Pro-repair macrophages driven by CGRP rescue white matter integrity following intracerebral hemorrhage

Intracerebral hemorrhage (ICH) triggers a dynamic immune response involving macrophages, However, the functional heterogeneity of these cells and the mechanisms through which they promote repair remain unclear. Although the neuropeptide CGRP has been shown to modulate macrophage phenotypes in other pathological contexts, its role in ICH recovery and white matter repair remains unexplored.

Single-cell RNA sequencing (scRNA-seq) of CD45 + cells from ICH mice (GSE167593 and GSE230414 datasets) identified macrophage subsets. Flow cytometry, diffusion tensor imaging (DTI), behavioral assays, and immunofluorescence validated macrophage dynamics and white matter (WM) integrity. Bone marrow and skull analyses traced macrophage origins. The role of calcitonin gene-related peptide (CGRP) was tested via intraperitoneal administration in ICH mice, with outcomes assessed through transcriptomics, ultrastructural imaging, and functional recovery.

scRNA-seq revealed sustained accumulation of THBS1 + macrophages post-ICH, correlating with WM repair and neurological recovery. These macrophages exhibited pro-repair and remyelination gene signatures (e.g., Arg1, Tgm2). Bone marrow-derived myeloid cells, particularly skull-resident populations, served as the primary source of THBS1 + macrophages. CGRP, elevated in meninges and bone marrow post-ICH, drove macrophage polarization toward THBS + phenotypes. CGRP administration expanded THBS1-positive macrophages in the bone marrow and brain, improving WM integrity (reduced radial diffusivity, higher fractional anisotropy) and sensorimotor function. Ultrastructural analysis confirmed enhanced myelin regeneration (lower g-ratio) in CGRP-treated mice.

This study identifies a neuroimmune axis wherein CGRP promotes bone marrow-derived THBS1 + macrophages to facilitate WM repair and functional recovery after ICH. Targeting CGRP-macrophage signaling offers a therapeutic strategy to enhance recovery in hemorrhagic brain injury.

Intracerebral hemorrhage (ICH) is a severe neurological disorder characterized by the accumulation of blood within the brain tissue, leading to significant morbidity and mortality [1, 2]. Despite advances in surgical technique and clinical management, the long-term consequences of ICH, particularly the inability of the brain to fully recover functional integrity, remain a critical challenge in clinical neuroscience [3, 4].

ICH triggers a significant immune response involving various immune cells. Studies have identified gene expression changes in peripheral blood, highlighting the involvement of neutrophils, macrophages, and T cells in the inflammatory and immune pathways following ICH [5]. Beyond their well-characterized role in initiating the neuroinflammatory cascade following cerebral injury [6,7,8], emerging evidence underscores the indispensable neuroprotective functions of these immune cells in orchestrating tissue repair during the recovery phase [9,10,11,12]. Yet the complexity of immune cell dynamics and their functional contributions in ICH has not been fully elucidated. Specifically, the heterogeneity of macrophages and their dynamic changes throughout the course [13] of the ICH challenge a ‘one-size-fits-all’ approach to macrophage-targeted treatment in ICH, underscoring the need for a deep understanding of their remarkable plasticity.

The role of neuropeptide CGRP (calcitonin gene-related peptide) in modulating immune and neural functions has been recognized in various pathological conditions [14,15,16,17]. After tissue injury, CGRP may play a role in modulating the transition of macrophages toward a reparative phenotype [17]. Moreover, it has demonstrated therapeutic potential in models of traumatic brain injury and subarachnoid hemorrhage [18, 19]. However, the specific role of CGRP in ICH remains unclear, which is particularly critical given its implicated dual functions in both anti-inflammatory and neuroprotective pathways.

This study employed single-cell transcriptomic approaches to investigate the dynamic changes in immune cell populations following ICH. Our findings reveal that macrophages, particularly those expressing THBS1, exhibit sustained contributions to the brain’s repair process. Furthermore, we identified a novel role for CGRP in modulating the phenotypic transition of macrophages toward a remyelination phenotype in bone marrow. These discoveries provide new insights into the immune and neural signaling pathways that drive brain repair after ICH, highlighting potential therapeutic targets for improving functional recovery in this condition.

Mice

Male C57BL/6J mice, aged 8–10 weeks and weighing 20–25 g, were sourced from Charles River Laboratory (Zhejiang, China). The animals were housed in specific pathogen-free(SPF) conditions under a 12-hour light/dark cycle, with ad libitum access to food and water. All procedures in this study complied with the Guide for the Care and Use of Laboratory Animals, as outlined by the National Institutes of Health. The Institutional Ethics Committee of the Second Affiliated Hospital, Zhejiang University School of Medicine, approved and oversaw experimental protocols.

Establishment of the ICH model

During the perioperative period, mice received an intraperitoneal injection of ketoprofen (1 mg/kg) for analgesia before surgery. Anesthesia was induced with 5% isoflurane and maintained at 2% isoflurane using a homeothermic monitoring system to regulate body temperature. Pain levels were assessed via toe clamping during the procedure. Subsequently, collagenase (type VII, derived from Clostridium histolyticum; Sigma-Aldrich) was prepared at a concentration of 0.05 U in 0.5 µl of saline and stereotactically injected into the right basal ganglia (2.5 mm lateral to the bregma, 3 mm deep at a 5° angle) over 5 min [20]. Afterward, a 5-minute interval was allowed to permit any potential reflux. Rectal temperature was continuously monitored throughout the ICH induction, maintained at 37.0 °C ± 0.5 °C. Mice in the sham group underwent the same procedure, including needle insertion, except for the collagenase injection.

