BCR-Associated Protein 31 Regulates Macrophages Polarization and Wound Healing Function via Early Growth Response 2/C/EBPβ and IL-4Rα/C/EBPβ Pathways Free

The BCR-associated protein 31 (BAP31), with m.w. of 28 kDa, is an indispensable transmembrane protein in the endoplasmic reticulum (ER) (1). BAP31 has an important role in immunomodulation and is mainly involved in the transport regulation of a variety of immune cell membrane proteins, including membrane IgD (mIgD), MHC class I, and leukocyte surface β-type 2 integrin CD11b/CD. Previous studies in our laboratory have shown that BAP31 deficiency upregulates LPS-induced proinflammatory cytokines in BV2 cells and mice through the IRAK1 pathway and regulates the functions of T cells through the TCR signaling pathway (2, 3). To date, the effect of BAP31 on macrophages has not been investigated. In this study, Lyz2 Cre-BAP31^flox/flox mice were bred to study the role of BAP31 in macrophages.

Macrophages are important in the regulation of host defense in organisms (4). Resting macrophages (M0) may polarize to form different phenotypes under different influencing factors, such as the proinflammatory phenotype (M1) and anti-inflammatory phenotype (M2) (5, 6). In vitro, M2 macrophages may be induced from M0 by CSF (M-CSF), IL-4, and IL-13, and express or release specific markers (Arg1, CD206, Chitinase-like 3 [Ym1], and Resistin-like alpha [Fizz1]) and anti-inflammatory cytokines (IL-10, TGF-β1) that are involved in tissue repair (7, 8). IL-4 produced by Th2 cells activates M2 polarization through the IL-4Rα/JAK/STAT6 signaling pathway (7). In addition, M2 macrophages play a decisive role in tissue fibrosis and regulation of tissue responsiveness to angiogenic therapies (9). M2 macrophages have a reparative phenotype and can participate in the entire process of repairing skin injury: (1) hemostasis, (2) inflammation, (3) epithelial remodeling, and (4) remodeling (10–13). Several key transcription factors are clearly associated with M2 macrophages polarization, such as peroxisome proliferator–activated receptor (PPAR), CREB-C/EBP, hypoxia-inducible factors (HIFs), NF-κB, and IFN regulatory factors (IRFs). PPAR-γ is a ligand-activated nuclear receptor with potent anti-inflammatory properties that modulate M2 macrophages (14). IRFs bind to coactivators, corepressors, and modifiers of M2 macrophages, resulting in specific and diverse molecular events (15). In addition, C/EBPβ regulates many M2-related genes, and CREB-C/EBPβ activity is required for wound healing (16, 17). C/EBPβ is a transcription factor and an important member of the C/EBPs family. C/EBPβ is involved in the regulation of cell proliferation and differentiation, tumorigenesis and apoptosis, and hematopoietic cell formation (18–20). Currently, it is known that C/EBPβ is important in M1 and M2 macrophages polarization. High expression of C/EBPβ in M1 macrophages and changes in the TLR4-HMGB1/C/EBPβ pathway promote M2 polarization (21). Early growth response 2 (Egr-2), a member of a family of zinc-finger proteins, has been found to play a critical role in hindbrain development and myelination in the peripheral nervous system (21). A study indicated that Egr-2 recognized three sites of the C/EBPβ promotor elements (22).

We found that BAP31 regulates macrophages polarization via Egr-2/C/EBPβ and IL-4Rα/C/EBPβ pathways. In addition, BAP31 knockdown inhibits IL-4Rα degradation via the ubiquitin–proteasome pathway. In adeno-associated virus (AAV)–short hairpin (sh)- IL-4Rα–transfected Lyz2 Cre-BAP31^flox/flox mice, we confirmed that BAP31 affected macrophages functions, including angiogenesis, skin fibrosis, and the speed of wound healing during the wound healing process through IL-4Rα. These novel roles of BAP31 in macrophages might shed light on the future treatment of chronic wounds.

First, we constructed a targeting vector containing intron of BAP31 and flanked with two loxp sites. The neomycin-resistant gene cassette was flanked with two flippase recognition target sites. The targeting vector was electroporated into embryonic stem cells in C57BL/6 mice (23). Then, BAP31^flox/flox mice were mated with age-matched Lyz2 Cre transgenic mice to generate Lyz2 Cre-BAP31^flox/flox and BAP31^flox/flox mice. Same-sex and littermate mice (Lyz2 Cre-BAP31^flox/flox and BAP31^flox/flox) were used in all experiments. All experiments were performed using mice aged 8–10 wk. Animal experiments were reviewed and approved by the Institutional Animal Care and Use Committee of Northeastern University. All mice were treated in accordance with the Guide for the Care and Use of Laboratory Animals.

