RGC32 Promotes Bleomycin-Induced Systemic Sclerosis in a Murine Disease Model by Modulating Classically Activated Macrophage Function Free

Systemic sclerosis (SSc) is a multisystem autoimmune disease that is characterized by fibrosis of the skin and other internal organs, including lung, gastrointestinal tract, heart, and kidney, which is accompanied by abnormalities in innate and adaptive immunity (1, 2). Although the etiology of SSc remains uncertain, several advances have been made in understanding the immune response involved in SSc development, including immunological activation with infiltration of mononuclear immune cells, microvascular endothelium injury by immune cells, and fibroblast activation in organs by proinflammatory cytokines, which leads to excessive deposition of extracellular matrix (3, 4). In particular, recent studies highlighted the presence of macrophages as a prerequisite for SSc, as shown in bleomycin-induced dermal and lung fibrosis in nude, rag-deficient, or SCID mice (5–7).

Macrophages, as the key regulator cells in the immune system, display diverse plasticity and physiology, giving rise to different populations with distinct functions. The classically activated macrophages (M1 macrophages) stimulated by IFN-γ and LPS are a macrophage population with a strong capacity to secrete high levels of proinflammatory cytokines, including IL-1β and inducible NO synthase (iNOS), through the NF-κB signaling pathway (8–10). Mononuclear inflammatory cells and proinflammatory cytokines, such as IL-1β and TNF-α, are significantly higher in the bronchoalveolar lavage fluid and peripheral blood of SSc patients (11). In addition, SSc patients show enhanced NF-κB activity (12) and increased expression of a cluster of IFN-regulated genes (13). Although these findings suggest a potential function of macrophages in SSc, the precise roles of inflammatory macrophages and NF-κB signaling in the pathogenesis of SSc remain poorly defined. In contrast, alternatively activated macrophages (M2 macrophages) are stimulated by IL-4 to promote arginase activity (8). Abundant M2 macrophages are observed in SSc patient skin (14), suggesting that they also play a role in the development of SSc.

Response gene to complement 32 (RGC32) is a cell cycle regulator that is involved in cell cycle regulation, cell migration, and differentiation (15–17). We have previously reported that RGC32-deficient macrophages exhibit decreased phagocytosis (18). RGC32 expression is upregulated in CSF (M-CSF) and/or IL-4–dependent tumor-associated macrophages (19). In addition, by interacting with Smad3, RGC32 promotes fibroblast activation in renal tubulointerstitial fibrosis (20). However, the regulatory role of RGC32 in M1 macrophage differentiation and inflammatory response in SSc development remains to be determined.

In the current study, we found that RGC32 played a critical role in the bleomycin-induced onset of SSc. RGC32 deficiency significantly ameliorated skin and lung sclerosis. RGC32 mediated SSc development primarily by promoting the polarization of M1 macrophages and the inflammatory response by interacting with NF-κB and enhancing iNOS and IL-1β gene expression via binding to their promoters.

Male and female mice were used in the studies. RGC32-deficient (RGC32^−/−) mice on the C57BL/6 background were generated and genotyped as described previously (21). Wild-type (WT) littermates were used as controls. All animals were housed in compliance with the Principles of Laboratory Animal Care. Animal surgical procedures were approved by the Institutional Animal Care and Use Committee of the University of Georgia.

Bleomycin was purchased from Thermo Fisher Scientific. The following Abs were used in Western blot and immunofluorescent staining. RGC32 polyclonal Ab was produced by Proteintech Group (Chicago, IL) (16). Collagen type I α 1 (COL1A1) (D-13) and Lamin B (C-20) were obtained from Santa Cruz Biotechnology. iNOS (4E5), Arginase (4E6), and CD3 were purchased from Abcam. IL-1β (3A6), NF-κB p65 (D14E12), phospho-NF-κB p65 (Ser⁵³⁶), IκB (44D4), and phospho-IκB (Ser³²) were from Cell Signaling Technology. GAPDH was from Proteintech, and F4/80 (BM8) was from BioLegend. Nuclei were stained with DAPI (Vector Laboratories). The secondary Abs were from Cell Signaling Technology. M-CSF and IFN-γ were purchased from R&D Systems. M-CSF was used at 10 ng/ml, and IFN-γ was used at 100 ng/ml. LPS was obtained from Sigma-Aldrich (St. Louis, MO) and used at 100 ng/ml.