Immunofluorescence staining

Mice were first anesthetized with an intraperitoneal injection of 1% pentobarbital sodium and transcardially perfused with 20 mL of ice-cold 0.1 M phosphate-buffered saline (PBS), followed by 4% paraformaldehyde (PFA) (BL539A, Biosharp, China). The brains were removed, fixed in 4% PFA for 24 h, and then dehydrated in 30% sucrose at 4 °C until they sank. After fixation and dehydration, the brains were embedded in an optimal cutting temperature compound, and coronal Sect. (10 μm) were prepared using a cryostat. The cryosections were blocked and permeabilized with a blocking buffer (P0260, Beyotime, China) for one hour at room temperature. The sections were then incubated with the primary antibody overnight at 4 °C, including F4/80(1:200, Thermo Fisher 13-4801-82), THBS1(1:200, Thermo Fisher, MA5-13398), NF(1:200, Abcam, ab207176), and MBP (1:500, Thermo Fisher,83683 S). Wash the cryosections with PBS and incubate them with species-specific secondary antibodies for 1 h at room temperature.

As described previously, the skull and femur were harvested and subjected to immunofluorescence staining [21]. Mice were euthanized via transcardial perfusion with PBS and 4% PFA. The skulls and femurs were then collected and immediately fixed in ice-cold 4% PFA for 6–8 h under gentle agitation. Following fixation, the bones were decalcified in 0.5 M EDTA for 3 days (skulls) or 7 days (femurs) at 4 °C with gentle shaking. Next, 10-µm-thick cryosections were prepared for immunofluorescence staining. These sections were blocked and permeabilized using blocking buffer (P0260, Beyotime, China) for 1 h at room temperature, then incubated overnight at 4 °C with primary antibodies—CD11b (1:100, Abcam, ab128797) and THBS1 (1:200, Thermo Fisher, MA5-13398). Finally, the sections were washed with PBS and incubated with the appropriate secondary antibodies.

Dissect the meninges and perform immunofluorescence staining as previously described [15]. The cortical meninges were dissected and post-fixed in PBS/4% PFA solution at 4 °C for 24 h. Before immunostaining, samples were transferred to PBS and incubated for 24 h at 4 °C to remove PFA. Free-floating samples were incubated, blocked, and permeabilized with blocking buffer (P0260, Beyotime, China) in 24-well plates for one hour at room temperature with agitation. The blocking solution was then replaced by a staining solution containing the anti-CGRP antibodies (1:100, Abcam, ab81887) and anti-Tubb3 antibodies (1:500, Sigma-Aldrich, T8578), followed by incubation of the samples for 24 hours at 4 °C with agitation incubated for 24 h at 4 °C with agitation. Samples were washed five times with PBS to remove unbound primary antibodies and then incubated with secondary antibodies.

Subsequently, the sections were incubated with DAPI (ab104139, Abcam, UK) to counterstain the nuclei. Fluorescence imaging was conducted using LAS X software with a confocal microscope (Leica DMI8). The system was optimized for full tissue depth (z-axis) and tile-stitching (xy-axes), and merged maximum projection images were exported. Three-dimensional analysis of brain slices was conducted using the z-stack feature of a confocal microscope, with images reconstructed using the Vaa3D 3D rendering software [22].

Western blot analysis

Anesthetized mice were perfused with PBS. Brains were sliced into 1-mm coronal sections, and basal ganglia tissue with hematoma was homogenized in RIPA buffer (Beyotime, Shanghai, China). Protein concentrations were measured using a BCA assay (Thermo Fisher Scientific, Waltham, MA, USA). For Western blot, 40 µg protein samples were separated on 10% SDS-PAGE gels, transferred to nitrocellulose membranes (GE Healthcare) at 300 mA for 90 min, and blocked in 5% non-fat milk in TBST (25 mM Tris-HCl, 137 mM NaCl, 0.1% Tween-20, pH 7.6) for 1 h. Membranes were incubated with MBP (Proteintech, 10458-1-AP, China) primary antibodies overnight at 4 °C, followed by HRP-conjugated secondary antibodies (Beyotime, Shanghai, China) for 1 h. Bands were visualized using ECL reagent (Pierce, Thermo Fisher Scientific) on a ChemiDoc XRS+ (Bio-Rad, Hercules, CA, USA) and quantified with ImageJ (NIH).

Generation and adaptive transfer of bone Marrow-derived macrophages (BMDMs)

BMDMs were generated as previously described [23, 24]. Bone marrow cells from mice were cultured in DMEM (Gibco, 11965092, USA) supplemented with 10% fetal bovine serum (FBS), 1% penicillin-streptomycin, and 20 ng/mL M-CSF (MCE, HY-P7085A, USA). On days 4 and 7, half of the medium was replaced with fresh medium. After 7 days, mature BMDMs were pretreated with CGRP or PBS and dissociated into single-cell suspensions. For in vivo studies, mice were administered intravenous injections of 3 × 10^6 CGRP- or PBS-treated BMDMs via the retro-orbital venous sinus on day 7 post-ICH induction.

ELISA assay

Following euthanasia, the skulls (with intact meningeal layers) and femoral tissues of the experimental animals were harvested. Samples were promptly transferred to 24-well culture dishes containing 1 ml of DMEM. The tissue cultures were kept at 32 °C for 30 min with constant agitation at 150 rotations per minute. Following incubation, the culture medium was collected and subjected to CGRP quantification using a specific enzyme immunoassay kit (Cayman Chemical), following the supplier’s recommended procedures.