Bone marrow–derived macrophages (BMDMs) were extracted from mice 6–9 wk of age and then cultured in RPMI 1640 medium (Life Technologies BRL, CA) with 10% (v/v) FBS, penicillin (100 U/ml), streptomycin (100 μg/ml), and 10% L929 cellular supernatants (24). L929 cells were cultured for 5 d in RPMI 1640 medium (Life Technologies BRL) with 10% (v/v) FBS, penicillin (100 U/ml), and streptomycin (100 μg/ml); then we collected the supernatants. BMDMs were cultured for 7 d and fully differentiated in L929 supernatants for experiment. RAW 264.7 mouse mononuclear macrophages leukemia cell line was obtained from the American Type Culture Collection (Manassas, VA). RAW264.7 cells were maintained in DMEM (Life Technologies BRL) with GlutaMAX containing 10% (v/v) FBS, penicillin (100 U/ml), and streptomycin (100 μg/ml). The cultures were maintained at 37°C in a humid incubator with a 5% (v/v) CO₂ atmosphere.

For knockdown, the cells were transfected with 5 μg of plasmid DNA (1.5 μg PSPAX2, 1.5 μg PMD2G, and 2 μg PL/short hairpin RNA/GFP-mouse-BAP31). For overexpression, the cells were transfected with a total of 10 μg of plasmid DNA (2.5 μg PLP1, 1.75 μg PLP2, 2.5 μg PCMV, and 3.25 μg PLBAP31-flag or 3.25 μg PLEgr-2-flag). A total of 3 × 10⁶ 293T cells were seeded in a 60-mm dish for 24 h before transfection in the earlier plasmid mixture. Cells transfected with control vectors were used as controls. 293T cell supernatants with the packaged virus were collected after transfection for 48 or 72 h and filtered through 0.45-μm cellulose acetate filters. Then the supernatant was transferred to a six-well plate containing 1 × 10⁶ RAW264.7 cells or BMDMs. Seventy-two hours later, the cells were observed using a Leica DMI3000 B fluorescence microscope (Wetzlar, Germany). Transfection efficiency was quantified using Western blotting.

A total of 1 × 10⁶ BMDM cells were seeded in a six-well plate, then transfected small interfering RNA (siRNA) was mixed with Lipofectamine 8000 according to the instructions of the manufacturer. The negative siRNA was used to generate control cells. After transfection for 72 h, real-time PCR and Western blot were used to detect the transcription levels and protein expression of the targets. The sequences of siRNA used in this study were as follows: siRNA–Egr-2 sense, 5′-CCU CGA AAG UAC CCU AAC ATT-3′ and siRNA–Egr-2 antisense, 5′-UGU UAG GGU ACU UUC GAG GTT-3′; siRNA–IL-4Rα sense, 5′-GCC AGG AGU CAA CCA AGU ATT-3′ and siRNA–IL-4Rα antisense, 5′-UAC UUG GUU GAC UCC UGG CTT-3′ (Gene Pharma, Shanghai, China).

BMDMs were collected, and the cell concentration was adjusted to 5 × 10⁴/100 µl. After the cells were fixed in 2% paraformaldehyde for 30 min and washed twice with cold PBS with 2% BSA, Fc receptor blocker (Miltenyi Biotec) was added and incubated in an ice-water bath for 10 min. The cells were washed once with cold PBS with 2% BSA, followed by the addition of FITC-CD11b (BD) and PE-F4/80 (BD) mAbs and incubated in 4°C for 45 min. The cells were then analyzed using Fortessa or Accuri C6 software (BD).

Total RNA was isolated from the earlier two cells using TRIzol reagent (Thermo Fisher Scientific) according to the manufacturer’s instructions. The RNA samples (2 μg) were used for the synthesis of cDNA by reverse transcription using a GoScript Reverse Transcriptase kit (Promega, Madison, WI). Real-time PCR was performed with a CFX96 Touch Real-Time PCR Detection System (Bio-Rad Laboratories). The PCR mixture was at a volume of 10 μl containing 5 μl SYBR Premix Ex Taq (Promega), 0.5 μM each of the primers, and 1 μl cDNA prepared as described and using the following conditions: 95°C for 2 min, 95°C for 15 s, and 60°C for 60 s for 40 cycles. The results were analyzed according to the 2^−ΔΔCq formula. The sequences of primers are listed in Supplemental Table I.

