Efferocytosis is restricted by axon guidance molecule EphA4 via ERK/Stat6/MERTK signaling following brain injury

Background Efferocytosis is a process that removes apoptotic cells and cellular debris. Clearance of these cells alleviates neuroinflammation, prevents the release of inflammatory molecules, and promotes the production of anti-inflammatory cytokines to help maintain tissue homeostasis. The underlying mechanisms by which this occurs in the brain after injury remain ill-defined. Methods We used GFP bone marrow chimeric knockout (KO) mice to demonstrate that the axon guidance molecule EphA4 receptor tyrosine kinase is involved in suppressing MERTK in the brain to restrict efferocytosis of resident microglia and peripheral-derived monocyte/macrophages. Results Single-cell RNAseq identified MERTK expression, the primary receptor involved in efferocytosis, on monocytes, microglia, and a subset of astrocytes in the damaged cortex following brain injury. Loss of EphA4 on infiltrating GFP-expressing immune cells improved functional outcome concomitant with enhanced efferocytosis and overall protein expression of p-MERTK, p-ERK, and p-Stat6. The percentage of GFP+ monocyte/macrophages and resident microglia engulfing NeuN+ or TUNEL+ cells was significantly higher in KO chimeric mice. Importantly, mRNA expression of Mertk and its cognate ligand Gas6 was significantly elevated in these mice compared to the wild-type. Analysis of cell-specific expression showed that p-ERK and p-Stat6 co-localized with MERTK-expressing GFP + cells in the peri-lesional area of the cortex following brain injury. Using an in vitro efferocytosis assay, co-culturing pHrodo-labeled apoptotic Jurkat cells and bone marrow (BM)-derived macrophages, we demonstrate that efferocytosis efficiency and mRNA expression of Mertk and Gas6 was enhanced in the absence of EphA4. Selective inhibitors of ERK and Stat6 attenuated this effect, confirming that EphA4 suppresses monocyte/macrophage efferocytosis via inhibition of the ERK/Stat6 pathway. Conclusions Our findings implicate the ERK/Stat6/MERTK axis as a novel regulator of apoptotic debris clearance in brain injury that is restricted by peripheral myeloid-derived EphA4 to prevent the resolution of inflammation. Supplementary Information The online version contains supplementary material available at 10.1186/s12974-023-02940-5.


Background
Traumatic brain injury (TBI) is a major cause of morbidity and mortality worldwide [1].Neuroinflammation is the primary driver of secondary injury after TBI, with microglia and peripheral-derived macrophages (PDMs) playing critical roles in regulating the production of proinflammatory mediators [2].These two ontogenically distinct phagocytes mediate neuronal dysfunction by producing inflammatory and cytotoxic molecules while promoting neuroprotection through efferocytosis, the process of clearing apoptotic cell debris formed after tissue damage [3][4][5].
Efferocytosis is a complex, well-orchestrated process to remove apoptotic cell debris by phagocytes to enable the resolution of inflammation and tissue repair.However, a prolonged or excessive inflammatory response reduces efferocytosis efficiency and delays tissue recovery [5].Recognition and engulfment of apoptotic cells are mediated by a network of interactions, including "find-me", "eat-me", and engulfment signals.During inflammation, apoptotic cells secrete molecules as "find-me" signals that attract phagocytes to eliminate the corpses.The cell surface expression of "eat-me" signals ensures the recognition of apoptotic cells by phagocytes.Phosphatidyl serine (PtdSer) is the most recognized "eat-me" signal expressed on the surface of apoptotic cells, which binds directly or indirectly to phagocytic receptors to initiate the engulfment.The indirect binding of PtdSer to the phagocytic receptors is mediated by engulfment signals secreted by phagocytes.PtdSer/(Gas6 or Pros1)/MERTK are wellcharacterized binding complexes for phagocyte efferocytosis, including macrophages and microglia.Although apoptosis is a hallmark of brain trauma and a determinant of the severity of TBI outcomes [6][7][8], the mechanisms regulating efferocytosis following injury remain ill-defined.
The erythropoietin-producing hepatocellular carcinoma receptors type A (EphA) are tyrosine kinase receptors which, upon activation by cell surface-bound ephrin ligands, initiate bidirectional signaling implicated in neurodevelopment [9].Following brain trauma, EphA4 is overexpressed in the damaged cortical tissues, and global deficiency of this receptor demonstrates neuroprotection [10].In prior work, we found that EphA4 is upregulated on both microglia and PDMs.Interestingly, conditional deletion of EphA4 in resident microglia failed to improve TBI outcomes [11]; however, peripheral EphA4 deficiency reduced neuroinflammation and showed neuroprotection [12].The current study reveals a key mechanism by which peripheral-derived EphA4 limits the innate immune response to resolving tissue inflammation in the brain.This is the first evidence that efferocytosis occurs in a limited fashion in the brain following a traumatic insult and that enhancing this process via targeting MERTK/EphA4 may be a novel strategy to improve outcomes.