Flow cytometry

Animals were euthanized and perfused with cold saline. The brains were dissected, and the ipsilateral hemispheres were collected. Brain homogenates were then prepared using the Neural Tissue Dissociation Kit (T) in combination with a GentleMACS Dissociator with heaters (Miltenyi Biotec). The cell suspension was passed through a 70 μm cell strainer (Thermo Fisher Scientific) and subjected to a 30% Percoll (17089109, Cytiva, USA) density gradient centrifugation to remove myelin. Bone marrow and spleen cells were prepared as described previously [15, 21]. These cells were resuspended in RBC lysing buffer (Thermo Fisher, 00-4333-57) and treated with an anti-CD16/CD32 antibody to block Fc receptors. Then, the cells were incubated with cell surface antibodies for 30 min at 4 °C: CD45 (1:100, Biolegend,103132), ly6G (1:100, Biolegend,127616), CD3 (1:100, Biolegend,100204), CD11b (1:100, Thermo Fisher,25-0112-82), F4/80 (1:100, Thermo Fisher,48-4801-82). After fixation and permeabilization with an Intracellular Fixation and Permeabilization Buffer Set (Thermo Fisher, 88-8824-00), cells were incubated with THBS1 (1:100, Thermo Fisher, MA5‑13398) or an isotype control for 60 min at room temperature. Following this, the cells were blocked in a 2% BSA‑PBS solution for 30 min at room temperature, incubated in the dark with a Dylight 488-conjugated goat anti‑mouse IgG (H + L) secondary antibody, and finally resuspended in PBS. Different immune cell populations were analyzed by flow cytometry using a Beckman Coulter flow cytometer (Brea, CA, USA) equipped with Cytexpert Software (Version 2.4.0.28). Different immune cell populations were analyzed by flow cytometry using a Beckman Coulter flow cytometer (Brea, CA, USA) equipped with Cytexpert Software (2.4.0.28). Data were further processed and analyzed using FlowJo software (10.1.8).

Behavioral tests

Accelerated rotarod test [25] and adhesive removal test [26] were used to detect neurological function in ICH mice. The rotating rod began at 5 rpm, accelerated at 0.2 rpm/s, and reached a maximum speed of 65 rpm; the time and speed at which each mouse fell were recorded. For the adhesive removal test, a 3 mm × 3 mm adhesive tape was applied to the left (affected) forepaw, and the latency to contact and remove the tape was recorded, with a maximum duration of 120 s. Before modeling, mice were pre-trained on the rotarod and adhesive removal tests for three days, with the final test results as the baseline.

Transmission electron microscopy (TEM)

Following the procedure described in [27], mice underwent intracardiac perfusion with PBS, 4% PFA, and 2.5% glutaraldehyde. Tissue cubes (1 mm³) were carefully microdissected from the striatum and subsequently immersed in 2.5% glutaraldehyde for 24 h. The samples were then post-fixed in 1% osmium tetroxide for 2 h before being sequentially dehydrated in ethanol and acetone. After embedding in epoxy resin, ultrathin sections were prepared and stained with uranyl acetate and lead citrate. Randomly selected fields within the striatum were imaged using TEM (Philips Tecnai 10), and the G-ratio—a critical measure of myelination—was determined and analyzed with ImageJ software.

Diffusion tensor imaging (DTI)

As previously described, white matter integrity was assessed using DTI [28]. Mice were anesthetized with 1–3% isoflurane delivered in compressed air (1.5 L/min), with body temperature consistently maintained at 37°C and respiratory rates continuously monitored during all magnetic resonance imaging (MRI) scans conducted on a 7.0 T scanner (BioSpec 70/16 USR, Bruker BioSpin MRI, Ettlingen, Germany).T2-weighted imaging was performed with the following parameters: TR/TE = 3000 ms/35 ms, matrix size = 256 × 256, FOV = 18 × 18 mm², slice thickness = 0.5 mm, 35 slices, excitation/refocusing flip angles = 90°/180°, two averages, scan time of 3’12” seconds. DTI utilized an EPI sequence with TR/TE = 1800 ms/24.33 ms, matrix size = 96 × 96, FOV = 15 × 15 mm², 9 slices, slice thickness = 0.8 mm, flip angle = 90°, scan time 16 ‘48 ‘’, and a b-value of 600 s/mm². DTI data were processed using DSI Studio [29] software (http://dsi-studio.labsolver.org) to generate directionally encoded color (DEC), fractional anisotropy (FA), and radial diffusivity (RD) maps. Regions of interest (ROIs) were manually delineated in a blinded manner to encompass the EC and IC in ipsilateral hemispheres for FA and RD quantification.

Single-cell RNA sequencing data and RNAseq data analysis

Single-cell transcriptomic datasets (GSE167593 and GSE230414) [7, 30]were reanalyzed through Seurat (v5.1.0). Specifically, CD45 + immune cell populations from the 14-day ICH were isolated from GSE167593 and subsequently integrated with GSE230414. Quality control measures were implemented to exclude cells exhibiting insufficient or excessive gene detection (< 400 or > 4,000 expressed genes). Additionally, cellular samples demonstrating elevated mitochondrial gene content (> 30% of total gene expression) were systematically removed to ensure data integrity. To address potential technical variability between datasets, batch effect correction was performed using the Harmony(1.2.1). Dimensionality reduction and cellular population visualization were achieved through Uniform Manifold Approximation and Projection (UMAP) embedding. Following initial clustering analysis, a focused investigation of macrophage subpopulations was conducted by applying identical computational parameters as those utilized in the primary analysis.

The raw count data from the GSE255049 [17] dataset were analyzed using the Degust online platform. Transcriptomic profiling was conducted employing the limma/voom [31] framework within the Degust interface, with statistical significance determined by a false discovery rate (FDR)-adjusted P-value threshold of < 0.05 for identifying differentially expressed genes.

Gene set enrichment analysis (GSEA)

GSEA analyses were performed using the ClusterProfiler (4.8.4) to identify the enriched gene set clusters in macrophage subpopulations.

Statistical analyses

Results are presented as mean ± standard deviation (SD). GraphPad Prism software (version 8.0.2) was used for statistical analyses. The Student’s t-test (equal variances) or Welch’s t-test (unequal variances) was used to compare two groups for continuous variables with normal distributions. The Mann-Whitney U rank sum test was used for continuous variables with non-normal distributions. The results from the neurological function were analyzed using two-way repeated-measures ANOVA.

