury and brain tissue water content during the acute phase of hemorrhage. Although it did not prevent neurodegeneration, the therapy significantly reduced neuroinflammation around the hemorrhage site following acute ICH. Specifically, it suppressed microglia and macrophage activation, limited neutrophil infiltration, and decreased levels of pro-inflammatory cytokines, including IL-1β, IFN-γ, and RANTES, which are associated with the Th1 immune response [^[8]]. These results are consistent with the inhibitory effects of the anti-BTLA agonist on T cell infiltration and activation in the hemorrhagic brain of mice with acute ICH.

Since no effect on neurodegeneration was observed during the acute phase, we then investigated the impact of anti-BTLA agonist antibody on long-term brain injury and prognosis after ICH to further evaluate its therapeutic value. In the meantime, we observed the influence of this antibody on mortality, rectal temperature, and body weight changes in ICH mice. While it did not reduce the degree of brain atrophy or white matter injury, this antibody decreased the size of the residual lesion at 28 days after ICH. Meanwhile, it also promoted long-term neurological recovery without affecting mouse mortality, rectal temperature, or body weight, demonstrating a favorable safety profile. In this study, the pathological changes during the acute and long-term phases after ICH were consistent, suggesting that the antibody may exert its therapeutic effects primarily by limiting injury.

While BTLA is known to be expressed on various immune cells, including T cells, dendritic cells, and monocytes [^[58], ^[59]], our data demonstrate that it is also highly expressed in microglia. Given the documented crosstalk between microglia and T cells, where microglia activate T cells via antigen presentation and are in turn activated by T-cell-derived IFN-γ [^[59]], we used the CSF1R inhibitor PLX5622 to deplete microglia to determine whether they are essential for the anti-BTLA agonist’s mechanism of action [^[32], ^[60]]. The results showed that depleting microglia abolished the anti-BTLA agonist’s ability to suppress pro-inflammatory cytokines (IL-1β, IFN-γ, RANTES) in the perihematomal region during acute ICH. This suggests that BTLA has anti-inflammatory effects that depend on microglia. However, it is still unclear whether BTLA acts directly on microglia or indirectly by reducing T-cell infiltration. To clarify this, studying activated microglia requires more precise morphometry, such as measuring soma size, circularity, ramification index, and using Sholl analysis [^[61]]. The current approach, which uses a 7.5 μm size threshold and morphological assessment, is only qualitative or semi-quantitative.

This study has several limitations. First, the use of only adult male mice limits the ability to assess potential differences in ICH susceptibility and responses to the anti-BTLA agonist antibody across sex and age groups. Second, although a single intraperitoneal antibody injection has been shown to suppress brain-infiltrating T lymphocytes for six days [^[30]], the dose–response relationship and therapeutic window remain undefined. Clarifying these parameters is essential for clinical translation. Third, as an immunomodulatory therapy, the systemic administration of the anti-BTLA antibody may increase the risk of post-ICH infections [^[62]], which was not directly evaluated in this study. Since BTLA is expressed on both infiltrating T cells and resident microglia in the brain, we used microglia-depleted mice to further investigate the interaction between T lymphocytes and microglia in this process. However, the exact mechanism by which the BTLA agonist antibody regulates immune-inflammatory responses in the hemorrhagic brain remains unclear.

Current evidence suggests that the blood-brain barrier is compromised following an ICH [^[24], ^[63]]. However, the effectiveness of antibody passage across this barrier during this time remains uncertain. The possibility that the BTLA antibody can act as either an agonist or an antagonist highlights its complex role. While most studies focus on the anti-BTLA agonistic antibody (clone 6A6, BioXcell) primarily as a BTLA agonist [^[20], ^[31]], its datasheet al.so indicates a potential for use as a BTLA blocker under certain experimental conditions. Our findings are consistent with observations from BTLA knockout models in systemic inflammatory contexts, such as endotoxemia and sepsis [^[20], ^[31]]. Nevertheless, the specific role of this antibody in collagenase-induced ICH remains unclear. Future research using BTLA knockout models or in vitro T cell activation assays is necessary to better understand the antibody’s ability to penetrate and function within the altered microenvironment of ICH. Additionally, using microglia-specific BTLA knockout mice would help clarify the relative contributions of T lymphocytes and microglia in this process.

In summary, this study demonstrates that systemic administration of an anti-BTLA agonist antibody inhibits the infiltration and activation of circulating T lymphocytes into the hemorrhagic brain following acute ICH. The treatment also reduced molecular and cellular neuroinflammation, mitigated both short- and long-term brain injury, and promoted long-term functional recovery. Additionally, we found that BTLA is co-expressed in microglia within the hemorrhagic brain. Depletion of microglia abolished the antibody’s inhibitory effects on molecular neuroinflammation during the acute phase of ICH. While several limitations exist, these findings suggest that BTLA, an essential immune checkpoint, may be a target for immunoregulation in the treatment of ICH. They also have implications for the development of therapeutic strategies targeting infiltrating T lymphocytes or innate microglia in the lesioned brain to treat ICH and other acute stroke conditions.

Supplementary Information

Supplementary Material 1.