Animals
Male Cx3cr1 CreER/+ , Epha4 f/f , Epha4 +/+ /ROSA mTmG / Tie2-Cre, and Epha4 f/f / ROSA mTmG /Tie2-Cre mice were housed in an AAALAC-accredited animal facility with a 12-h light-dark cycle and food and water provided ad libitum.Animal experiments were performed following the NIH Guide for the Care and Use of Laboratory Animals and approved by the Virginia Tech Institutional Animal Care and Use Committee (IACUC; #21-044).Epha4 f/f recipient mice,Epha4 +/+ /ROSA mTmG /Tie2-Cre and Epha4 f/f / ROSA mTmG /Tie2-Cre donor mice were used for the chimeric generation, which were bred and genotyped as previously described [10,13,14].

Generation of bone marrow chimeric mice
Bone marrow ablation was induced by exposing EphA4 f/f recipient mice (6-8 weeks) to X-ray irradiation (two doses of 550 rad at six h apart).Within 24 h of irradiation, recipient mice were intravenously injected with 3-4 million bone marrow cells (BMCs) isolated from femur and tibia of donor Rosa26 mtmg /Tie2-Cre/EphA4 +/+ or Rosa26 mtmg /Tie2-Cre/EphA4 f/f mice as previously described [10,12].Recipient mice were placed on gentamycin sulfate water (1 mg/ml) for 3 days before irradiation and 2 weeks after adoptive transfer.Controlled cortical impact (CCI) injury for adoptive transfer WT (WT +WTBMCs ) and KO (WT +KOBMCs ) was performed at four weeks post-chimera generation.

Controlled cortical impact injury
CCI injury was induced as previously described [10,11].Briefly, mice were anesthetized with subcutaneous ketamine (100 mg/kg) and xylazine (10 mg/kg) and securely positioned in a stereotaxic frame, maintaining a body temperature of 37 °C.A midline incision was made on the sanitized and shaved scalp to expose the skull.Then, a craniectomy (4 mm diameter) was performed over the right parietal-temporal cortex (− 2.5 mm A/P and 2.0 mm lateral from the bregma) using a portable drill.The injury was induced with a velocity of 5 m/s and a depth of 2 mm.The incision was sutured, and animals were monitored until recovery.

Perfusion fixation and brain serial sectioning
At 1 or 3 days post-CCI injury (dpi), mice were euthanized by subcutaneous (s.c.) injection of a combination of 150 mg/kg ketamine and 20 mg/kg xylazine and perfusion fixation was performed as previously described [11].Five minutes before perfusion, heparin (2000 units/ kg, s.c.) and sodium nitroprusside (0.75 mg/kg, s.c.) were injected.After confirming the loss of pedal reflex, cardiac perfusion of heparin (20 units/ml) in phosphate-buffered saline (PBS) was performed to clear the circulation of blood, followed by perfusion with ice-cold 4% paraformaldehyde (PFA) in PBS.Fixed brains were cryoprotected with gradient sucrose solutions, snap-frozen, embedded in OCT compound with 30% sucrose, and stored at − 80 °C until sectioning.Serial coronal section (30 µm) were obtained using a cryostat (− 1.1 to − 2.6 mm posterior from bregma), mounted on positively charged slides (with five sections spaced 450 μm apart), and stored at − 80 °C for further analysis.

Single-cell sequencing, library generation, gene ontology enrichment, and computational analyses
At 1-day post-CCI injury, CD1 mice were euthanized using isoflurane, and 4 mm × 4 mm ipsilateral (injured) cortical tissues were harvested and dissociated using a papain digest neural dissociation kit (Miltenyi Biotec).Cells were then cryopreserved in 1 ml of CryoStore ® CS10 media (Stem cell technologies, Seattle, WA, USA) and sent to Medgenome for scRNAseq (Foster City, California, USA).ScRNA-seq libraries were generated using the Chromium Next GEM Single Cell 5' v2 chemistry (10 × Genomics) and sequenced on a NovaSeq 6000 (Illumina).An alignment of the libraries and read counts were performed using the mouse RNA-Seq Database consisting of expression values of 358 bulk Mouse RNAseq samples of sorted cell populations (Cell Ranger V7.0, 10X Genomics).Quality control was performed, excluding genes not expressed in at least 3 cells, and cells that do not express > 200 genes (Seurat, V 4.1.0Read10X function).Doublet Finder package, v2.0.3, was used to filter for doublets, LogNormalize was used for globalscaling normalization, and Seurat was used for clustering.Unbiased cell type recognition from scRNAseq data was performed using Cellenics.The following markers were used for the clustering of endothelial cells (Cd31, Tie2, Cdh5, Glut1), microglia (Tmem119, Ccr2-, cx3cr1), astrocytes (GFAP, Aldhl1, sox9), and monocytes/macrophages (Ccr2, Mmp8), as well as program automated clusters by brain tissue type.We applied modularity optimization by the Louvain algorithm to iteratively group cells together and visualize the data using UMAP.

