Homer1 promotes the conversion of A1 astrocytes to A2 astrocytes and improves the recovery of transgenic mice after intracerebral hemorrhage

All animal experiments were performed in accordance with protocols approved by the Institutional Ethics Committee of Xijing Hospital. All experimental procedures were approved by the Institutional Animal Care and Use Committee of Air Force Military Medical University. C57BL/6 and Homer1^flox/flox/Nestin-Cre^+/− mice were purchased from the Shanghai Model Organisms Center, Inc. (Shanghai, China). All mice were maintained in the same environment.

Primary astrocytes were infected with lentivirus to stably overexpress Homer1 (Homer1-OE) (LV-hU6-Homer1-Ubiquitin-EGFP-IRES-puromycin) or knock-down Homer1 (Homer1-KD) (LV-Ubi-shRNAHomer1-3FLAG-SV40-EGFP-IRES-puromycin). The lentivirus was constructed with the assistance of GeneChem Co., Inc. (Shanghai, China).

The ICH model was established as previously described with minor modifications [25, 26]. The mice were anesthetized with 4% chloral hydrate (400 mg/kg) injected intraperitoneally. The rectal temperature was maintained at 37.5 °C. A stereotactic technique was used to make a scalp incision along the midline and a burr hole was drilled on the left side of the skull (0.2 mm anterior and 2.5 mm lateral to the bregma). Thirty microliters of autologous blood obtained from the femoral artery were transferred into a 50 μL Hamilton syringe. The syringe was connected to a microinjection pump and the needle was inserted into the brain through the burr hole (depth, 3.5 mm from the bone surface). Thirty microliters of autologous blood were injected within 10 min. The syringe was withdrawn after 10 min, and selumetinib (MCE, AZD6244) and Homer1 protein (EUPROTEIN, EP8767430) (5 mg/kg, dissolved in normal saline) were re-injected in situ for 10 min. After surgery, the skull hole was sealed with bone wax and the incision was closed with sutures. To avoid postsurgical dehydration, normal saline (0.5 mL) was subcutaneously injected into each mouse immediately after the surgery. The ICH model mice were killed at different time points for staining and western blot (WB) analyses.

Ten C57BL/6 mice were used as the sham group. Thirty C57BL/6 mice were used to establish the ICH model, and five mice per group were anesthetized on days 1, 3, and 7. The brain tissue around the hemorrhage site in each mouse was extracted and stored at  80 °C for WB, quantitative real-time polymerase chain reaction (qPCR), and enzyme-linked immunosorbent assay (ELISA) analyses. Each experiment was repeated three times (n = 5/group). In addition, five mice in each group were anesthetized and killed to obtain frozen sections. The frozen sections with the largest bleeding area in each mouse were selected for immunofluorescence stainings. One field at the same magnification was randomly selected from one section of each mouse for positive cell counting and quantification (n = 5/group). Mice were scored according to the Longa score standard before anesthesia in all groups (n = 10/group).

Three weeks before ICH modeling, AAV was injected into the cerebral hemorrhage site under the same coordinates, and ten Homer1-OE and ten Homer1-KD mice were generated. AAV was stably expressed after 3 weeks of normal feeding. Ten C57BL/6 mice were used in the sham group (Sham group). Ten ordinary C57BL/6 mice (ICH group), 10 Homer1-OE mice (Homer1-OE group), and 10 Homer1-KD mice (Homer1-KD group) were used to establish the ICH model. On the third day after ICH, all mice were anesthetized and killed. The hemorrhagic brain tissue of five mice in each group was used for WB, qPCR, and ELISA assays (n = 5/group). The brain tissue of five mice in each group was used to obtain frozen sections and for immunohistochemistry (n = 5/group). Mice were scored according to the Longa score standard before anesthesia in all groups (n = 10/group). In addition, 20 mice from each group were used for survival analysis (n = 20/group).