To induce skin fibrosis, bleomycin (0.02 U) dissolved in PBS was injected s.c. into a single location on the back of mice daily for 28 d. To induce pulmonary fibrosis, bleomycin (0.2 U) was applied intranasally once, and the mice were euthanized 24 d later. PBS was used as control in both models. The skin or lung tissues were isolated for further analyses.

Mouse peritoneal macrophages (PEMs) were obtained from the peritoneum by PBS flushing and were cultured in DMEM, as previously described (22). Briefly, peritoneal cells were flushed out from the peritoneal cavity using cold PBS and cultured in DMEM supplemented with 10% heat-inactivated FBS in a humidified CO₂ incubator at 37°C for 2 h. Adherent cells, consisting of >90% macrophages, were harvested.

Bone marrow–derived macrophages (BMDMs) were generated from bone marrow by M-CSF induction, as described previously (22). Bone marrow was aseptically flushed from the tibiae and femurs of euthanized mice and depleted of RBCs using Red Blood Cell Lysis Buffer (Roche). Cells were incubated in DMEM in a cell culture dish at 37°C for 2 h to remove macrophages. Nonadherent cells were resuspended in DMEM supplemented with 10% heat-inactivated FBS, 100 U/ml penicillin, 100 mg/ml streptomycin, 2 mM L-Glutamine (Thermo Fisher Scientific), and 10 ng/ml M-CSF and cultured for 7 d. Nonadherent cells were removed, and M-CSF–conditioned medium was changed on days 3 and 5. To acquire M1 macrophages, macrophages were stimulated with 100 ng/ml IFN-γ and 100 ng/ml LPS for 3 h for mRNA analysis or for 6 h for protein assays.

Bone marrow cells (BMCs) from 8-wk-old RGC32^−/− mice and WT littermates were prepared. A total of 1 × 10⁷ cells was injected into the tail vein of lethally irradiated 8-wk-old WT mice to generate full chimeras, as described (23). Eight weeks after reconstitution, recipient mice were treated with bleomycin or PBS. Mice were euthanized at the indicated time points, and the skin or lung samples were isolated for further analyses.

Skin and lung tissues were fixed in 4% paraformaldehyde and embedded in paraffin. Tissue sections (5 μm thick) were stained with H&E or Masson’s trichrome, using commercial kits (Dako) and following standard procedures, for histopathological analyses. For immunofluorescent staining, serial sections (10 μm) from OCT-embedded frozen tissues or primary cultured cells were fixed in cold acetone or 4% paraformaldehyde. After blocking with 1% goat serum, sections were incubated at room temperature with primary Abs for 2 h and then with fluorescent dye–conjugated secondary Abs for 1 h. Images were acquired with a fluorescence microscope (Nikon Instruments).

TRIzol Reagent (Invitrogen) was used to extract total RNA, following the manufacturer’s instructions. cDNA was synthesized using an iScript cDNA synthesis kit (Bio-Rad). RT-PCR was performed on a Bio-Rad C1000 Thermal Cycler. Quantitative PCR (qPCR) was performed on an MX3000P qPCR machine using SYBR Green QPCR Master Mix (Agilent). The primers used were described previously (18).

PEMs, BMDMs, or skin and lung tissues were lysed in RIPA lysis buffer (1% Nonidet P-40, 0.1% SDS, 0.5% sodium deoxycholate, 1 mM sodium orthovanadate, and protease inhibitors) to extract the total proteins. Samples were separated on SDS-polyacrylamide gels and electrotransferred onto nitrocellulose membranes (Amersham Biosciences). After blocking with 5% BSA, the membranes were incubated with various primary Abs at 4°C overnight and then incubated with IRDye secondary Abs (LI-COR Biosciences) at room temperature for 1 h. Protein expression was detected using an Odyssey CLx Infrared Imaging System (LI-COR Biosciences).