The collected cells were suspended in PBS and fixed in 1 ml 2% paraformaldehyde on ice for 1 h. Fc receptor blocker (Miltenyi Biotec) was added and incubated for 10 min on ice-water bath. The cells were washed once with cold PBS, followed by adding FITC-CD11b (BD) and PE-F4/80 (BD) mAb to the cells; then cells were suspended in 100 μl of 0.1% saponin (Sigma, St. Louis, MO) mixed with 1:1000 each of purified anti-Ym1 rabbit Ab, anti-CD206 rabbit Ab (EB), or anti-C/EBPβ rabbit Ab (Abcam, Cambridge, MA). After washing, samples were subjected to analysis on an FACS Accuri C6 (BD) and were analyzed using FlowJo 7.6 software (Tree Star, Ashland, OR).

Proteins were isolated from the cells in buffer radio immunoprecipitation assay at 4°C for 30 min before being centrifuged at 12,000 × g/15 min. Then the supernatant was collected and added with 5× NaDodSO₄ (SDS) sample buffer and boiled for 10 min. Lysates (20 μg) were electroblotted onto a polyvinylidene difluoride membrane (Millipore) after electrophoretic separation in 10 or 12% SDS-polyacrylamide gels. Then the membranes were blocked with 5% BSA-TBST at room temperature for 1 h, followed by incubation overnight at 4°C with the primary Abs and then washed with TBST three times and incubated with the secondary Ab for 1 h at room temperature. The signal was detected with ECL (Tanon, Shanghai, China). Protein bands were documented on a chemiDoc XRS+ System (Bio-Rad, Munich, Germany). OD analysis of signals was performed with Image Lab software (Bio-Rad). The band intensities of the target proteins were calculated and normalized to that of β-actin. Ab sources were listed: anti-BAP31, anti-C/EBPβ, anti-CREB, anti–Egr-2, anti–IL-4Rα, anti–α-SMA, and anti-ubiquitin were from Abcam; anti–Arginase-1 (anti–Arg-1), anti-JAK3, anti–p-JAK3, anti-STAT6, anti–p-STAT6 were from Cell Signaling Technology (Danvers, MA); and anti-Ym1 was from STEMCELL.

Cells were treated with radioimmunoprecipitation lysis buffer (Beyotime Biotechnology, Shanghai, China) and mixed with protein G (Beyotime Biotechnology) for 30 min. The lysates were incubated with Abs at 4°C for 2 h; then we washed the complexes bound to protein G three times and eluted in 5× SDS sample buffer. All samples were consistently analyzed by Western blot analysis.

An AAV carrying sh-IL-4Rα under the control of a CD68 promoter was purchased from Miaolingbio (Wuhan, China). 293T cells were transfected with the plasmid; the virus supernatant was collected, purified, and counted; and 1010 viral particles of AAV–sh-IL-4Rα were administered to the Lyz2 Cre-BAP31^flox/flox mice and BAP31^flox/flox mice through intraspinal injection. The skin injury test was performed after 4 wk.

Mice were anesthetized with an i.p. injection of chloral hydrate (3%, 350 mg/kg). Wounds were created using sterile ophthalmic scissors in the back deep to the fascia and the square side length of 1 cm. In the trial group, IL-4 (5 ng/g; Sigma-Aldrich) was injected around the wound to induce the recruitment and polarization of M2 cells. The wound area was measured using ImageJ software (National Institutes of Health, Bethesda, MD) by a blinded researcher at the indicated time points (0, 1, 4, 7, 10, and 14 d). The percentage of wound area was calculated as follows: wound area (%) = (wound area on day n/wound area on day 0) × 100.

The blood vessels in wound sections were taken by Laser Speckle Imaging Instrument (RFLSI III) at day 4 after injury. Skin samples were harvested and fixed on days 4 and 14 after the injury. H&E and Masson's staining were performed to detect collagen deposition at the injury sites. The expression of DC31 was visualized immunohistochemically using mouse anti-mouse CD31 (Abcam). Immunofluorescence staining was performed to evaluate the angiogenesis and macrophages effects using mouse anti-mouse CD31 and mouse anti-F4/80 (Invitrogen, Carlsbad, CA). Moreover, anti–α-SMA Ab (Abcam) and anti-Col1A1 Ab (Abcam) were used to evaluate the myofibroblast infiltration at the injury sites. Cell nuclei were counterstained with DAPI.