Dynamic changes in immune cells following ICH

The scRNA-seq dataset was retrieved from the GEO database. CD45 + cells from ICH 14-day mice brain in the GSE167593 dataset were extracted and subsequently merged with the GSE230414 dataset (including sham and ICH 3-day) to systematically analyze the major immune cell types in the brain parenchyma and their temporal dynamics in ICH mice. A UMAP embedding space and a graph-based clustering approach were also utilized to identify distinct cell clusters. The cells were classified into five major cell types (Fig. 1A) based on the enriched genes for each cluster (Fig. 1B, C). As expected, we observed that each cell cluster’s relative abundance and genes (Figure S1) changed significantly with ICH. For example, T lymphocytes increase considerably at 14 days (Fig. 1D). Although multiple studies have reported neutrophil infiltration after ICH [32,33,34], the changes between the sham group and 3 days after ICH are not significant (Fig. 1D), which may be since neutrophil accumulation peaks at earlier time points and becomes less prominent compared to other cell types by day 3 [35, 36]. Conversely, Macrophages rose after ICH and continued until 14 days (Fig. 1D). This result is consistent with that of cerebral infarction [37].To verify this result, we took the ipsilateral brain tissue of mice after ICH. We performed flow cytometry to measure the proportion of infiltrating immune cells, and the flow cytometry results were consistent with the above changes (Fig. 1E). The sustained rise of macrophages at 3 and 14 days suggests a vital role in the post-ICH.

Dynamic changes in immune cells following ICH. A UMAP visualization of immune cell heterogeneity in the brain following ICH at sham, 3 days, and 14 days post-injury. B Heatmap illustrates the top marker genes for each cluster. C Expression levels of selected known marker genes across unsorted cells, illustrated in UMAP plots, in Sham and ICH mice brains. D Composition of immune cell clusters in the brain following ICH. E Flow cytometry analysis of immune cell infiltration in the ICH brain at 3 days, 14 days post-ICH, and sham brains, *p < 0.05, ***p < 0.001; ns, not significant, one-way ANOVA and Bonferroni or Kruskal-Wallis and Dunn’s, n = 3–6. Data are mean ± SD

Full size image

Identification of THBS1 + macrophages in ICH

We next explored the different subtypes and functions of macrophages after ICH. Unsupervised clustering of macrophages separates these cells into four subtypes (Fig. 2A). We examined the transcriptional profiles of these distinct macrophage identities across molecular subtypes. Based on previous studies [38, 39] and the top significantly differentially expressed (DE) genes identified in our analysis, we revealed a spectrum of phenotypes, ranging from chemotaxis-related monocyte-derived macrophages (MDM1-CCL4) and inflammation-associated (MDM2-SPP1) to border-associated macrophage (BAM-MRC1) and pro-repair-associated (MDM3-THBS1) phenotypes (Fig. 2B, C, Figure S2, 4). To further determine the function of THBS1 + macrophages, we analyzed the DE gene of THBS1 + macrophages. We showed that THBS1 + macrophages were significantly enriched in repair and myelination-related pathways, suggesting that THBS1 + macrophages may play an essential role in post-ICH repair (Fig. 2D-H, Figure S3). Immunofluorescence demonstrated THBS1 + macrophages infiltrating mouse brain tissue after ICH (Fig. 2I).

THBS1 + macrophages expressing pro-repair phenotype aggregate in the periphery of the hematoma following ICH. A UMAP representation of macrophage sub-clusters. B Manhattan plot of gene expression log2 fold-change values in macrophage sub-clusters. C Heatmap of selected genes representing standardized gene expression values (z-scores) in macrophage sub-clusters. Genes are classified by known functions, as indicated by colored circles below the heatmap for functional classification. D, E GO BP (D) and GO CC (E) enrichment analysis. Results were generated by gene set enrichment analyses of the significant DE genes for THBS1 + macrophage (log2FC > 1, adjusted < 0.05). F-H GSEA of THBS1 + macrophages reveals enrichment of myelin sheath, distal axon, and wound healing pathways. I Representative immunofluorescence images showing THBS1⁺ macrophages in the peri-hematomal region at 3 days after ICH, n = 3. J, K Flow cytometry quantification of CD11b + F4/80 + THBS1 + macrophages in ipsilateral brain tissue 3, 7, 14 days after ICH versus sham with mean ± SD, *p < 0.05, **p < 0.01, ***p < 0.001, Welch’s ANOVA and Dunnett’s or Kruskal-Wallis and Dunn’s, n = 6–9. Data are mean ± SD

Full size image

THBS1 + macrophages significantly correlate with enhanced long-term functional recovery and maintained white matter (WM) Integrity in ICH. Integrity

Based on the analysis that suggested a potential role of THBS1 + macrophages in myelin sheath, we designed experiments (Fig. 3A) to further investigate their functional correlation in a mouse ICH model. Following behavioral testing, brain tissue was collected for flow cytometry analysis to evaluate temporal changes in the CD11b + F4/80 + THBS1 + macrophage (Figure S4) population at multiple time points (0, 3, 7, and 14 days) post-ICH. The 7-day time point post-ICH was selected for DTI to assess white matter integrity, as this period represents a critical transition from acute inflammation to chronic repair. The population of infiltrating THBS1 + macrophages showed a significant increase starting from 3 days onward, continuing to escalate until at least 14 days post-ICH (Fig. 2J, K), highlighting the importance of this macrophage population in ICH recovery.

Following the induction of ICH, mice exhibited significant behavioral impairments in neurological assessments: the adhesive removal test (Fig. 3B, C) and the rotarod test (Fig. 3D). Pearson correlation analysis demonstrated significant correlations between the THBS1 + macrophage ratio and improvements in neurological function (Fig. 3E-G). Animals were anesthetized at 7 days post-ICH and subjected to DTI scanning. WM integrity was assessed using FA maps, with RD as an indicator of myelin damage (Fig. 3H). At 7 days post-stroke, FA values were significantly reduced, while RD values increased in both the EC and IC (Fig. 3I). Pearson correlation analysis revealed that the THBS1 + macrophage/total macrophage ratio was significantly correlated with FA and RD values in the EC region (Fig. 3J, K). While no significant correlation was observed between the THBS1 + macrophage/total macrophage ratio and FA values in the IC, RD values in this region demonstrated a significant correlation (Fig. 3L, M). These findings collectively highlight the association between the THBS1 + macrophage/total macrophage ratio and both neurological recovery and white matter integrity following ICH.