Protein A/G PLUS-Agarose beads (Santa Cruz Biotechnology) were incubated with IgG, RGC32, or NF-κB Ab at 4°C for 2 h. BMDMs were lysed in 500 μl of coimmunoprecipitation (co-IP) lysis buffer (Pierce) on ice for 5 min. Following rapid centrifugation, the supernatants were incubated with beads at 4°C overnight. After washing with co-IP buffer, proteins were eluted from the beads and boiled in SDS loading buffer. Western blotting was performed to detect the precipitation of proteins.

Chromatin immunoprecipitation (ChIP) assays were performed using a ChIP kit (Millipore), according to the manufacturer’s protocol. BMDMs were treated with IFN-γ and LPS for 30 min. Chromatin complexes were immunoprecipitated with 3 μg of IgG (negative control), NF-κB, or RGC32 Abs. Semiquantitative PCR was performed to amplify the iNOS and IL-1β promoter regions containing the NF-κB binding site. The primers for iNOS promoter were 5′-AAA GGA GAA ACA GCC ACC AAG C-3′ (forward) and 5′-AGC ACC CAC AAC CCA AAG AAC-3′ (reverse). The primers for IL-1β promoter were 5′-TCC CTG GAA GTC AAG GGG TGG-3′ (forward) and 5′-TCT GGG TGT GCA TCT ACG TGC C-3′ (reverse).

All data are presented as mean + SD. An unpaired Student t test or one-way ANOVA, followed by the Fisher t test for comparison of means, was used to compare groups. A p value < 0.05 is considered statistically significant.

Bleomycin-induced skin fibrosis in mice was used to study SSc (24). RGC32 expression was significantly upregulated at the mRNA (Fig. 1A) and protein (Fig. 1B) levels in skin tissues, along with SSc progression. Remarkably, a substantial increase in RGC32 expression was observed at the initial stage following 1 d of bleomycin injection (Fig. 1A, 1B), suggesting a role for RGC32 in SSc development.

One hallmark of skin fibrosis is the thickened dermis that results from collagen deposition. Bleomycin caused significant skin thickening 28 d after the injection, as measured by ultrasonography; however, RGC32 deletion diminished the skin thickening (Fig. 1C, 1D). Histopathological analyses of skin sections confirmed that the bleomycin-induced increased skin thickness seen in WT mice was significantly reduced in RGC32^−/− mice (Fig. 1E, left panel). Because fibrosis is characterized by excessive collagen deposition, we detected collagen content using Masson’s trichrome staining. Bleomycin induced a large amount of collagen deposition in the skin tissues of WT mice, but it was attenuated in RGC32^−/− mice (Fig. 1E, right panel). Consistently, the expression of collagen protein COL1A1 was also decreased considerably in bleomycin-treated RGC32^−/− mice (Fig. 1F).

Pulmonary lesions in patients with SSc are strongly associated with mortality (25); thus, we sought to determine whether RGC32 plays a role in lung sclerosis. Lung fibrosis was induced by bleomycin in WT and RGC32^−/− mice. Bleomycin treatment significantly impaired the lung structure, as analyzed by H&E staining (Fig. 1G). However, the bleomycin-induced lung damage was remarkably alleviated in RGC32^−/− mice (Fig. 1G). Furthermore, bleomycin-treated RGC32^−/− mice exhibited a marked reduction in collagen deposition and expression in the lung compared with WT mice (Fig. 1G, right panel, 1H). These results further demonstrated that RGC32 played a critical role in the development of SSc.

Because RGC32 is abundantly expressed during the initial stage of SSc, it may regulate early events of SSc. Macrophage infiltration is one of the essential factors initiating SSc (26). In bleomycin-induced SSs, we observed a rapid appearance of macrophages in skin, as evidenced by the increase in macrophage marker F4/80 (Fig. 2A). Notably, the highest F4/80 level was observed at day 1 after bleomycin injection, concomitant with RGC32 expression (Fig. 1A). Thus, we supposed that RGC32 was expressed in macrophages. Indeed, RGC32 was costained with F4/80 (Fig. 2B) but not CD3 (Supplemental Fig. 1A), indicating that RGC32 was primarily expressed in macrophages during the initiation phase of skin sclerosis. Moreover, RGC32^−/− mice exhibited decreased macrophage infiltration in bleomycin-treated skin, as shown by flow cytometry analysis (Fig. 2C, 2D), suggesting that RGC32 promoted bleomycin-induced SSc by modulating macrophage function.