Fifty Homer1^flox/flox/Nestin-Cre^+/− mice were divided into five groups: sham, ICH, Homer1 protein, selumetinib, and Homer1 + selumetinib group (Both). ICH models were established in all groups, except the sham group, and different drugs were administered according to the design. On the third day after ICH, all mice were anesthetized and killed. The hemorrhagic brain tissue of five mice in each group was used for WB, qPCR, and ELISA assays (n = 5/group), and mouse cytokine array detection (n = 3/group). The brain tissue of five mice was used to obtain frozen sections and for immunohistochemistry (n = 5/group). The mice were scored according to the Longa score standard before anesthesia in all groups (n = 10/group). In addition, 20 Homer1^flox/flox/Nestin-Cre^+/− mice in each group were used for survival analysis.

Neonatal C57 mice were used to extract primary astrocytes. The brain tissue was minced with sterile ophthalmic scissors, digested with 0.25% trypsin for 5 min at 37 °C before the brain tissue was centrifuged at 1000 rpm for 5 min. The digested tissue was cultured in F-12 medium enriched with 10% fetal bovine serum (FBS), 0.224% NaHCO₂, and 1% penicillin/streptomycin at 37 °C in the presence of 5% CO₂. Astrocyte monolayers were obtained at the bottom of the dish after 2 weeks. The astrocytes were identified by morphological analysis and GFAP staining. Before further treatment, the medium containing 0.5% FBS was replaced to ensure that the cells were in a resting state.

The ICH cell model was established as previously described [27]. Cells were treated with erythrocyte lysate (1 μL of red blood cell lysate per mL of medium) to create an in vitro ICH inflammation model. The cells were incubated for different durations and used in different experiments. Erythrocyte lysates were prepared with red blood cell lysis buffer (Solarbio, Beijing), and the experimental process was in strict accordance with the manufacturer’s instructions.

A1 astrocytes were generated as described by Liddelow et al. [6]. Astrocyte conditional medium (ACM) was prepared in DMEM supplemented with TNF-α (30 ng/mL, CST, 8902), IL-1α (3 ng/mL, Peprotech, 400-01), and C1q (400 ng/mL, Novus protein, NBP2-62410). Conversion of astrocytes from A1 to A2 was examined by protein and mRNA expression of C3 and S100A10 after the cells were cultured in ACM for 24 h.

Primary astrocytes were harvested for RNA extraction after different treatments using TRIzol reagent. Mice were anesthetized at different time points after ICH induction, and the brain tissue around the bleeding site was used for qPCR. Reverse transcription was performed according to the protocol of the HiScript II Q Select RT SuperMix for qPCR (+ gDNA wiper) kit (Vazyme, China). qPCR was performed according to the protocol of the ChamQ SYBR Color qPCR Master Mix (Low ROX Premixed) kit (Vazyme, China). The primers for mRNA were as follows: Homer1-F: “AATGGTTAGGGGGCACTGTTT,” Homer1-R: “CCCATCTGCCACAGTCACAA”; C3-F: “AACAAGCTCTGCCGTGATGA,” C3-R: “GCCTGACTTGATGGTCTGCT”; S100A10-F: “TGAGAGTGCTCATGGAACGG,” S100A10-R: “AGAAAGCTCTGGAAGCCCAC”; GAPDH-F: “AGAGGCCCTATCCCAACTCG,” and GAPDH-R: “GTGGGTGCAGCGAACTTTATT”. The expression of related RNAs was calculated using the 2^−ΔΔCt method, and GAPDH was used as a control. The experiment was repeated thrice.

Western blotting was performed as previously described [25]. Primary astrocytes were harvested for protein extraction after the different treatments. Mice were anesthetized at different time points after ICH induction, and the brain tissue around the bleeding site was used for WB. The following antibodies were used: rabbit anti-Homer1 (1:1000; Abcam), rabbit anti-C3 (1:1000; Abcam), rabbit anti-S100A10 (1:1000; Proteintech), mouse anti-GAPDH (1:10,000; Abcam), rabbit anti-Ras (1:1000; CST), rabbit anti-Phospho-c-Raf (Ser338) (1:1000; CST), mouse anti-Raf-1 (1:1000; Proteintech), rabbit anti-Phospho-MEK1/2 (Ser217/221) (1:1000; CST), mouse anti-MEK1/2 (1:1000; CST), rabbit anti-Phospho-ERK1/2 (Thr202/Tyr204) (1:1000; Proteintech), rabbit anti-ERK1/2 (1:1000; Proteintech), and goat anti-mouse/rabbit secondary antibody (1:10,000; Abcam).