Statistical analysis was performed using GraphPad Prism 7.0 software (GraphPad Software, San Diego, CA). All experimental data are reported as the mean ± SD. One-way ANOVA was used for comparison among multiple groups. The Student t test was used to compare two groups (significance: *p < 0.05, **p < 0.01, and ***p < 0.001).

To investigate the role of BAP31 in macrophages, we generated BAP31 conditional knockdown mice. BAP31^flox/flox were bred with Lyz2 Cre transgenic mice to induce the specific deletion of BAP31 in bone marrow macrophages (Supplemental Fig. 1). Bone marrow cells of Lyz2 Cre-BAP31^flox/flox (^−/−) mice and BAP31^flox/flox (^+/+) mice were extracted and induced with the supernatant of L929 cells to allow them to differentiate into BMDMs. FACS analysis revealed that the proportion of macrophages markers, F4/80 and CD11b, double-positive cells from both mice was >98% (Fig. 1A). The mRNA and protein levels of BAP31 were reduced ∼80% in BMDMs of ^−/− mice compared with that of ^+/+ mice using quantitative real-time RT-PCR (RT-qPCR) and Western blot analysis (Fig. 1B, 1C). We detected macrophage-associated genes expression involving M2 macrophages markers (Arg-1, Fizz1, Ym1, and CD206), angiogenesis-related genes (VEGFA, PLGF, b-FGF, Ang-1, Ang-2, and PDGF-BB), and fibrosis-related genes (Col1A1, TGF-β, Fibronectin). RT-qPCR analysis revealed that the mRNA level of these genes significantly decreased in BMDMs of ^−/− mice compared with that of ^+/+ mice (Fig. 1D–F). Similar results were demonstrated in peritoneal macrophages in ^−/− and ^+/+ mice (Supplemental Fig. 2A). FACS analysis indicated that BAP31 deficiency suppresses the expression of CD206 and Ym1 (Fig. 1G). In addition, from the BAP31-overexpressed RAW264.7 cells (Supplemental Fig. 2B, 2C), we found that the mRNA levels of Arg-1, Fizz1, Ym1, and CD206 increased (Supplemental Fig. 2D). These results indicate that BAP31 is involved in the expression of M2 macrophage–associated genes.

To clarify the involvement of BAP31 in the regulation of macrophages polarization, we investigated the transcription factors of M2 macrophages, including IRFs, C/EBPs, PPAR-γ, P50, KLF4, HIF-2α, IL-21, FABP4, and LXRα, by RT-qPCR analysis in BMDMs (25). We found that mRNA levels of IRF1, IRF2, IRF6, C/EBPβ, and LXRα were decreased in BMDMs of ^−/− mice, in which C/EBPβ was significantly affected (Fig. 2A–C). Similarly, BAP31 deficiency downregulated the protein levels of C/EBPβ and Arg-1 (Fig. 2D, Supplemental Fig. 2G). Conversely, with overexpression of BAP31 in RAW264.7 cells, the protein and mRNA levels of C/EBPβ increased (Supplemental Fig. 2E, 2F). Our findings indicate that BAP31 regulates macrophages polarization, perhaps through the transcription factor C/EBPβ.

Trying to make sense of the regulatory mechanisms of BAP31 on C/EBPβ expression, we first examined the binding of the two proteins. The immunoprecipitation (IP) analysis showed that there was no binding between BAP31 and C/EBPβ (Supplemental Fig. 3A). Therefore, the regulation of C/EBPβ by BAP31 is mediated by an intermediate protein. Our attention was drawn to CREB and Egr-2, which may regulate the expression of C/EBPβ. Egr-2 is an upstream regulator of C/EBPβ and capable of recognizing three sites of the C/EBPβ promotor elements (22, 26, 27) (Supplemental Fig. 3B). Western blot analysis showed that the expression of CREB was not altered, but the Egr-2 protein level was decreased in BMDMs of ^−/− mice compared with that of ^+/+ mice (Fig. 2E). Interestingly, the mRNA level of Egr-2 remained unchanged (Supplemental Fig. 3C). Therefore, it is suggested that BAP31 deficiency downregulated the Egr-2 protein level, possibly through posttranscriptional mechanisms. These results may imply that BAP31 regulates C/EBPβ through Egr-2. To confirm our speculation, we constructed Egr-2 overexpression plasmid and transferred it into BMDMs of ^−/− mice and found that overexpression of Egr-2 significantly recovered C/EBPβ expression (Fig. 2F, Supplemental Fig. 3D). Collectively, these data indicate that BAP31 regulates macrophages polarization through the Egr-2/C/EBPβ pathway.