Quantitative real-time PCR
Total RNA was freshly isolated from cultured bone marrow-derived macrophages, ipsilateral (4 × 4 mm), or contralateral (4 × 4 mm) cortical tissue using TRIzol ® reagent (Ambion) according to the manufacturer's instructions.RNA concentration was measured using a spectrophotometer ND-1000 (NanoDrop).DNAse I treatment (for a 1000 ng RNA) was performed to degrade the genomic DNA using an amplification grade kit (Sigma Aldrich, St. Lois, MO).To synthesize cDNA, reverse transcription was performed using the iScript ™ cDNA synthesis kit (Biorad, Hercules, CA).For qRT-PCR, iTaq ™ Universal SYBR ® Green Supermix (Biorad, Hercules, CA) and specific primers (Table 1) were used to amplify cDNA (50 ng) following the manufacturer's instructions.mRNA expression (fold change) was calculated relative to GAPDH as an internal control using the 2 −ΔΔCT method.

Immunohistochemistry and confocal image analysis
Serial coronal sections were washed in PBS, subsequently blocked 2% cold water fish gelatin/0.2%Triton X-100 (Sigma, Inc.) for 2 h, and then incubated overnight with the following primary antibodies diluted in blocking solutions: Rb anti-IBA1 (Wako) antibody

Stereological cell counts
Cell quantification was assessed by a blinded investigator using the optical fractionator probe function of Stereoinvestigator (MicroBrightField, Williston, VT, USA) on an upright Olympus BX51TRF motorized microscope (Olympus America, Center Valley, PA, USA) with a grid size set at 450 × 450 mm and 150 × 150 mm counting frame for cortex as previously described [15,16].

TUNEL staining, imaging and counting
TUNEL staining of serial coronal sections of perfused fixed brains was performed using the Click-iT Plus TUNEL Assay 647 (Thermo Fisher Scientific) according to the manufacturer's instructions.Serial coronal section (300 μm) were permeabilized with 0.2% Triton X-100 in phosphate-buffered saline (PBS) and incubated with the TUNEL reaction mixture containing terminal deoxynucleotidyl transferase enzyme, followed by AlexaFluor 647-conjugated nucleotides.Following the TUNEL reaction, the sections were blocked in 2% cold water fish gelatin (Sigma, Inc.) in 0.2% triton for 1 h, then incubated with Rb anti-IBA1 (Wako) or Rb anti-NeuN (cell signaling) antibody (1:200 in blocking solution) overnight.Slides were then washed with 1 × PBS, treated with AlexaFluor donkey anti-rabbit-555 (1:250 in block) for 1 h, further washed with 1 × PBS, and then mounted in media with DAPI counterstain (SouthernBiotech).Z-stack images were acquired using a Zeiss 880 confocal microscope (Carl-Zeiss, Oberkochen, Germany).

Western blot analysis
At 3 days post-CCI injury, mice were euthanized using isoflurane, and 4 mm x 4 mm ipsilateral (injured) or contralateral cortical tissues were harvested and freshly homogenized in RIPA lysis buffer (Pierce, ThermoFisher) containing protease/phosphatase inhibitor cocktail mix (Halt, ThermoFisher).Tissue lysate was centrifuged at 12,000 rpm for 20 min, and protein concentration was determined using BCA reagents (Thermo Fisher Scientific, Inc., Rockford, IL).Equal protein concentrations of each sample were loaded onto SDS-PAGE gels and transferred to PVDF (BioRad, Hercules, CA) using transfer buffer (containing 190 mM glycine, 25 mM Tris, 0.5% SDS, and 20% methanol).Membranes were blocked at room temperature using Intercept ® (TBS) blocking buffer (LI-COR Biosciences) for 2 h and then incubated overnight with the following primary antibodies (1:1000 in blocking solution): Rb anti-Phosphorylated MERTK (PhosphoSolutions, Aurora, CO), Rb anti-MERTK (Abcam), Rb anti-Phosphorylated ERK 1/2 (Cell signaling), Ms anti-ERK 1/2 (Cell signaling), Rb anti-Phosphorylated Stat6 (Cell signaling), Rb anti-Stat6 (Cell signaling), and Ms anti-actin (Cell signaling).Membranes were washed with TBST and then incubated with LICOR IRDye anti-rabbit and/or anti-mouse IgG (LI-COR Biosciences) for one hour.Protein bands were imaged with the LI-COR Odyssey system and quantified using Fiji Image-J software.