Inflammation is known to be a critical factor driving the pathogenesis of SSc (27). Therefore, we investigated the inflammatory response in bleomycin-induced skin by analyzing the expression of iNOS and IL-1β. Bleomycin treatment markedly increased the expression of iNOS and IL-1β mRNAs (Supplemental Fig. 1B) and proteins (Fig. 2E–G) in WT mouse skin tissues. Immunostaining of skin sections showed that iNOS (Fig. 2F) and IL-1β (Fig. 2G) were colocalized with F4/80, indicating that these inflammatory mediators were mainly expressed by F4/80⁺ macrophages. Notably, bleomycin-treated RGC32^−/− skin showed reduced inflammation compared with WT skin, as evidenced by the attenuated expression of iNOS and IL-1β at the mRNA (Supplemental Fig. 1B) and protein (Fig. 2E) levels. Reduced iNOS and IL-1β expression coincided with the lower number of F4/80⁺ macrophages in RGC32^−/− mouse skin (Fig. 2F, 2G).

In addition to inflammatory macrophages, M2 macrophages were reported to play an important role in perpetuating Ssc (13). Therefore, we tested whether RGC32 affects M2 macrophage content in Ssc. Treatment with bleomycin for 28 d caused an accumulation of numerous arginase⁺ cells in WT mouse skin (Fig. 2H), suggesting an increased M2 macrophage infiltration. However, RGC32 deficiency significantly attenuated the presence of arginase⁺ cells (Fig. 2H). The reduction in arginase⁺ cells coincided with decreased F4/80⁺ macrophages in RGC32^−/− mouse skin, which further indicated that RGC32 promoted the infiltration of M2 macrophages (Fig. 2H). The critical role for RGC32 in arginase expression was also confirmed by Western blot analyses (Fig. 2I), consistent with our previous studies showing that RGC32 promotes M2 macrophage polarization in vitro (18). These results suggested that RGC32 regulated M1 and M2 macrophage function in the development of SSc.

To determine whether RGC32 in macrophages is essential for bleomycin-induced skin fibrosis, BMCs from WT or RGC32^−/− mice were adoptively transferred into lethally irradiated WT mice to establish full bone marrow chimeras. Eight weeks after transfer, recipient mice were treated with bleomycin, and skin fibrosis was assessed. Similar to WT mice, bleomycin treatment of mice receiving WT BMCs caused intensive skin sclerosis, as reflected by the thickened dermis (Fig. 3A, left panel) and exacerbated collagen deposition and expression (Fig. 3A, right panel, 3B). However, mice receiving RGC32^−/− BMCs exhibited significantly thinner skin (Fig. 3A, left panel) and a marked reduction in collagen deposition under bleomycin treatment (Fig. 3A, 3B). Moreover, mice receiving RGC32^−/− BMCs showed an alleviated inflammatory response in the skin after bleomycin treatment because the iNOS and IL-1β expression was significantly decreased compared with mice receiving WT BMCs (Fig. 3C). These results indicated that macrophage RGC32-mediated inflammation is essential for SSc development.

The role of macrophage RGC32 in lung impairment was also assessed in bleomycin-induced lung fibrosis in full bone marrow chimeras. The lung fibrosis induced by bleomycin in mice receiving RGC32^−/− BMCs was significantly attenuated compared with mice receiving WT BMCs, as shown by the improved lung morphology (Fig. 3D, left panel) and the decreased collagen deposition (Fig. 3D, right panel, 3E). These data demonstrated that RGC32 also promoted lung sclerosis via the modulation of macrophage function.

Because RGC32 regulated M1 macrophage inflammation in SSc, we further studied its role in inflammatory macrophage differentiation using PEMs and BMDMs. RGC32 was highly expressed in PEMs and M-CSF–induced BMDMs (Supplemental Fig. 1C, 1D). Importantly, RGC32 protein levels were significantly elevated in PEMs upon stimulation with IFN-γ and LPS, factors that are well known for inducing M1 macrophages (Fig. 4A). Similar results were observed in IFN-γ + LPS–treated BMDMs (Fig. 4B).