Previous studies have reported that C/EBPβ plays an important role in both M1 and M2 macrophages (28, 29). IFN-γ and IL-4 could induce M1 and M2 polarization in macrophages, respectively. Therefore, we added the corresponding stimulation to further elucidate the impact of BAP31 on C/EBPβ. After 7 d of culture, BMDMs were stimulated with IFN-γ (20 ng/ml) and IL-4 (20 ng/ml) for 24 h, and the proteins were collected for Western blot analysis. The results showed that C/EBPβ and Arg-1 expression remained unchanged in IFN-γ–stimulated BMDMs of ^−/− mice compared with that of ^+/+ mice. Unexpectedly, the downregulation of C/EBPβ and Arg-1 induced by BAP31 deficiency recovered in IL-4–stimulated BMDMs of ^−/− mice (Fig. 3A). RT-qPCR analysis in IL-4–stimulated BMDMs also showed the same changes (Fig. 3B–D). Similarly, the FACS results revealed that IL-4 simulation recovered the downregulation of C/EBPβ induced by BAP31 deficiency (Supplemental Fig. 2H). Furthermore, we constructed RAW264.7 cells lines with BAP31 knockdown. The expression of BAP31 mRNA and protein in RAW264.7 cells was reduced by ∼70% compared with that of the control cells (Fig. 3E, 3F). As before, we evaluated the expression of C/EBPβ in RAW264.7 cells under IFN-γ and IL-4 stimulation, and the results of RAW264.7 cells were consistent with those of BMDMs (Fig. 3G–J). We realized that BAP31 regulates C/EBPβ, and M2 macrophages might be involved in the IL-4 pathway.

We sought to explore why the expression of C/EBPβ could recover in IL-4–stimulated BMDMs. We investigated the expression of p-JAK3, p-STAT6, and IL-4Rα in the IL-4 signaling pathway. The results demonstrated that the protein levels of p-JAK3, p-STAT6, and IL-4Rα were significantly increased in BMDMs of ^−/− mice (Fig. 4A), while the IL-4Rα mRNA level remained unchanged (Supplemental Fig. 3E). We then transferred IL-4Rα siRNA into BMDMs of ^−/− mice and found that IL-4Rα knockdown blocked the recovery of C/EBPβ in IL-4–stimulated BMDMs (Fig. 4B, Supplemental Fig. 3F). Due to the negative regulation of BAP31 on IL-4Rα, the effect of BAP31 on its degradation was examined. The BMDMs were treated with cycloheximide to inhibit protein biosynthesis, and IL-4Rα protein levels were detected at various time points. The results showed that knockdown of BAP31 in BMDMs resulted in a significant inhibition of IL-4Rα degradation rate (Fig. 4C). In contrast, with MG132 (a protease inhibitor) treatment, the protein level of IL-4Rα increased in a time-dependent manner, indicating that IL-4Rα is a fast-degrading protein that is mainly degraded through the proteasome pathway, and the effect of BAP31 on IL-4Rα could be blocked by MG132 (Fig. 4D). This suggested that the regulation may be related to the ubiquitination process. The ubiquitination experiment indicated that knockdown of BAP31 was accompanied by reduced ubiquitination of IL-4Rα in BMDMs (Fig. 4E). Therefore, the molecular mechanism underlying how BAP31 regulates IL-4Rα is that BAP31 affects IL-4Rα degradation via the ubiquitination-mediated proteasome-dependent pathway.

These results motivated us to speculate that BAP31 coregulates C/EBPβ through both Egr-2 and IL-4/IL-4Rα pathways. To confirm our speculation, we knocked down or overexpressed Egr-2 and IL-4Rα with different combinations in BMDMs of ^−/− mice. As expected, knockdown of Egr-2 partially blocked the recovery of C/EBPβ expression, while knockdown of both Egr-2 and IL-4Rα completely blocked the recovery of C/EBPβ expression. Simultaneous overexpression of Egr-2 and knockdown of IL-4Rα restored C/EBPβ expression to normal in IL-4–stimulated BMDMs of ^−/− mice (Fig. 4F, Supplemental Fig. 3G). These findings indicate that BAP31 coregulates C/EBPβ by Egr-2 and IL-4/IL-4Rα pathways in IL-4–stimulated macrophages.