Cerebral blood flow analysis
A laser speckle contrast imaging system (RFLSI III Laser Speckle Imaging System, RWD Life Science, Dover, DE, USA) was used to scan cerebral blood flow over the right parietal cortex after performing a craniectomy (4 mm diameter).Cerebral blood flow was recorded in the region of interest (ROI, 2.5 mm diameter) for 30 s, and 4 readings for each recording were taken before injury (baseline), at 10 min, and 3 dpi.The average of the 4 readings and the percentage from the baseline were calculated for each mouse.Laser speckle contrast (LSC) and bright field images were captured at every recording.

Blood-brain barrier permeability (IgG deposition)
Serial coronal sections of perfused fixed brains were blocked in 2% cold water fish gelatin (Sigma, Inc.) in 0.2% triton for 1 h, incubated with AlexaFluor donkey antimouse-594 (1:250 in block) for 1 h, washed in 1 × PBS, and then mounted in media with DAPI counterstain (SouthernBiotech).The volume of IgG deposition (mm 3 ) was measured using the Cavalieri Estimator from Stere-oInvestigator software.

Statistical analysis
Data are presented as mean ± standard error of the mean (SEM), graphed, and analyzed using the Graph-Pad Prism program, version 9 (GraphPad Software, Inc., San Diego, CA).Shapiro-Wilk test of normal distribution was conducted for all data, and the results confirmed that all data sets follow a normal distribution of variance.The intergroup variations were analyzed using Student's unpaired two-tailed t-test (to compare between two experimental groups), one-way analysis of variance (ANOVA) followed by Bonferroni's multiple comparisons test (to compare multiple groups with one independent variable), or two-way ANOVA followed by Šídák's multiple comparisons test (to compare multiple groups with two independent variables).The variations were considered significant at P < 0.05.

Efferocytosis-related gene expression across single cells in the damaged cortex
Efferocytosis has not been fully characterized in brain injury following trauma.To identify specific cell types across the damaged cortex that express efferocytosisrelated genes, we conducted scRNA-seq expression analysis on cells dissociated from the ipsilateral cortex at 1-day post-CCI injury.A total of 5,000 cells were included in our analysis.Uniform Manifold Approximation and Projection (UMAP) plot shows different clusters of microglia, astrocytes, endothelial cells, pericytes, neutrophils, and monocytes/macrophages isolated from the injured cortex based on RNA gene expression (Fig. 1A).The analysis of differential expression of efferocytosis signals in different cell clusters revealed that the expression of the efferocytosis receptor Mertk is present predominately in microglia and astrocytes, as well as monocyte/ macrophages, and endothelial cells.Its cognate ligands, Gas6 and Pros1, predominate in microglia and endothelial cells (Fig. 1B).Feature UMAPs highlight the spatial location of these genes in the different clusters, confirming their enriched expression (Fig. 1C-F).

Upregulation of Mertk in the ipsilateral cortex following CCI injury
To investigate the effect of CCI injury on efferocytosisrelated genes, we measured temporal mRNA expression of receptors involved in apoptotic cell recognition and engulfment in the ipsilateral cortex at 1, 3, and 7 dpi.We observed a significant increase in recognition receptors (S1pr1 and Cx3cr1), engulfment receptors (Mertk), and bridging molecules (Gas6 and Pros1) at 1, 3, and 7dpi (Fig. 2A-E).To determine if MERTK protein is upregulated in Cx3cr1-expressing cells following CCI injury, we used Cx3cr1 CreER mice, which express EYFP in Cx3cr1positive cells.Immunohistochemical analysis using Iba1 (a marker for microglia and PDM) revealed increased MERTK expression on IBA1 + /Cx3cr1 EYFP+ cells located in the peri-lesion at 3 dpi compared to the contralateral cortex (Fig. 2G-O).To distinguish between microglia and PDMs, we used GFP bone marrow chimeric wildtype mice.Quantification of MERTK was measured in serial coronal sections immuno-stained for IBA1 at 1 and 3 dpi.The estimated number of MERTK + /GFP + / IBA1 + PDMs and MERTK + /GFP-/IBA + microglia was measured in the ipsilateral cortex using non-biased stereology via an optical fractionator probe in the Stereoinvestigator program.Consistent with scRNAseq data, MERTK expression was present on microglia and PDMs; however, more microglia were positive for MERTK at 1dpi.MERTK + microglia count remained consistent at 3dpi, however, the estimated number of PDMs expressing MERTK was much higher.This could reflect the increased trafficking of monocytes to the brain at this time point (Fig. 2P-Z).Therefore, we show a distinct temporal change in the expression of MERTK across two distinct myeloid populations following brain injury.
Single-cell RNAseq analysis revealed an astrocyte cluster expressing Mertk mRNA at 1 dpi, implying a potential role in the engulfment process following CCI injury.Therefore, we conducted immunostaining of brain sections using GFAP and MERTK to assess whether MERTK protein is upregulated in astrocytes at 3 dpi.Our findings revealed that astrocytes constituted less than 10% of MERTK-positive cells, in contrast to the 85% prevalence among PDMs and 13% in microglia (Additional file 1: Fig. S1A-K), indicating a limited role of astrocyte MERTK in the efferocytosis process.