To determine how RGC32 is regulated in macrophages, pathway-specific inhibitors were used to identify signaling pathways that mediated the induction of RGC32 in M1 macrophages. Inhibition of the p38 MAPK activity did not affect RGC32 mRNA or protein expression (Supplemental Fig. 1E–G) in BMDMs. However, blockade of PI3K, ERK, and JNK significantly inhibited RGC32 expression (Supplemental Fig. 1E–G), suggesting that the PI3K, ERK, and JNK signaling pathways regulated RGC32 expression in M1 macrophages.

To determine whether RGC32 is critical for the proinflammatory function of M1 macrophages we compared iNOS and IL-1β expression in classically activated PEMs and BMDMs in WT and RGC32^−/− mice. IFN-γ and LPS stimulation induced a remarkable increase in iNOS and IL-1β protein expression in WT, but not RGC32^−/−, PEMs (Fig. 4C, 4D). Consistently, iNOS and IL-1β expression was markedly decreased in RGC32^−/− BMDMs compared with WT BMDMs stimulated with IFN-γ and LPS (Fig. 4E, 4F). These results supported that RGC32 promoted the polarization of classically activated inflammatory macrophages to initiate SSc.

Previous studies have shown that NF-κB activation promotes M1 macrophage polarization (28). We speculated that RGC32 activated M1 macrophages by modulating the NF-κB signaling pathway; therefore, we assessed phosphorylation of IκB and NF-κB (p65) in bleomycin-induced SSc. In agreement with the expression of iNOS and IL-1β (Fig. 2E–G), bleomycin significantly augmented the phosphorylation of IκB and NF-κB in WT, but not RGC32^−/−, skin (Fig. 5A, 5B).

The involvement of NF-κB signaling in the RGC32-promoted classical activation of macrophages was assessed in PEMs and BMDMs stimulated with IFN-γ + LPS. Although phosphorylation of IκB and NF-κB was induced by IFN-γ + LPS in WT PEMs (Fig. 5C, 5D) and BMDMs (Fig. 5E, 5F), RGC32 deficiency markedly blocked phosphorylation in both cells (Fig. 5C, 5F). Because NF-κB is sequestered in an inactive state in the cytoplasm by IκB and is translocated into the nuclei upon activation to regulate target gene expression (29), we compared the cytoplasmic and nuclear levels of NF-κB in WT and RGC32^−/− BMDMs upon IFN-γ + LPS induction. IFN-γ + LPS–induced NF-κB nuclear translocation was significantly impaired in RGC32-deficient BMDMs, as evidenced by the lower level of nuclear NF-κB and the relatively higher level of cytoplasmic NF-κB (Fig. 5G–I) compared with WT BMDMs. These data indicated that RGC32 regulated the inflammatory response of macrophages during the early stage of SSc via the classical NF-κB signaling pathway.

In macrophages, NF-κB activates iNOS and IL-1β expression by direct binding to their promoters as a transcriptional factor (30, 31). Therefore, RGC32 may regulate iNOS and IL-1β expression in M1 macrophages by cooperation with NF-κB. Co-IP assays were performed to test the direct interaction between endogenous RGC32 and NF-κB in macrophages. In classically activated WT BMDMs, NF-κB was pulled down along with RGC32 (Fig. 6A, 6B) and vice versa (Fig. 6C, 6D), indicating a strong interaction between RGC32 and NF-κB in M1 macrophages. The interaction between RGC32 and NF-κB was minimal in undifferentiated BMDMs.

Because RGC32 interacted with NF-κB, and both are critical for mRNA expression of the inflammatory mediators iNOS and IL-1β (30, 31) (Supplemental Fig 1B), we performed ChIP assays to test whether RGC32 binds to NF-κB binding regions in iNOS and IL-1β promoters. As shown in Fig. 6E and 6F, RGC32 indeed bound to the same promoter regions of iNOS and IL-1β as NF-κB. Due to the lack of a DNA-binding domain, we tested whether NF-κB is required for RGC32 binding to iNOS and IL-1β promoters. Blockade of NF-κB nuclear translocation by ammonium pyrrolidine dithiocarbamate (PDTC) diminished RGC32 binding to the promoters and inhibited RGC32-enhanced iNOS and IL-1β expression in macrophages (Fig. 6G, 6H), demonstrating that RGC32 promoted the inflammatory response in M1 macrophages through NF-κB signaling.