Frozen sections with the largest bleeding areas were selected for Nissl staining. The sections were washed with PBS three times for 5 min each time and stained with 1% toluidine blue (Beyotime, Shanghai) at 56 °C for 20 min. The slices were washed with PBS three times for 5 min each. The slices were soaked in 75% alcohol for 1 min, differentiated in 95% alcohol, and the degree of differentiation was observed under a microscope. The sections were then dehydrated in absolute ethanol and soaked in xylene for 5 min. After sealing with neutral resin, the samples were observed under an optical microscope.

Frozen sections with the largest areas of bleeding were selected for TUNEL staining. Experimental procedures for TUNEL staining were performed in strict accordance with the manufacturer’s instructions (Beyotime, Shanghai, China). Stained sections were observed under a fluorescence microscope. The excitation and emission wavelengths were 550 nm and 570 nm, respectively.

Immunohistochemical analysis was performed as previously described [25]. Mice were anesthetized at different time points after ICH induction, and frozen sections with the largest bleeding area were selected for immunofluorescence analysis. The following antibodies were used: mouse anti-GFAP (1:300; CST), mouse anti-Iba1 (1:250; GeneTex), mouse anti-NeuN (1:300; CST), rabbit anti-Homer1 (1:200; Abcam), rabbit anti-C3 (1:200; Abcam), rabbit anti-S100A10 (1:300; Proteintech), donkey anti-rabbit IgG (H + L) highly cross-adsorbed secondary antibody, Alexa Fluor Plus 488 (1:1000; Invitrogen), and donkey anti-mouse IgG (H + L) highly cross-adsorbed secondary antibody, Alexa Fluor Plus 555 (1:1000; Invitrogen).

Cell supernatants from the different treatment groups were harvested for ELISA. Mice were anesthetized at different time points after ICH induction, and the brain tissue around the bleeding site was used for ELISA. ELISA was performed in strict accordance with the manufacturer’s instructions. The following ELISA kits were used for detection: Mouse IL-1 beta ELISA Kit (Abcam, UK), Mouse TNF alpha ELISA Kit (Abcam, UK), Mouse Tweak ELISA Kit (Abcam, UK), Mouse Activin A ELISA Kit (Abcam, UK), Mouse TRAIL ELISA Kit (Abcam, UK), and Mouse Persephin (Pspn) ELISA Kit (4A Biotech, China).

The brain tissue around the bleeding site was used for the detection of inflammatory factors and cytokines. The mouse cytokine array Q4000 was purchased from RayBiotech (QAM-CAA-400; USA). The experimental procedures were performed in strict accordance with the manufacturer’s instructions.

Mouse tails were cut, digested with proteinase K for 20 min at 55 °C, and further inactivated with proteinase K for 5 min at 100 °C. Polymerase chain reaction (PCR) was performed according to the protocol of the One Step Mouse Genotyping Kit (Vazyme, China). Identification of Homer1 Flox (wild type: 232 bp; mutant: 300 bp): Primer 1 (Forward): 5′-TGAGCTGGACACCCCCTGCC-3′, Primer 2 (Reverse): 5′- TGTTAAAACAATTACACCCGATTCTT -3′; identification of Nestin-Cre (wild type: 246 bp; mutant: 150 bp): Primer 1 (wild type): 5-TTGCTAAAGCGCTACATAGGA-3′, Primer 2 (mutant): 5-CCTTCCTGAAGCAGTAGAGCA-3′, Primer 3 (common): 5-GCCTTATTGTGGAAGGACTG-3′.

Mice were scored behaviorally at different time points after ICH induction. The scores were calculated using the Longa method. No neurological deficit: 0 points; inability to fully extend the forepaws on the paralyzed side: 1 point; circling to the paralyzed side during walking: 2 points; dumping to the paralyzed side during walking: 3 points; inability to walk automatically, with loss of consciousness: 4 points.

SPSS software (version 17.0) was used for the statistical analyses. All values for each group are presented as mean ± SD. Parametric and nonparametric tests were used according to the homogeneity of variance. According to different comparison situations, statistical differences were analyzed using Student’s t-test or one-way ANOVA, as appropriate, with Sidak’s or Turkey’s multiple comparisons test. P < 0.05 indicated that the difference was statistically significant.