Initiation of efferocytosis by Cx3cr1 + microglia/ macrophages following CCI injury
To determine if the change in MERTK expression in microglia and PDMs is correlated with the initiation of apoptotic neuron engulfment, coronal sections of CCIinjured Cx3cr1 CreER mice were stained with TUNEL and anti-NeuN antibody (Fig. 3).Z-stacked confocal images show Cx3cr1 EYFP+ efferocytes containing DAPI + / TUNEL +, DAPI + /NeuN +, and DAPI + /TUNEL + / NeuN + nuclei along with DAPI + /NeuN-/TUNELnuclei in the peri-lesion cortex at 3 dpi (Fig. 3F-O).We also observed several GFAP + astrocytes approaching GFP + and NeuN + cells within the peri-lesion cortex (Additional file 2: Fig. S2A-E).However, due to the complex morphology of astrocytes and their non-amoeboid shape, confirming the actual engulfment of other cells by astrocytes presents a challenging endeavor and requires further investigation.Therefore, our data provide the first evidence for the engulfment of dead/dying neurons by microglia and PDMs in traumatic braininjured mice.

EphA4 deficiency in PDMs enhances the clearance of apoptotic debris in the damaged cortex following CCI injury
To evaluate the role of peripheral-derived EphA4 in this response, we used WT (WT +WTBMCs ) and EphA4 KO (WT +KOBMCs ) GFP BM chimeric KO mice.Bone marrow chimerism was verified by observing 94% and 96% of peripheral GFP + /CD45 + immune cells in WT +WTBMCs and WT +KOBMCs mice, respectively.A non-significant difference in GFP + CD45 + count was observed in WT +WTBMCs and WT +KOBMCs mice (Additional file 3: Fig. S3).We stained serial coronal sections of CCI-injured WT +WTBMCs and WT +KOBMCs GFP BM chimeric mice with anti-IBA1, anti-NeuN, or TUNEL and counted the percentage of PDMs (GFP + /IBA1 +) and microglia (GFP-/IBA +) engulfing other cells, including peripheral immune cells (cells contain two or more DAPI + nuclei including GFP cells; Fig. 4A-D), neurons (cells contain at least one NeuN + cell and one DAPI + nuclei; Fig. 4E-H), or TUNEL + cells (cells contain at least one TUNEL + and one TUNEL-nuclei; Fig. 4I-K) in the core and/or perilesion of the injured cortex.We found that the percentage of PDMs and microglia containing 2 + nuclei is significantly higher in the lesion core (Fig. 4C) and perilesion (Fig. 4D) of WT +KOBMCs than in WT +WTBMCs .The percentage of PDMs engulfing NeuN + cells in the lesion core (Fig. 4G) and the peri-lesion (Fig. 4H), as well as TUNEL + cells in the lesion core (Fig. 4K), is higher in WT +KOBMCs than WT +WTBMCs .Microglia engulfing NeuN + cells are significantly higher in the peri-lesion of WT +KOBMCs than WT +WTBMCs .No significant difference was observed in microglia engulfing NeuN + or TUNEL + cells in the lesion core.This data suggests that EphA4 deficiency in BM-derived immune cells enhances the efferocytosis capacity of PDMs and peri-lesional microglia.

Deficiency of peripheral-derived EphA4 enhances ERK/ Stat6/MERTK signaling in the CCI-injured cortex
Given that myeloid MERTK-mediated efferocytosis signaling may be predominant in the brain following CCI injury and is enhanced in the absence of EphA4 (Figs. 1, 2), we sought to investigate the mRNA and protein phosphorylation status of key players in this pathway.We found that mRNA expression of Mertk and Gas6 was significantly higher in the ipsilateral cortex of WT +KOBMCs at 3dpi when compared to WT +WTBMCs (Fig. 5A, B).No significant difference was observed in Pros1 transcript (Fig. 5C).Further, Western blot analysis showed increased phosphorylation of MERTK (P-MERTK), ERK 1/2 (P-ERK), and Stat6 (P-Stat6) in the ipsilateral cortex of WT +KOBMCs at 3 dpi compared to WT +WTBMCs (Fig. 5D and Additional file 4: Fig. S4).To specifically address whether ERK and Stat6 are present on MERTKexpressing cells, coronal sections were immune-stained with anti-MERTK and anti-P-ERK (Fig. 5E and F) or anti-MERTK and anti-P-Stat6 (Fig. 5G and H).Confocal images show colocalization of P-ERK and P-Stat6 with MERTK in the ipsilateral cortex of WT +KOBMCs mice and that Stat6 is increased on GFP-expressing cells in KO mice compared to WT.This indicates the coexpression of these key signaling molecules on infiltrating efferocytes.