Tissue repair is the final stage of wound healing, mediated by IL-4–activated alternatively activated macrophages (30). Angiogenesis, fibroblast regeneration, myofibroblast differentiation, and collagen production are the key processes in the repair of deep wounds (31, 32). Based on the results of BAP31-regulated M2 macrophages via IL-4Rα, we explored whether BAP31 affected angiogenesis during the wound healing process in vivo. We knocked down IL-4Rα in ^−/− mice using the AAV. Because the common AAV vector cannot locate macrophages, we designed and synthesized an AAV plasmid constructed with the macrophages-specific promoter CD68. After verifying that IL-4Rα was knocked down in macrophages of ^−/− mice (Fig. 5A), we divided the mice into four groups: ^+/+ mice,^−/− mice, IL-4–treated ^−/− mice, and IL-4–treated + IL-4Rα^−/− mice. Each group contained five mice. To establish a model of cutaneous injury for each mouse, we needed to inject IL-4 (5 ng/g) near the wound daily. To investigate the angiogenesis of the wound healing, we took the blood vessels in wound sites by Laser Speckle Imaging Instrument (RFLSI III) at day 4 after injury. The number of neovessels in ^−/− mice was markedly reduced compared with ^+/+ mice and returned to normal with IL-4 treatment, but the reduction of blood vessels was not significantly improved in IL-4–treated + IL-4Rα^−/− mice (Fig. 5B). PECAM-1 (CD31) is widely distributed in vascular-associated cells and is related to angiogenesis (33). The blood vessels were identified by CD31 immunohistochemical staining. Quantitative analysis showed that BAP31 knockdown significantly decreased the number of blood vessels and returned to normal after IL-4 treatment, but the reduction of blood vessels was not significantly improved on IL-4Rα knockdown (Fig. 5C, 5D). Similar results were observed in double-immunofluorescence labeling for CD31 and CD206 (Fig. 5E, 5F). In conclusion, BAP31 affects the angiogenesis during the wound healing process via IL-4Rα in vivo.

Wound healing can lead to skin fibrosis and effective accumulation of extracellular matrix components, particularly collagen, which eventually leads to scar formation. Therefore, on the 14th day after injury, immunofluorescence staining was performed by fibrotic markers, α smooth muscle actin (α-SMA) and type 1 collagen α1 (Col1A1), to evaluate the skin fibrosis during the wound healing. The positive immunostaining area of α-SMA (Fig. 6A, 6B) and Col1A1 (Fig. 6C, 6D) was significantly elevated in the ^−/− mice and IL-4–treated + IL-4Rα^−/− mice compared with ^+/+ mice and IL-4–treated ^−/− mice. Western blot analysis demonstrated that α-SMA expression of the cutaneous wound tissue was consistent with the earlier results (Fig. 6F). Furthermore, analysis of Masson’s trichrome showed that collagen deposition increased in ^−/− mice and IL-4–treated + IL-4Rα^−/− mice compared with ^+/+ mice and IL-4–treated^−/− mice (Fig. 6E). In conclusion, BAP31 affects the fibrosis during the wound healing process via IL-4Rα in vivo.

To evaluate the impact of BAP31 on the speed of wound healing, we measured the wound area every 3 d after injury. We observed that the wound healing process was slowed down in ^−/− mice and IL-4–treated + IL-4Rα^−/− mice compared with ^+/+ mice and IL-4–treated ^−/− mice (Fig. 7A, 7B). Western blot analysis demonstrated that Ym1 expression of the cutaneous wound tissue in ^−/− mice and IL-4–treated + IL-4Rα^−/− mice was significantly reduced (Fig. 7C). Morphological analysis using H&E staining of the skin tissue at 14 d after injury indicated that the cutaneous wounds of the ^−/− mice and IL-4–treated + IL-4Rα^−/− mice were not completely repaired compared with the ^+/+ mice and IL-4–treated ^−/− mice (Fig. 7D). These results imply that BAP31 knockdown slows down the wound healing process. In IL-4–treated mice, the speed of wound healing was expedited, but it slowed on IL-4Rα knockdown. Therefore, BAP31 affects the speed of wound healing via IL-4Rα in vivo.