THBS1 + macrophages correlate with improved neurological outcomes and enhanced white matter function. A Experimental design to investigate the function of THBS1 + macrophages. B-D Assessment of neurological function following ICH by adhesive removal (B, C) and rotarod (D) tests, *p < 0.05, **p < 0.01, and ***p < 0.001, two-way repeated-measures ANOVA and Bonferroni, n = 6–7. E-G Correlation analysis revealed a significant positive association between the THBS1 + macrophage ratio and adhesive removal test results and rotarod test performance following ICH (n = 7). H Representative axial views of T2W and DTI FA maps acquired 7 days post-ICH. I Quantification of FA and RD in the EC and IC at three levels (n = 4), *p < 0.05, **p < 0.01, ***p < 0.001; Student’s t-test, Welch’s t-test, or Mann-Whitney test. J-M Pearson correlation analysis revealed a significant positive association between the THBS1 + macrophage ratio and the quantification of FA and RD in the EC and IC following ICH (n = 4). Data are mean ± SD

Full size image

Bone marrow response following ICH serves as the critical source of THBS1 + cells

Subsequently, we conducted pseudo-time analysis on all macrophage populations, revealing that THBS1 + macrophages were predominantly localized at the initial state of the trajectory path. In contrast, BAM and CCL4 + macrophages were distributed at the terminal state (Fig. 4A). To investigate the origin of THBS1 + macrophages, we collected the skull, spleen, and femur from mice at 7 days following ICH and employed flow cytometry to quantify the proportion of THBS1 + cells within the myeloid cell population (Fig. 4B). Flow cytometry analysis revealed a significant increase in THBS1 + myeloid cells, primarily in the bone marrow, following ICH (Fig. 4C, E, and F). Additionally, a lateralized distribution of this increase was observed in the skull (Fig. 4D), a finding consistent with prior reports [40]. The immunofluorescence findings from the skull and femur bone marrow were consistent with the cell count results obtained from flow cytometry (Fig. 4H-I). These results suggest that, following ICH, the bone marrow undergoes persistent changes and serves as an immune reservoir, supplying THBS1 + MDM.

Skull bone marrow remodeling following ICH is a major source of THBS1 + cells. A Pseudotime analysis of four macrophage subclusters utilizing Monocle. The trajectories of the clusters are represented by lines (above), while the corresponding changes in pseudotime are indicated by color gradients (below). B-F Flow cytometry analysis of THBS1 + myeloid cells in the skull (C, D), spleen(E), and femur(F) following ICH. **P < 0.01, ***P < 0.001, ns representing non-significant, Student t-test, Paired t-test, or Welch’s t-test, n = 5–7. G Representative images depicting THBS1 + myeloid cells in the skull and femur bone marrow at 7 days following ICH.H, I Quantification of THBS1 + CD11b + cells in the skull (H) and femur (I) bone marrow; **p < 0.01, ***p < 0.001, Welch’s t-test or Student’s t-test, n = 6. Data are mean ± SD

Full size image

CGRP promotes the phenotypic transition of macrophages toward a THBS1 + pro-repair phenotype

Extensive research has demonstrated that CGRP, derived from nociceptors, is pivotal in modulating macrophage differentiation into the THBS1 + phenotype through intricate neuroimmune regulatory mechanisms and promoting tissue repair processes [17]. This neuroimmunomodulatory phenomenon exhibits remarkable tissue-specific distribution, as evidenced by its regulatory effects on bone marrow immune cells [14] within the femoral compartment and meningeal macrophages [15] residing in the meningeal layers. Transcriptomic analysis revealed a significant overlap between THBS1 + macrophages and CGRP-induced macrophages at the transcriptional level, strongly suggesting that CGRP is a key inducer of this specific macrophage phenotype (Fig. 5A). AND the expression of anti-inflammatory/pro-repair genes(Fig. 5B), such as Thbs1 [17], Arg1 [41], Clu [42] and Stat3 [43], as well as remyelination-related genes(Fig. 5C), including Tgm2 [44], Vegfa [45], Tgfb3 [46], and Osm [47], was elevated in CGRP-stimulated macrophages compared to those treated with PBS.

CGRP promotes the phenotypic transition of macrophages toward a THBS1 + pro-repair phenotype. A Violin plot depicting macrophage subset transcriptional enrichment for gene signatures identified in RNA-seq analyses from GSE255049. B, C The expression of anti-inflammatory/pro-repair (B) and remyelination-related(C) genes was increased in CGRP-stimulated macrophages versus PBS, *adjusted p < 0.05, moderated t-test and Benjamini-Hochberg false discovery rate, n = 3. D Whole-mount confocal imaging of CGRP + nociceptor innervation in the dura mater in sham (above) and ICH (below) mice. E llustration of skull and femur CGRP assays from sham or ICH mice. F CGRP levels released from the skull and femur were measured by ELISA, **p < 0.01, ***p < 0.001, one-way ANOVA and Bonferroni, n = 5. G Quantification of THBS1 + myeloid cells(CD45 + CD11b+)ratio in CGRP-treated and PBS-treated mice skull (left) and femur bone(right) marrow at 5 days post-ICH, **p < 0.01, Student’s t test or Welch’s t test, n = 6. H Quantification of THBS1 + macrophage ratio in CGRP-treated and PBS-treated mice ipsilateral brain at 5 days post-ICH, **p < 0.01, Student’s t test, n = 6. Data are mean ± SD

Full size image

Given the pronounced and lateralized alterations observed in immune cell populations within the skull (Fig. 4C, D) and anatomical proximity, we hypothesized that CGRP may predominantly originate from the meningeal tissues adjacent to the cranial bone. Immunofluorescence analysis substantiated this hypothesis, demonstrating a significant enhancement in CGRP expression localized within the meningeal nociceptive nerve fibers following ICH (Fig. 5D, Figure S6). Ex vivo culture analyses of skull and femur tissues harvested from ICH models revealed significantly elevated concentrations of CGRP compared to sham (Fig. 5E, F).