Enhanced efferocytosis in EphA4-deficient bone marrow-derived macrophages (BMDMSs) is mediated by ERK/Stat6 pathway
To confirm that EphA4 deficiency in bone marrowderived macrophages enhances the clearance of apoptotic debris, an in vitro efferocytosis assay was performed.GFP + BMDMSs from Epha4 +/+ /ROSA mTmG /Tie2-Cre (WT) or Epha4 f/f / ROSA mTmG /Tie2-Cre (KO) mice were co-cultured with pHrodo-stained live or apoptotic Jurkat cells.Apoptosis was induced by treating Jurkat cells with staurosporine.After 3 h of treatment, 90% of Jurkat cells were in the early phase of apoptosis (Annexin V + and PI − , Additional file 5: Fig. S5).No engulfment of live Jurkat cells was observed in EphA4 WT or EphA4 KO macrophages (Fig. 6A-H).More engulfment of apoptotic Jurkat cells was observed in untreated, LPS-treated, and HMGB1-treated EphA4 KO macrophages when compared to WT (Fig. 6I).Importantly, treating macrophages with EphA4-FC clusters (to activate the reverse ephrin signaling) did not reduce the efferocytosis efficiency of EphA4 KO macrophages, confirming that the enhanced efferocytosis in KO macrophages is mediated by the blockade of EphA4 forward signals (Fig. 6J).In addition, Mertk and Gas6 expression is significantly upregulated in EphA4 KO macrophages in the absence of apoptotic Jurkat cells compared to WT macrophages.After engulfing apoptotic Jurkat cells, Mertk and Gas6 expression increased in WT and KO macrophages; however, the expression is significantly higher in KO than in WT macrophages (Fig. 6K, L).Pros1 expression increased only in EphA4 KO macrophages after apoptotic Jurkat cell engulfment (Fig. 6M).To determine if enhanced efferocytosis in EphA4 KO macrophages is mediated by MERTK, ERK, and Stat6 activation, selective inhibitors were used.MERTK inhibitor (UNC2250) reduced efferocytosis efficiency of both WT and KO macrophages; however, ERK inhibitor (FR18020R) and Stat6 inhibitor (AS1517499) selectively reduced efferocytosis efficiency of EphA4 KO macrophages (Fig. 6N).Interestingly, ERK inhibition reduced Mertk expression in EphA4 KO macrophages engulfing apoptotic cells, and Stat6 inhibition reduced Mertk and Gas6 expression.A non-significant difference was observed in Pros1 expression in the presence of ERK or Stat6 inhibitors (Fig. 6O).Data suggest that EphA4 forward signaling reduces macrophage efferocytosis by inhibiting ERK and Stat6 activation, which in turn reduces the expression of efferocytosis receptor (MERTK) and its ligand (Gas6) and restricts the efferocytosis process (Fig. 6P).

Peripheral EphA4 deficiency reduced apoptotic cell counts and improved functional outcomes following CCI injury
To determine if enhanced efferocytosis in bone marrow chimeric EphA4 KO mice is associated with a reduction in the number of remaining uncleared apoptotic cells, the number of TUNEL + cells was counted in serial coronal sections at 3 dpi using the optical fractionator probe function on Stereoinvestigator (Fig. 7A).The total number of apoptotic cells (total TUNEL +) and apoptotic peripheral-derived immune cells (GFP + TUNEL +) were significantly reduced in WT +KOBMCs compared to WT +WTBMCs mice (Fig. 7B-G, M).This effect was correlated with improved cerebral blood flow and reduced blood-brain barrier permeability.Cerebral blood flow was measured at 10 min and 3dpi using Laser Speckle Contrast Imaging (LSCI) and calculated as percentage flow from the baseline (pre-injury).WT +KOBMCs showed a significant increase in cerebral blood flow at 3dpi compared to WT +WTBMCs (Fig. 7H, I, N).Blood-brain barrier permeability was measured at 3 dpi by estimating the volume of IgG deposition in serial coronal section using the Cavalieri Estimator from the Stereoinvestigator software.Significant reduction in IgG deposition was observed in WT +KOBMCs compared to WT +WTBMCs , indicating improved blood-brain barrier function and neuroprotection (Fig. 7J-L).We find greater improvements in functional outcome correlate with efferocytosis enhancement by gene deletion of EphA4 on infiltrating innate immune cells.