First, according to a previously described method, autologous blood from the mouse femoral artery was extracted for intracranial injection to induce ICH [26] (Fig. 1A). The expression of Homer1 protein increased after ICH, peaked on the third day after the induction, and returned to the level of the sham group on the seventh day (Fig. 1B, C). The change in the trend of the Homer1 transcription level was consistent with that of the protein level (Fig. 1D). After staining the brain tissue at the peak of Homer1 expression on the third day after ICH induction, we found that Homer1 was mainly located in astrocytes at the site of ICH (Fig. 1E). To further verify the expression of Homer1 in astrocytes, we extracted primary astrocytes from neonatal mice and treated them with erythrocyte lysate (1 μL of red blood cell lysate per mL of medium) to create an in vitro ICH inflammation model. After erythrocyte lysate treatment, the protein (Fig. 1F, G) and mRNA (Fig. 1H) levels of Homer1 initially increased and then decreased, and the expression level reached its peak 48 h after ICH induction.

To detect changes in astrocytes after ICH, we detected the marker C3 in A1 phenotype astrocytes and S100A10 in A2 phenotype astrocytes. WB analysis showed that the expression of C3 gradually increased on the first and third day after ICH, and the expression of S100A10 gradually increased on the third and seventh days after ICH (Fig. 2A). The mRNA levels of C3 (Fig. 2B) and S100A10 (Fig. 2C) were consistent with the protein levels. A1 phenotype astrocytes also regulate the expression of classical inflammatory factors IL-1β and TNF-α. ELISA revealed that the expression of IL-1β and TNF-α in hemorrhagic brain tissue peaked on the third day, which was consistent with the expression trend of C3 (Fig. 2D, E). Meanwhile, immunohistochemical analysis of frozen tissue sections showed that the number of C3⁺GFAP⁺ cells was the largest on the third day after ICH induction, and the number of S100A10⁺GFAP⁺ cells increased on the third and seventh days after induction (Fig. 2F, G) (C3⁺GFAP⁺ cells: Sham vs. 1d vs. 3d vs. 7d: 18.40 ± 2.88 vs. 40.20 ± 2.864 vs. 61.40 ± 3.507 vs. 22.80 ± 2.864, respectively) (S100A10⁺GFAP⁺ cells: Sham vs. 1d vs. 3d vs. 7d: 13.60 ± 2.510 vs. 20.60 ± 3.286 vs. 51.00 ± 5.612 vs. 64.40 ± 3.715, respectively). In addition, the Longa score of ICH mice showed that hemiplegia was more serious on the third day (Fig. 2H) (Sham vs. 1d vs. 3d vs. 7d: 0.000 ± 0.000 vs. 1.400 ± 0.516 vs. 2.500 ± 0.527 vs. 1.500 ± 0.710, respectively).

In the primary astrocyte ICH model, WB (Fig. 3A) and qPCR (Fig. 3B, C) results showed that the expression of C3 began to increase after 24 h of erythrocyte lysate stimulation and reached its peak at 48 h, while the expression of S100A10 began to increase after 48 h and remained high until 72 h. TNF-α (Fig. 3D) and IL-1β (Fig. 3E) in the cell supernatant also began to increase after erythrocyte lysate stimulation and were most highly expressed 48 h after stimulation. Together with the in vivo results, these data illustrate that mice had the most severe inflammatory response on the third day after ICH, and the phenotype of astrocytes changed from A1 to A2.

To investigate the effect of Homer1 on the phenotypic conversion of astrocytes, primary astrocytes were transduced with Homer1-OE and Homer1-KD using lentiviral transduction. In addition, we also cultured astrocytes in ACM in vitro to simulate the microenvironment dominated by astrocytes with the A1 phenotype on the third day after ICH in vivo. First, the results of the cell activity test showed that Homer1-OE and Homer1-KD had no effect on cell activity in either common medium or ACM (Fig. 4A). According to previous experimental results, the modified cells were recultured in ACM and cultured for 48 h to allow the cells to reach the most severe state of inflammation. qPCR (Fig. 4B–D) and WB (Fig. 4E, F) results showed that the expression of Homer1, C3, and S100A10 increased after astrocytes were cultured in ACM. In astrocytes cultured with ACM, Homer1-OE significantly decreased the expression of C3 and increased the expression of S100A10 while Homer1-KD increased C3 levels and decreased S100A10 levels. In addition, Homer1-OE decreased the release of TNF-α (Fig. 4G) and IL-1β (Fig. 4H), whereas Homer1-KD exacerbated its expression.