BAP31 is the principal protein expressed in the ER and has an important role in regulating the intracellular transport of multiple proteins, including mIgD, cellubrevin, MHC class I, BCL-2/BCL-X(L), and membrane-associated RING-CH (MARCH) (34, 35). BAP31 regulated MARCH-I and MARCH-VIII on the macrophages surface through ubiquitination to participate in the immune response (36). Moreover, BAP31 also affects the activation and secretion of cytokines by T cells that may influence macrophages activation (3). These findings indicate that BAP31 is likely to affect the function and polarization of macrophages. Previous studies have demonstrated that macrophages can induce immune tolerance to transplantation and promote wound healing (37). During the healing process, abundant apoptotic debris and anti-inflammatory cytokines induce the production of M2 macrophages, which contribute to tissue regeneration and repair, accompanied by angiogenesis and skin fibrosis (38–41). In this study, we reveal that BAP31 deficiency decreased the expression of M2 macrophages markers (CD206, Ym1, Arg-1, and Fizz1), angiogenesis-related genes (VEGFA, PLGF, b-FGF, Ang-1, Ang-2, and PDGF-BB), and fibrosis-related genes (Col1A1, TGF-β, Fibronectin) in BMDMs. These results suggest that BAP31 affects the polarization of M0 macrophages to M2 macrophages. High expression of multiple transcriptional regulators promotes M2 polarization of macrophages. Therefore, one of the explanations for how BAP31 regulates macrophages is that BAP31 is an important transporter regulatory protein that may affect the expression of transcription factors of macrophages. Our further investigations confirmed this speculation and demonstrated that BAP31 knockdown downregulated many transcription factors related to M2 macrophages containing IRFs, C/EBPs, LXRα, and C/EBPβ, and has the most significant inhibitory effect on C/EBPβ expression. In contrast, overexpression of BAP31 could upregulate the levels of C/EBPβ and macrophage-associated genes in RAW264.7. Collectively, the regulation of macrophages by BAP31 is more associated with the expression of C/EBPβ.

C/EBPβ is an important regulatory transcription factor for M2 macrophages polarization. Previous studies found that C/EBPβ regulated the expression of Arg-1 and several other specific genes in the M2 macrophages (42). To characterize the mechanism underlying the regulation of C/EBPβ by BAP31, we performed the IP analysis and found that there was no direct binding between BAP31 and C/EBPβ. Therefore, we considered that BAP31 affects C/EBPβ expression via regulating intermediary proteins. Regulatory factors affecting C/EBPβ transcription include Sp1, CREB, SREBP1c, RARα, Myβ, Fra-2, Egr-2, STAT-6, and NF-κB. We focused on CREB and Egr-2, which are closely related to macrophages. There were two cAMP-like responsive elements (CRE-like sites) in the region close to the TATA box of the C/EBPβ gene, and the PKA/CREB pathway targets these two CRE-binding sites, thereby regulating the transcription of C/EBPβ (43). Egr-2 is a transcription factor of the Egr family; it has been shown to recognize the promotor elements of C/EBPβ, and regulated C/EBPβ may be related to the “leaky ribosome scanning mechanism” (44). Our results revealed that the protein level of Egr-2 decreased significantly after BAP31 knockdown, and restoring the level of Egr-2 resulted in the recovery of the downregulation of C/EBPβ induced by BAP31 deficiency. Therefore, we concluded that the regulation of BAP31 on C/EBPβ was through Egr-2. Because BAP31 knockdown did not affect Egr-2 mRNA level, we consider that BAP31 regulates Egr-2 possibly through posttranscriptional mechanisms. Nevertheless, the molecular mechanisms of this pathway remain to be clarified.