Discussion
Apoptosis is a prominent mode of cell death that mediates tissue damage and may contribute to the propagation of neuroinflammation following brain injury.In cases of excessive apoptotic cell death, an overwhelmed clearance mechanism may result in the accumulation of apoptotic cells and cellular debris, prolonging the exposure to pro-inflammatory signals and activated immune cells.Therefore, a greater understanding of the efferocytotic process may aid in developing strategies targeting the timely removal of dying cells in the brain.Our results demonstrate an upregulation of "find-me" signal receptors (S1pr1 and Cx3cr1), engulfment receptor (MERTK), and bridging molecules (Gas6 and Pros1) in the damaged cortex, indicating the onset of efferocytosis acutely following trauma.We identified that resident (microglia) and peripheral-derived (monocyte/macrophages; PDMs) myeloid cells show distinct temporal expression patterns of MERTK.Notably, we find that ephrin type-A receptor 4 (EphA4) deficiency on infiltrating immune cells enhanced the capacity of PDMs and peri-lesional microglia for efferocytosis and enhanced cortical expression of Mertk and Gas6 transcripts.Our investigation into the molecular mechanisms revealed increased levels of phosphorylated p-MERTK, p-ERK, and p-Stat6 in the injured cortex of chimeric EphA4 KO mice.In addition, inhibition of ERK and Stat6 attenuated the enhanced efferocytosis and Mertk expression in EphA4-deficient BMDMSs in vitro.This demonstrates that EphA4 suppresses efferocytosis by inhibiting the ERK/Stat6 pathway.These findings highlight a new and novel role for the axon guidance molecule EphA4 in regulating the coordinated process of efferocytosis, which may contribute to the overall neuroinflammatory milieu after brain injury.
EphA4 has been identified as a regulator of neuroinflammation and secondary injury following brain trauma.Our previous studies utilizing mouse models and bone marrow chimeric approaches have demonstrated that the absence or inhibition of EphA4 in peripheral-derived monocytes/macrophages results in neuroprotection, improved motor function, reduced cortical infiltration of monocytes/macrophages, and a shift in their gene profile from a pro-inflammatory to an anti-inflammatory state [10,12].These findings underscore the pivotal role of EphA4 in mediating the pro-inflammatory phenotypic state of PDM, thereby prompting further investigation into whether EphA4 hampers the adequate clearance of apoptotic debris by modulating the phenotypic state of efferocytes following CCI injury.Interestingly, the deletion of EphA4 on infiltrating immune cells fosters the augmentation of efferocytosis in both PDMs and resident microglia.This may be because EphA4 deficiency in monocytes enhances the expression of Gas6, which may fuel the neighboring microglia and promote their efferocytotic abilities.This also suggests that monocytes may have critical properties, allowing them to communicate with and regulate microglial function.
Eph receptors and ephrin ligands facilitate intercellular communication to regulate diverse processes, including adhesion, repulsion, migration, survival, proliferation, remodeling, and differentiation.Upon binding, the clustering of Eph receptors and ligands triggers a cascade of signaling events in both the cells expressing the receptor, and those bearing the ligand [18].EphA4-null BMD-MCs, treated with clustered EphA4-FC, showed no effect on efferocytosis efficiency, confirming that forward EphA4 signaling on monocyte/macrophages mediates the suppression of ERK/Stat6 signaling and subsequent enhancement of efferocytosis.Whether the genetic deletion of PDM EphA4 may impair reverse ephrin signaling on apoptotic cells, and its relevance in the current study remains unknown and requires additional investigation.Forward Eph signaling exhibits a remarkable ability to attenuate the RAS-ERK pathway, overriding its activation by other receptor tyrosine kinases [18,19].This inhibition has been observed in diverse scenarios, including the modulation of growth cone motility in neurons [20] and the suppression of tumorigenicity in cancer cells [21,22].The mechanism underlying Eph receptor-dependent ERK inhibition often involves the activation of the RAS GTPase, p120 RAS GAP, which leads to the subsequent inactivation of H-RAS [23].Conversely, Eph receptors can activate the RAS-ERK pathway in specific contexts, promoting cellular processes such as proliferation, early gene transcription, cell migration, or repulsion [24,25].Eph receptor signaling exerts a complex and contextdependent influence on the RAS-ERK pathway, however, our novel findings indicate that EphA4 suppresses this signaling to limit efferocytosis in the brain following injury.
ERK1/2 signaling modulates multiple aspects of efferocytosis and inflammation resolution, including the regulation of phagocytic receptor expression, cytoskeletal rearrangement, and the production of anti-inflammatory molecules [26,27].Activation of the ERK1/2 signaling pathway can enhance the expression of phagocytic receptors, including MERTK, Gas6, MFG-E8, and integrins (such as αvβ3) [28][29][30].In addition, ERK1/2 signaling is implicated in promoting actin polymerization and cytoskeletal rearrangement, thereby facilitating the process of efferocytosis [31].Importantly, ERK1/2 activation also regulates the production of anti-inflammatory mediators during efferocytosis, promoting the polarization of macrophages toward a proresolving phenotype, which contributes to sustained efferocytosis and the resolution of inflammation [32,33].One crucial transcription factor influenced by ERK1/2 activation to regulate macrophage polarization is Stat6 [34,35].The phosphorylation of Stat6 promotes the transcription of genes involved in macrophage proresolving and efferocytosis response, such as Gas6 and MERTK [36,37].In the present study, we observed an activation of MERTK, ERK1/2, and Stat6 in the injured cortex of chimeric EphA4 KO mice.Furthermore, selective inhibition of ERK and Stat6 reduced efferocytosis and MERTK expression, specifically in EphA4-deficient BMDMS in vitro.These findings suggest that EphA4 impedes efferocytosis by inhibiting the P-ERK/P-Stat6/ MERTK signaling pathway.