In vivo, we first used AAV to stably overexpress Homer1 or knock-down Homer1 at the ICH site in mice. The mouse ICH model was induced 3 weeks after AAV injection, and AAV regulation efficiency and molecular biological experiments were detected on the third day after ICH induction. Compared with ordinary ICH mice, Homer1-OE mice expressed more S100A10 in the brain at the transcriptional (Fig. 4K–M) and translational levels (Fig. 4I, J), whereas Homer1-KD mice expressed more C3. In addition, the results of immunohistochemistry experiments showed that the proportion of S100A10⁺GFAP⁺ cells in Homer1-OE mice was higher (Fig. 5B and D) (Sham vs. 1d vs. 3d vs. 7d: 14.40 ± 3.51 vs. 46.20 ± 6.34 vs. 65.00 ± 2.45 vs. 21.60 ± 5.46, respectively), while the proportion of C3⁺GFAP⁺ cells in Homer1-KD mice was higher (Fig. 5A and C) (Sham vs. 1d vs. 3d vs. 7d: 16.60 ± 1.67 vs. 50.20 ± 5.17 vs. 21.00 ± 2.45 vs. 62.60 ± 4.93, respectively). ELISA results of bleeding tissues also showed that Homer1-OE effectively reduced the expression of TNF-α (Fig. 5E) and IL-1β (Fig. 5F), whereas Homer1-KD exacerbated the inflammatory response. Finally, behavioral tests and survival analysis showed that Homer1-OE decreased the average Longa score of ICH mice (Fig. 5G) (Sham vs. 1d vs. 3d vs. 7d: 0.00 ± 0.00 vs. 2.60 ± 0.52 vs. 1.50 ± 0.53 vs. 2.80 ± 0.63) and prolonged the survival time of mice (Fig. 5H), while Homer1-KD had the opposite effect. These data illustrate that Homer1 plays a protective role by inducing a shift in astrocyte phenotype from A1 to A2 in ICH.

The MAPK canonical signaling pathways include Ras, Raf-1, MEK1/2, and ERK1/2. While MAPK signaling plays an important role in tumor growth [28], the regulatory mechanisms of MAPK signaling and inflammation have been reported in an increasing number of studies [29, 30]. In addition, MAPK is an important initiator and modulator of astrocyte reactivity [31, 32]. To investigate whether Homer1 regulates inflammatory responses via MAPK signaling after ICH, we examined the expression of MAPK signaling-related proteins in in vivo and in vitro ICH models. WB results showed that after primary astrocytes were cultured in ACM for 48 h, the expression of Ras and the ratios of P-Raf-1^Ser338 /Raf-1, P-MEK1/2/MEK1/2, and P-ERK1/2/ERK1/2 were increased. Homer1-OE inhibited the increase in protein expression induced by ACM, while Homer1-KD further promoted the increase in protein expression induced by ACM (Fig. 6A, B). Meanwhile, the results of animal experiments showed that MAPK signaling-related protein expression was significantly lower in mice with Homer1-OE than in the normal ICH group, whereas related protein expression was significantly higher in mice with Homer1-KD than in the ICH group (Fig. 6C, D). These data illustrate that Homer1 may play a protective role after ICH by inhibiting MAPK signaling.