SSc is a chronic autoimmune disorder that is characterized by diffuse fibrosis in the skin, joints, and internal organs (e.g., lungs and kidneys) (32, 33). It is widely accepted that the vascular endothelial activation and excessive deposition of extracellular matrix in SSs are highly associated with the inflammatory abnormalities (34). Previous studies showed that RGC32 is involved in renal fibrosis by mediating fibroblast activation (20). By using RGC32^−/− mice, we found that RGC32 deficiency significantly ameliorated the bleomycin-induced skin and lung sclerosis, as evidenced by the reduction in dermis thickness and collagen deposition, demonstrating that RGC32 is an essential mediator of SSc. Because the notable expression of RGC32 is induced at the very early stage of SSc, the major function of RGC32 in the onset of SSc is likely to be independent of fibroblasts that are activated and function during the late stage of SSc (35, 36).

Macrophages and T cells are involved in the onset of SSc (37–42). Macrophage accumulation was revealed at the early stage of bleomycin-induced skin sclerosis. RGC32 was abundantly expressed in macrophages, but not in T cells, during the early stage. RGC32 appeared to play an essential role in inducing macrophage infiltration and in the inflammatory response of skin fibrosis, because RGC32 deficiency reduced macrophage accumulation and markedly diminished the expression of two potent fibrosis inducers (iNOS and IL-1β) in skin tissues (43–46). The importance of macrophage RGC32 in SSc was also supported by the outcome in full bone marrow chimeric mice; RGC32 deficiency in hematopoietic macrophages attenuated skin and lung sclerosis.

RGC32 regulated the proinflammatory response of M1 macrophages. IFN-γ and LPS stimulation of PEMs and BMDMs induced RGC32 expression through the PI3K, ERK, and JNK signaling pathways. RGC32 deficiency significantly impaired the IFN-γ and LPS–stimulated increase in the levels of iNOS and IL-1β in PEMs and BMDMs. A previous article showed that LPS decreased RGC32 mRNA expression in macrophages (19). The discrepancy is likely due to the different treatment times. In the previous study, macrophages were treated with IFN-γ and LPS for 42 h, which might cause a negative-feedback mechanism, resulting in a reduction in the mRNA levels of RGC32. Our results clearly showed that increased expression of RGC32 was critical for the polarization and inflammatory response of M1 macrophages.

RGC32 interacted with NF-κB signaling to mediate M1 macrophage polarization (9, 10, 47) and skin fibrosis (48–51). Activation of NF-κB signaling was significantly inhibited in bleomycin-treated RGC32^−/− skin tissues, as well as in IFN-γ and LPS–treated RGC32^−/− PEMs and BMDMs. In contrast, RGC32 promoted iNOS and IL-1β expression in an NF-κB–dependent manner, because blockade of NF-κB nuclear translocation diminished RGC32-mediated iNOS and IL-1β expression. It appeared that RGC32 regulated iNOS and IL-1β expression by binding to their promoters (30, 31) via interaction with NF-κB.

Our previous studies showed that RGC32 deficiency inhibited arginase expression by IL-4–stimulated canonical M2 macrophages (18). Arginase expression was increased in WT mouse skin tissues but was attenuated in RGC32^−/− skin, suggesting that RGC32 was also important for M2 macrophage function in the development of Ssc. Thus, the marked reduction in skin fibrosis in RGC32^−/− mice was likely due to impaired M1 macrophage function during the early stage and impaired M2 macrophage function during the late stage of skin sclerosis.

Taken together, our studies revealed a crucial role for RGC32 in the development of SSc. RGC32 promoted M1 macrophage polarization and regulated the expression of inflammatory mediators through direct interaction with NF-κB. Therefore, targeting RGC32 is a novel therapeutic strategy in treating sclerosis.

We thank Dr. Wendy Watford for technical assistance with bone marrow transplantation.

The authors have no financial conflicts of interest.