Conclusions
Efferocytosis occurs acutely in the brain following trauma in a limited capacity.Blockade of EphA4 receptor function improves this activity, which coincides with tissue protection and restoration of cerebral blood flow and BBB stability.Unraveling the molecular mechanisms underlying efferocytosis in TBI will lead to new therapeutic avenues promoting this process, mitigating the harmful consequences of apoptotic death, and facilitating optimal recovery following brain trauma.

Fig. 1 Fig. 2 (
Fig. 1 ScRNA-seq analysis displays a differential expression of efferocytosis signals in the brain following CCI injury.A Uniform Manifold Approximation and Projection (UMAP) plot showing different cell clusters and their cell-specific annotation in the injured cortex at 1dpi.B Dot plot of efferocytosis genes expressed by each cluster type.C-F Feature maps highlighting the expression across clusters of MERTK (C), Gas 6 (D), Pros1 (E), and Stat6 (F)

Fig. 6
Fig. 6 EphA4-null BMDMs show enhanced efferocytosis regulated by ERK/ Stat6/MERTK.A-I EphA4 deletion improves the efferocytosis efficiency of BMDMS in vitro.A-H Representative images showing the engulfment of the pHrodo-stained apoptotic (but not live) Jurkat cells (red) by GFP + untreated WT (A-D) and EphA4 KO (E-H) BMDMS.I Quantification of the efferocytosis efficiency of WT and EphA4 KO BMDMS after stimulation with LPS and HMGB1 for 4 h.J Efferocytosis of EphA4 KO BMDMSs is mediated by the blockade of forward EphA4, not reverse ephrin signals.Treatment of WT and EphA4 KO BMDMS with clustered EphA4-FC to activate reverse ephrin signals did not reduce the efferocytosis of EphA4 KO BMDMS.K-M mRNA expression of Mertk (K), Gas6 (L), and Pros1 (M) with and without engulfment of apoptotic Jurkat cells.EphA4 KO BMDMSs have higher expression of Mertk and Gas6 than WT.N The use of MERTK inhibitor reduced efferocytosis of both WT and EphA4 KO BMDMS; however, Stat6 and ERK inhibitors selectively reduced efferocytosis in EphA4 KO BMDMS.O Stat6 inhibitor reduced Mertk and Gas6 expression, and ERK inhibitor reduced Gas6 expression in KO BMDMS engulfing apoptotic Jurkat cells.P Suggested pathway for the regulation of efferocytosis by EphA4.N = 5-6 mice/group.Ns = non-significant; *P < 0.05; **P < 0.01; ***P < 0.001, ****P < 0.0001.Two-way ANOVA followed by Šídák's multiple comparisons test (I-N) or one-way ANOVA followed by Tukey's multiple comparisons test (O).Scale bar = 50 µm in A-H

Fig. 7
Fig. 7 Peripheral EphA4 deficiency reduces apoptotic cell count and improves cerebral blood flow and BBB stability after CCI injury.A Experimental outline.B-G, M Apoptotic TUNEL + (purple) cell count is reduced in WT +KO BMCs (B, D, E) compared to WT +WT BMCs (C, F, G) mice.The total number of TUNEL + and GFP + /TUNEL + cells (M) was counted using image J in 3 serial sections using 5 × 7 tiled z-stack confocal images.H, I, N Cerebral blood flow is improved in WT +KO BMCs compared to WT +WT BMCs .Blood flow was measured using laser speckle contrast imaging (LSCI) at 10 min and 3 days post-injury and presented as a percentage perfusion of the baseline.J-L Blood-brain barrier permeability is improved in WT +KO BMCs (K) compared to WT +WT BMCs (J).The volume of IgG deposition was measured at 3 dpi in serial coronal sections using Cavalieri Estimator from StereoInvestigator (L).N = 5-6 mice/group.Ns = non-significant; *P < 0.05; **P < 0.01; ***P < 0.001.t-tests (L) or multiple t-tests (M, N).Scale bar = 500 µm in B, C, J, K and 50 µm D-G

Table 1
Sequence of primers used in qRT-PCR