The results of previous experiments suggest that Homer1 plays a protective role in ICH by affecting the MAPK signaling pathway. Therefore, we chose Homer1 and selumetinib, a non-competitive ATP MEK1/2 inhibitor, to explore effective treatments for ICH. Astrocytes were cultured in normal DMEM and ACM and the medium was supplemented with 1 μM Homer1 protein or selumetinib, and cell viability was measured after 48 h of culture. The results showed that Homer1 protein and selumetinib did not affect the activity of primary astrocytes (Fig. 7A). To better reflect the role of Homer1 in ICH, conditional knockout mice were generated. Genotypes (Homer1^flox/flox/Nestin-Cre^+/−) were identified by PCR (Fig. 7B). Mice were used to create ICH models and after in situ injection of Homer1 protein (5 mg/kg) and selumetinib (5 mg/kg), mouse brain tissues were collected for molecular biological and pathological examination on the third day after ICH. Compared to the ICH model of normal mice (Fig. 2B), the expression of C3 in Homer1^flox/flox/Nestin-Cre^+/− mice was higher on the third day after ICH. In addition, compared with the ICH group, the expression of C3 in the brain tissue decreased significantly, and the expression of S100A10 increased after injection of Homer1 protein or selumetinib. The effect of the combined administration of Homer1 protein and selumetinib was more obvious than that of the injection of each drug individually (Fig. 7C–E). The results of ELISA assays showed that both drugs reduced the expression of IL-1β and TNF-α in the brain (Fig. 7F, G). Additionally, mouse brain tissue was used to prepare frozen sections for pathological detection. HE staining showed that Homer1 protein and selumetinib injection reduced the bleeding area of brain tissue (Fig. 7H, I) (Sham vs. ICH vs. Homer1 protein vs. Selumetinib vs. Both: 0.00 ± 0.00 vs. 0.77 ± 0.03 vs. 0.41 ± 0.02 vs. 0.44 ± 0.07 vs. 0.22 ± 0.23, respectively). Nissl staining showed that Homer1 protein and selumetinib injection improved the activity of neurons at the bleeding site (Fig. 7J, K) (Sham vs. ICH vs. Homer1 protein vs. Selumetinib vs. Both: 0.00 ± 0.00 vs. 0.85 ± 0.04 vs. 0.47 ± 0.04 vs. 0.48 ± 0.09 vs. 0.28 ± 0.04, respectively), while TUNEL staining showed that Homer1 protein and selumetinib injection significantly reduced the number of apoptotic cells in brain tissue (Fig. 7L, M) (Sham vs. ICH vs. Homer1 protein vs. Selumetinib vs. Both: 5.60 ± 3.71 vs. 1039 ± 134.6 vs. 293.8 ± 55.86 vs. 398.2 ± 41.84 vs. 98.20 ± 53.49, respectively). Finally, we found that the drugs effectively reduced the Longa score (Fig. 7N) (Sham vs. ICH vs. Homer1 protein vs. Selumetinib vs. Both: 0.00 ± 0.00 vs. 2.80 ± 0.42 vs. 2.00 ± 0.67 vs. 1.80 ± 0.63 vs. 1.60 ± 0.70, respectively) on the third day after ICH and significantly prolonged the survival time of mice after ICH (Fig. 7O).

To study the detailed mechanism of Homer1 protein in the treatment of ICH, hemorrhagic brain tissues of the Sham group, ICH group, and Homer1^flox/flox/Nestin-Cre^+/− mice injected with Homer1 protein were used for the detection of 200 index protein chips. The protein with the most obvious multiple changes between the ICH and Homer1 protein treatment groups was selected for further analysis (Fig. 8A, B). We found that compared with the ICH group, the expression levels of TNF-related weak inducer of apoptosis (TWEAK) (Fig. 8C), persephin (Fig. 8D), and activin A (Fig. 8E) in the Homer1 protein treatment group increased significantly, whereas the expression of TNF-α, tumor necrosis factor ligand superfamily member 10 (TNFSF10) (Fig. 8F), and IL-1β decreased significantly. However, using hemorrhagic brain tissue for clinical molecular detection is inconvenient. To verify whether the changes in these molecules specifically reflect the effectiveness of Homer1 protein in the treatment of ICH, we extracted the cerebrospinal fluid of mice for ELISA detection. The expression of TWEAK (Fig. 8G), persephin (Fig. 8H), and activin A (Fig. 8I) was upregulated after Homer1 protein treatment, while the expression of TNFSF10 (Fig. 8J) was downregulated. These data suggest that the mechanism of action of Homer1 protein in the treatment of ICH may be related to TWEAK, activin A, TNFSF10, and persephin. The expression changes of these proteins reflect the effectiveness of Homer1 protein in the treatment of ICH.