To investigate whether CGRP enhances the population of endogenous THBS1 + macrophages following ICH, we administered CGRP (1ug/g) or PBS via intraperitoneal injection in mice after inducing ICH (Fig. 6A) at 3 h post-ICH, followed by repeated administrations at 3, 7and 10 days post-ICH. Brain and bone marrow tissues were collected at 5 days and analyzed using flow cytometry. The results demonstrated that CGRP treatment significantly increased THBS1 + myeloid cells in the bone marrow (Fig. 5G) and elevated infiltration of THBS1 + macrophages (Fig. 5H) in the brain. This result suggests a release of this neuropeptide to regulate bone marrow immunity following ICH.

Enhancing THBS1 + macrophage populations post-treatment with CGRP promotes white matter repair following ICH. A Experimental design for evaluating the effects of CGRP treatment in a mouse model of ICH. B-D Adhesive removal tests (B, C) and rotarod performance (D) indicate that mice treated with CGRP exhibited better long-term recovery of sensory and motor functions compared to PBS following ICH (n = 8),*P < 0.05, two-way repeated-measures ANOVA and Bonferroni. E Representative axial views of DTI fractional FA maps and a single plane of the DEC map are presented for the same brains collected 14 days after ICH. F, G Quantitative analysis of FA and RD was conducted in the EC (F) and IC(G) at three distinct anatomical levels (n = 5), *p < 0.05, Student’s t-test, Welch’s t-test, or Mann-Whitney test. H MBP staining in brain Sect. 14 days post-ICH (n = 4–6). White dashed lines indicate the region of the ipsilateral EC. I Representative TEM images at 14 d post-ICH with red squares highlighting the areas that have been enlarged. Red arrows, defects in the myelin sheath. Blue arrowheads, remyelinated axons. J The g-ratios of myelinated axons were quantified to axon diameters. The counts of axons from three mice are indicated.***p < 0.001, Mann-Whitney test. K Scatterplot illustrating the individual g ratio values and the distribution of axonal sizes. Data are mean ± SD

Full size image

CGRP-induced THBS1 + pro-repair macrophages promote remyelination after ICH

Previous studies have demonstrated that macrophages participate in tissue repair at day 7 following ICH [13]. To investigate the impact of CGRP-induced macrophages on post-ICH recovery, we designed an experiment to compare the long-term outcomes of mice receiving BMDMs stimulated with either PBS or CGRP. (Fig. 7A). BMDMs were treated with CGRP (1 nM) or PBS for 4 h, dissociated into single cells, and then intravenously injected (3 × 10⁶ per mouse) into C57BL/6J mice on day 7 post-ICH to evaluate their role in repair. Compared to PBS-stimulated macrophages, CGRP-stimulated macrophages infused exhibit superior long-term sensorimotor recovery post-ICH, as assessed by rotarod (Fig. 7B) and adhesive removal (Fig. 7C,D) tests. Then, we assessed WM integrity by measuring the expression of MBP, a major structural component of the myelin sheath, using WB analysis in the basal ganglia. MBP levels were markedly decreased at both day 3 and day 7 following ICH, indicating progressive demyelination and white matter disruption in the perihematomal region (Fig. 7E). ICH mice receiving CGRP-stimulated BMDMs exhibited higher levels of MBP expression compared to those receiving PBS-treated BMDMs, indicating enhanced white matter integrity (Fig. 7F). To investigate whether CGRP-induced macrophages promote remyelination, we assessed the extent of myelination by evaluating the colocalization of MBP with axons (neurofilament, NF). Adoptive transfer of CGRP-stimulated BMDMs promoted enhanced remyelination, as evidenced by three-dimensional reconstruction of brain sections from the peri-lesional region of the basal ganglia (Fig. 7G, H). Collectively, these findings suggest that CGRP-induced BMDMs facilitate both neurological recovery and remyelination in ICH.

CGRP-induced repair-associated macrophages promote remyelination after ICH. A Experimental Design for the Adoptive Transfer of PBS- or CGRP-Induced Macrophages into ICH Mice. B-D Rotarod performance (B) and adhesive removal tests (C, D) showed that mice receiving CGRP-induced macrophages exhibited significantly improved long-term sensory and motor recovery compared to those receiving PBS-induced macrophages following ICH, *p < 0.05, Student’s t test, n = 6. E Western blot and heatmap show the protein expression of MBP in the hematoma region, ***p < 0.001, one-way ANOVA and Bonferroni, n = 5. F Western blot and heatmap show the protein expression of MBP in the adoptive transfer of PBS- or CGRP-induced Macrophages into 14 days post-ICH Mice, ***p < 0.001, Student’s t test, n = 6. G Representative immunofluorescence images of peri-hematomal region stained for NF and MBP. H Fold change MBP + NF + axons of total NF + axons, **p < 0.01, Student’s t test, n = 4–6. Data are mean±SD

Full size image

Enhancing the population of endogenous THBS1 + macrophages following ICH significantly rescues long-term stroke outcomes