IFN-γ or a combination of IFN-γ and LPS is responsible for the differentiation of classically activated M1 macrophages, while cytokines IL-4, IL-10, and IL-13 were responsible for the differentiation of alternatively activated M2 macrophages (45). The expression of C/EBPβ could increase by activation of macrophages (46). Consequently, we investigated the regulation of BAP31 on C/EBPβ expression in activated macrophages and found that the downregulation of BAP31 on C/EBPβ and M2 macrophages marker Arg-1 was recovered in IL-4–stimulated macrophages. Based on this, we speculated that the changes of the proteins in the IL-4 signaling pathways might counteract the decrease in C/EBPβ induced by BAP31 knockdown. When IL-4 binds to receptor IL-4Rα, JAK1 and JAK3 are activated, leading to the activation of STAT6, which is the classical signaling pathway of IL-4 (47). Based on this, we detected the receptors and adaptor proteins in IL-4 signaling pathways and found that BAP31 knockdown upregulated the expression of IL-4Rα, resulting in the increase of JAK3 and STAT6 phosphorylation. Hyperactivation of the IL-4 pathway can promote the expression of downstream transcription factor C/EBPβ to counterbalance its reduction caused by BAP31 knockdown. Furthermore, IL-4Rα knockdown blocked the recovery of C/EBPβ expression. BAP31 acts as a broad-specificity membrane protein chaperone and quality control factor, which can promote different fates for its clients, including ER retention, ER export, and ER-associated degradation (48). Due to the negative regulation of BAP31 on IL-4Rα protein expression, we speculated that the effect of BAP31 on IL-4Rα is perhaps related to its degradation. We found that IL-4Rα was a fast-degrading protein, and knockdown of BAP31 inhibited the degradation rate of IL-4Rα. Moreover, the expression of IL-4Rα gradually recovered after adding MG132 (a proteasome inhibitor), indicating its degradation via the ubiquitin-proteasome degradation system. In the ubiquitin-proteasome system, proteins tagged by certain types of polyubiquitin chain are selectively recognized and removed by proteasome to result in proteolytic degradation (49). These led us to consider that BAP31 may affect IL-4Rα ubiquitination. The results of IP showed that knockdown of BAP31 inhibited the binding of IL-4Rα with ubiquitin. Therefore, BAP31 regulates IL-4Rα degradation through the ubiquitination pathway. Next, we explored the pattern of BAP31 coregulated C/EBPβ by knockdown and overexpression of Egr-2 and IL-4Rα. We confirmed that BAP31 regulated C/EBPβ via two pathways in IL-4–stimulated macrophages. One is BAP31 deficiency downregulated the C/EBPβ level through Egr-2, and the other is BAP31 deficiency exacerbated the activation of the IL-4 signal pathway through the upregulation of IL-4Rα, which led to the restoration of the downstream protein C/EBPβ. In conclusion, BAP31 regulates the C/EBPβ-mediated polarization homeostasis of M2 macrophages through two direction cycles of mechanisms. When we overexpressed or knocked down the target proteins of the two signaling pathways, the homeostasis was disrupted.

When injury occurs, Th2-related cytokine IL-4 binds to IL-4Rα on the macrophages surface, and high expression of p-STAT6 and C/EBPβ lead to M2 polarization of macrophages (50). M2 macrophages affect the secretion of the cytokines related to angiogenesis and fibrosis and initiate the repair process (30). Based on the regulation of BAP31 on M2 macrophages and the IL-4 signaling pathway in vitro, we study the effect of BAP31 on the functions of M2 macrophages in vivo using macrophage–IL-4Rα knockdown and macrophage-BAP31 knockdown mice with skin injury model. Angiogenesis is a tightly orchestrated process in which proangiogenic and antiangiogenic factors are released to bind to cognate endothelial cell 2 surface receptors (51). PECAM-1/CD31 is a cell adhesion molecule that directly participates in the angiogenic process (52). To investigate wound angiogenesis, we took the blood vessels in the wound site by Laser Speckle Imaging Instrument at day 4 after injury and collected the skin tissue labeled CD31 and CD206. Our results indicated that the inhibition of angiogenesis by BAP31 knockdown was alleviated in IL-4–induced mice and aggravated on BAP31 and IL-4Rα double knockdown. Myofibroblast markers, α-SMA and type 1 collagen α1 (Col1A1), were labeled with skin tissue 14 d after injury to evaluate the process of skin fibrosis (53, 54). We found that the promoting of skin fibrosis by BAP31 knockdown was attenuated in IL-4–induced mice and aggravated on BAP31 and IL-4Rα double knockdown. We monitored the wound size and wound morphology during the wound healing process within 14 d and observed that BAP31 knockdown significantly slows down the speed of wound healing. Besides, the speed of wound healing was expedited in IL-4–induced mice and slowed after BAP31 and IL-4Rα double knockdown. In conclusion, BAP31 regulates wound healing function of macrophages via the IL-4 pathway signaling in vivo.

In summary, these findings demonstrate that BAP31 regulates C/EBPβ through Egr-2 and IL-4Rα and further affects M2 macrophages in vitro and in vivo. The data presented in this article illustrate a new potential function of BAP31 and provide novel targets for the prevention and treatment of chronic wounds.

We thank Shenyang Pharmaceutical University for providing the Laser Speckle Imaging Instrument (RFLSI III).

The authors have no financial conflicts of interest.