To assess the long-term influence of THBS + 1 macrophage boosting, mice were subjected to ICH and treated intraperitoneally with CGRP (1ug/g) or PBS starting at 3 h post-ICH, followed by repeated administrations at 3, 7and 10 days post-ICH (Fig. 6A). Sensorimotor functions, evaluated through the adhesive removal test (Fig. 6B, C) and rotarod test (Fig. 6D), showed marked improvement in CGRP-treated mice compared to PBS-treated mice, with the benefits emerging around 7 days following ICH. Animals were anesthetized 7 days after ICH and underwent DTI scanning (Figs. 6E-G). The reduction in FA values, indicative of WM injury in the EC and IC of PBS-treated mice, was significantly attenuated in CGRP-treated mice (Fig. 6F). Furthermore, CGRP treatment reduced RD values in the EC and IC (Fig. 6G), suggesting enhanced myelin integrity. Immunofluorescence staining for myelin basic protein (MBP), a significant myelin protein, was performed 14 days after ICH to evaluate WM lesions further. CGRP-treated mice exhibited significantly higher MBP levels in the EC and peri-hematomal areas than PBS-treated controls (Fig. 6H, Figure S7), providing additional evidence of improved WM integrity following the enhancement of THBS1 + macrophages. To evaluate the ultrastructural integrity of the myelin sheath in the basal ganglia in in the peri-lesional core of the basal ganglia, TEM was performed 14 days after ICH (Fig. 6I). In PBS-treated mice, extensive myelin sheath loss and fragmentation were observed in many axons (Fig. 6I, red arrows). In contrast, CGRP-treated mice exhibited a higher proportion of myelinated axons with well-defined, regenerated myelin structures (Fig. 6I, blue arrows). Quantitative analysis revealed that CGRP treatment significantly increased myelin sheath thickness, as evidenced by a reduction in the g-ratio (the ratio of axon diameter to the total outer diameter of the myelinated fiber) in axons of ICH-affected mice (Fig. 6J and K). Collectively, these findings suggest the therapeutic effects of CGRP through the augmentation of THBS1 + cells in promoting WM repair and rescuing long-term neurological function after ICH.

The complex interplay between neuroinflammation and tissue repair following ICH remains a critical area of investigation. This study elucidates a novel neuroimmune axis mediated by CGRP that drives the expansion of THBS1 + macrophages, a pro-repair macrophage subset that contributes to WM integrity and functional recovery (Fig. 8). After brain hemorrhage, monocytes infiltrate the brain during ICH in a CCR2 (C-C motif chemokine receptor 2)-dependent manner and differentiate into macrophages [13, 48]. Our investigation revealed a sustained infiltration of monocyte-derived macrophages in the post-intracerebral hemorrhage period, demonstrating a distinct temporal pattern compared to other immune cell populations. While early studies highlighted their detrimental role in acute brain injury ([49, 50]), more recent work has revealed their essential contributions to hematoma resolution and long-term functional recovery ([13, 51, 52]). Studies demonstrate that monocyte depletion initiated at 3 days post-ICH worsens neurological deficits, revealing a critical shift in monocyte-derived macrophages toward a protective phenotype. This transition coincides with their temporally regulated modulation of pro-inflammatory, anti-inflammatory, and hematoma clearance pathways [13]. This indicates that the conventional M1/M2 classification system may not fully capture macrophages’ complexity and functional heterogeneity. Through single-cell phenotypic classification of macrophages, researchers discovered that in addition to the subsets involved in immune attraction and inflammation, a subgroup of THBS1 + macrophages emerges during the chronic phase following brain hemorrhage for a long time. These macrophages, associated with tissue repair, consistently exhibit upregulation of markers such as Arg1 and Tgm2 during the late post-hemorrhagic phase. This indicates that macrophages exhibit multifunctionality in immune defense and tissue repair after brain hemorrhage.

Neuro-immune axis mediated by CGRP enhances macrophage-driven white matter repair in ICH

Full size image

THBS1 + macrophages have been previously reported to play a role in skin wound repair [17]. Our findings demonstrate that these cells are closely associated with brain hemorrhage prognosis and white matter function. Macrophages have been reported to play a crucial role in central nervous system (CNS) remyelination by clearing myelin debris and secreting growth factors and cholesterol that promote oligodendrocyte differentiation [53,54,55,56,57]. We found that THBS1 + cells express genes related to anti-inflammatory/pro-repair and myelination, which may be related to macrophages’ involvement in myelin repair after ICH. This discovery provides new insights into the role of macrophages in ICH.

Regarding the origin of macrophages, monocyte-driven macrophages are typically believed to originate from the spleen or bone marrow [58, 59]. However, the source of THBS1 + macrophages remains unclear after brain hemorrhage. Flow cytometry experiments have shown that after ICH, Myeloid cells in the bone marrow undergo THBS1 + immunophenotypic switching rather than in the spleen. This implies that bone marrow is an essential source of repair-related macrophages. Recent studies found that bone marrow-derived immune cells play a significant role in immune cell infiltration during CNS diseases, particularly in bone marrow alterations and side-specific responses observed in the skull bone [21, 40, 60,61,62]. Furthermore, CNS-infiltrating monocytes rely more on cranial bone marrow than neutrophils [61, 62], underscoring the critical role of bone marrow-derived immune cells in CNS immune recruitment. This dependence highlights the potential of targeting bone marrow monocytes as a therapeutic strategy following CNS injury.

Beyond its well-established role as a neuropeptide in migraine pathogenesis [63], CGRP has recently been identified as a key modulator in neuroimmune regulation [17, 64,65,66,67]. Our findings revealed a significant elevation of CGRP expression in the bone marrow following ICH. Notably, CGRP-induced repair-associated macrophages exhibited transcriptomic profiles closely resembling the THBS1 + macrophages identified in our study, expressing genes associated with anti-inflammatory/pro-repair and remyelination. These observations suggest that THBS1 + macrophages may originate from bone marrow-derived myeloid cells modulated by CGRP signaling. Previous studies have shown that promoting macrophage phenotypic switching improves brain hemorrhage outcomes. Further intraperitoneal administration of CGRP enhanced neurological recovery and preserved white matter integrity following ICH, with evidence suggesting a significant contribution from CGRP-induced THBS1 + macrophages. These findings align with prior studies on brain injury [68], underscoring the pivotal role of macrophages as a mechanistic link between CGRP signaling and neural repair processes. In addition, our results also demonstrate that early administration of CGRP modulates macrophages, thereby protecting neurological function in the early stages of intracerebral hemorrhage in mice. This dual role—initially protective and later pro-remyelinating—is consistent with previous studies [69, 70], which highlight inflammation resolution as a critical component of white matter repair processes, underscoring the therapeutic potential of CGRP. Given that THBS1 + macrophages primarily originate from the bone marrow and undergo conversion within the bone marrow, targeting CGRP in the bone marrow may offer a precise therapeutic approach. However, this hypothesis remains to be validated with further experimental and clinical applications.

In summary, our study highlights the pivotal role of CGRP in orchestrating a neuroimmune axis that promotes the expansion of THBS1 + macrophages, thereby rescuing white matter integrity and functional recovery following ICH. These macrophages, primarily derived from bone marrow, exhibit a pro-repair phenotype characterized by the expression of anti-inflammatory markers and genes associated with myelination. The modulation of CGRP signaling pathways offers a promising therapeutic avenue to enhance macrophage-mediated tissue repair processes in the context of ICH.

Limitations

Our study also has some limitations: sex plays a crucial role in brain hemorrhage prognosis, and future studies should investigate the role of macrophages in inflammatory tissue repair in female mice. Unclear mechanisms of THBS1 + macrophages in white matter repair after brain hemorrhage: Whether THBS1 + macrophages directly regulate oligodendrocyte development or promote myelination through other immune cells (e.g., microglia) needs to be further explored.

The datasets used and analyzed during this current study are available from the original article.

ICH:

Intracerebral Hemorrhage

CD45:

Cluster of Differentiation 45

Ly6G:

Lymphocyte Antigen 6 Complex, Locus G

CD3e:

Cluster of Differentiation 3 epsilon

CD11b:

Cluster of Differentiation 11b

F4/80:

EGF like Module-containing mucin-like Hormone Receptor-like 1

THBS1:

Thrombospondin 1

MBP:

Myelin Basic Protein

CGRP:

Calcitonin Gene-Related Peptide

BAM:

Border Associated Macrophage

MDM:

Monocyte-Derived Macrophage

EC:

External Capsule

IC:

Internal Capsule

PBS:

Phosphate-Buffered Saline

DTI:

Diffusion Tensor Imaging

TEM:

Transmission Electron Microscopy

PFA:

Paraformaldehyde

DAPI:

4’ 6-Diamidino-2-Phenylindole

DMEM:

Dulbecco’s Modified Eagle Medium

UMAP:

Uniform Manifold Approximation and Projection

FA:

Fractional Anisotropy

RD:

Radial Diffusivity

EDTA:

Ethylenediaminetetraacetic Acid

DEC:

Directionally Encoded Color

CNS:

Central Nervous System

BMDM:

Bone Marrow Derived Macrophages

This research was supported by the National Science Foundation of China (82001299,82201431, 82201617,82401724).

1. Huaping Huang and Yirui Kuang contributed equally to this work.

Authors and Affiliations

1. Department of Neurosurgery, The Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, China

   Huaping Huang, Yirui Kuang, Yi Zhang, Jiayin Zhou, Xian Yu, Yonghe Zheng, Lingxin Cai, Liansheng Gao, Haijian WU, Hui Ling, Xiao Dong, Hang Zhou, Xiaobo Yu, Yucong Peng, Gao Chen, Xiaoyu Wang & Wei Yan

2. Key Laboratory of Precise Treatment and Clinical Translational Research of Neurological Diseases, Hangzhou, Zhejiang, China

   Huaping Huang, Yirui Kuang, Yi Zhang, Jiayin Zhou, Xian Yu, Yonghe Zheng, Lingxin Cai, Liansheng Gao, Haijian WU, Hui Ling, Xiao Dong, Hang Zhou, Xiaobo Yu, Yucong Peng, Gao Chen, Xiaoyu Wang & Wei Yan

3. State Key Laboratory of Transvascular Implantation Devices, Hangzhou, China

   Huaping Huang, Yirui Kuang, Yi Zhang, Jiayin Zhou, Xian Yu, Yonghe Zheng, Lingxin Cai, Liansheng Gao, Haijian WU, Hui Ling, Xiao Dong, Hang Zhou, Xiaobo Yu, Yucong Peng, Gao Chen, Xiaoyu Wang & Wei Yan

4. Department of Neurosurgery, Taizhou Hospital of Zhejiang Province Affiliated to Wenzhou Medical University, Taizhou, Zhejiang, China

   Yang Chen

5. Department of Neurosurgery, Ningbo First Hospital, Ningbo, Zhejiang, China

   Wanglu Hu

Contributions

H.H., X.W., and W.Y. conceived and designed the study. H.H. and Y.Z. Underwent a DTI MRI scan. H.H., Y.K., and Y.C. perform data analysis and write manuscripts. H.H., X.Y., and Y.Z. established ICH models, intraperitoneal injection, and flow cytometry testing. H.H. and L.C. performed a neurological function assessment. J.Z. and H.H. were stained with immunofluorescence. H.H. and X.Y. checked the statistics. L.G., X.Y., H.Z., and Y.P. provide funding. All authors reviewed the manuscript.

Corresponding authors

Correspondence to Xiaoyu Wang or Wei Yan.

Ethics approval and consent to participate

All animal procedures were authorized by the Animal Ethics Committee of the Second Affiliated Hospital of Zhejiang University (Approval No. 2020 − 360) and conducted by the Guiding Principles for the Care and Use of Laboratory Animals, as endorsed by the National Science and Technology Committee of China.

Consent for publication

All authors have read and approved the publication of this manuscript.

Competing interests

The authors declare no competing interests.

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions