Linagliptin ameliorates pulmonary fibrosis in systemic sclerosis mouse model via inhibition of endothelial-to-mesenchymal transition

Systemic sclerosis (SSc) is a connective tissue disease that often causes pulmonary fibrosis. Dipeptidyl peptidase 4 (DPP4) inhibitor has shown anti-fibrotic properties in various fibrotic diseases. However, only two studies have reported its anti-fibrosis effects in pulmonary fibrosis, and the mechanism is not completely clear. In the present study, we further investigated the protective effects of linagliptin, a highly specific DPP4 inhibitor, on pulmonary fibrosis in SSc mouse model and the potential mechanisms. The results showed that linagliptin ameliorated pulmonary fibrosis in SSc mouse model, as evidenced by improved pathological changes of lung and body weight loss induced by BLM. Linagliptin also reduced BLM-induced oxidative stress, inflammation in lung in vivo. We revealed that linagliptin attenuated BLM-induced endothelial-to-mesenchymal transition (EndMT) in vitro and in vivo. BLM-induced enhanced migration ability of endothelial cells was also alleviated by linagliptin. Moreover, we confirmed that the Akt/mammalian target of rapamycin pathway was involved in BLM-induced EndMT in vivo, which was suppressed by linagliptin. In summary, we further confirmed the therapeutic effects of linagliptin on pulmonary fibrosis in SSc mouse model, which is based on its inhibitory effects on EndMT, oxidative stress, and inflammation.


Introduction
Systemic sclerosis (SSc), a rare autoimmune-mediated connective tissue disease, often causes progressive pulmonary fibrosis [1], which causes a high mortality [2]. A previous study has reported that pulmonary fibrosis or interstitial lung disease is present in 80% of patients with SSc [3]. However, treatment for SSc-related pulmonary fibrosis is limited. Therefore, development of novel therapeutic approaches is an urgent need to improve the clinical benefit and quality of life of patients with SSc-related pulmonary fibrosis.
The pathogenesis of SSc-related pulmonary fibrosis is complicated and poor understanding. Myofibroblasts, a unique population of mesenchymal cells, play a pivotal role in the pathogenesis of fibrotic diseases. The cells display a marked profibrotic cellular phenotype including an increased production of fibrillar type l and type lll collagen, initiation of the expression of α-smooth muscle actin (α-SMA) and reduced expression of genes encoding extracellular matrix degradative enzymes [4]. Zeisberg EM et al. reported that in cardiac fibrosis, a part of myofibroblasts were derived from endothelial cells via endothelial-to-mesenchymal transition (EndMT) [5], which is a transdifferentiation process first found in the embryonic developmental process. Subsequently, the process has been demonstrated to be involved in the progression of various fibrotic diseases [6], as well as SSc-related pulmonary fibrosis [7]. Given the critical role of myofibroblasts in the pathogenesis of SSc, treatments targeting EndMT are promising therapies.
Dipeptidyl peptidase 4 (DPP4), also known as CD26, is a serine protease and a type II transmembrane glycoprotein expressed in various cell types including endothelial cells [8]. DPP4 inactivates incretin hormones such as glucagonlike peptides and glucose-dependent insulinotropic peptide. DPP4 inhibitor has been widely used for the treatment of type 2 diabetes because of its activity against DPP4 [9]. DPP4 has also been found to be involved in fibrotic diseases. Previous investigations have demonstrated that the DPP4 1 3 expression and the number of DPP4-positive fibroblasts were increased in fibrotic skin of SSc and keloid patients and that DPP4-positive fibroblasts expressed higher levels of myofibroblast markers and collagen, indicating that DPP4 may be a marker of myofibroblasts [10,11]. Moreover, two studies have reported the therapeutic effects of DPP4 inhibitor on pulmonary fibrosis by inactivating fibroblasts, regulating extracellular matrix and anti-inflammatory effects in mice induced by intratracheal administration of bleomycin (BLM) [10,12]. However, relevant research is still lacking. Therefore, further elucidation of the mechanisms of its anti-fibrotic activity will contribute to its application in the treatment of SSc-related pulmonary fibrosis.
This study aimed to investigate the effects of linagliptin, a highly specific DPP4 inhibitor, on pulmonary fibrosis in SSc mouse model and potential underlying mechanisms.

Reagents
BLM, linagliptin, and rapamycin were purchased from Selleck (Houston, TX, USA). Primary antibodies against CD31, VE-cadherin, and α-smooth muscle actin (α-SMA) were purchased from Abcam (Cambridge, UK). Primary antibodies against snail, slug, twist, Akt, phosphorylated Akt, mammalian target of rapamycin (mTOR), and phosphorylated mTOR were purchased from cell signaling technology (Danvers, MA, USA). Primary antibodies against type I collagen were obtained from Wanleibio (Shenyang, China). Primary antibodies against β-actin were obtained from Proteintech (Wuhan, China). 32 6-8-week-old female C57BL/6 mice, weighing 20-22 g, were obtained from the Laboratory Animal Center of Anhui Medical University (Hefei, China) and housed in a SPF room with controlled temperature and humidity. All mice were supplied with a standard rodent diet and free water.The back of the mice was shaved, and then the mice were randomly divided into four groups (n = 8 each): control, BLM, BLM + linagliptin, and BLM + rapamycin. Rapamycin was used as a positive control for the involvement of the Akt/ mTOR pathway in both the BLM-induced SSc mouse model [13] and the BLM-induced EndMT in human umbilical vein endothelial cells (HUVECs) [14]. The mice in the control group were subcutaneously injected with 0.1 mL of phosphate-buffered saline (PBS) with a 4.5-gauge needle. The mice in the other groups were injected with 0.1 mL of BLM solution (1 mg/mL) daily for 4 weeks. After the first week of injections, the mice in the control, BLM + linagliptin, and BLM + rapamycin groups were orally administrated daily with PBS, linagliptin (10 mg/kg), and rapamycin (1.5 mg/ kg), respectively. All mice were anesthetized and sacrificed on day 28. All experiments were approved by the Animal Care and Use Committee of Anhui Medical University (Hefei, China).

Measurement of tumor necrosis factor (TNF)-α and interleukin (IL)-6 in bronchoalveolar lavage fluid (BALF)
BALF was harvested and centrifuged at 1000×g for 10 min. Then, the supernatant was collected and stored at − 80 °C for further analysis. The levels of TNF-α and IL-6 were determined using TNF-α enzyme-linked immunosorbent assay (ELISA) kits (PI326, Beyotime Biotechnology, Shanghai, China) and IL-6 ELISA kits (PT512, Beyotime Biotechnology) in accordance with the manufacturer's instructions.

Cell culture
HUVECs were purchased from Sciencell Research Laboratories (San Diego, CA, USA) and cultured in endothelial culture medium supplemented with 5% fetal bovine serum, 1% endothelial cell growth supplement, and 1% penicillin-streptomycin solution. The cells were maintained in an incubator containing 5% CO2 at 37 °C. Cells were used in passages 2-7. The cells were seeded in 6-well plates and treated as follows: control, BLM (2 μg/mL), BLM (2 μg/mL) + linagliptin (100 nM), or BLM (2 μg/mL) + rapamycin (200 nM). After 48 h of treatment, the cells were harvested for further analysis. An inverted microscope (CKX53, Olympus, Tokyo, Japan) was used to observe and photograph changes in cell morphology.

Scratch assay and migration assay
For scratch assay, cells were seeded in 6-well plates. A scratch was evenly created by 200 μl pipette tip crossing the surface of the cell monolayer and detached cells were washed off. Then cells were incubated with serum-free 1 3 medium containing indicated drugs. Images were taken at different times. For migration assay, cells were suspended with serum-free medium and then 200 ul cell suspension with indicated drugs was plated in Matrigel (BD Biosciences) pre-coated transwell chamber with 8-um pore. The chamber was then placed into 12-well plate containing 800 ul culture medium supplemented with 5% fetal bovine serum and cultured. After 24 h of migration, the migrated cells were fixed with 4% paraformaldehyde, stained with crystal violet, and photographed.

Histological analysis and immunohistochemistry
Lungs were fixed in 4% paraformaldehyde, embedded in paraffin, cut into 4 μm-thick slices, and then subjected to Masson's trichrome staining and hematoxylin and eosin (H&E) staining. For immunohistochemical staining, after the slices were dewaxed and rehydrated, they were incubated with citrate buffer for antigen retrieval. Endogenous peroxidase activity and non-specific binding were blocked with 3% hydrogen peroxide and 5% goat serum, respectively. Then, the slices were incubated with primary antibodies against CD31 (1:100), α-SMA (1:200) or slug (1:200) at 4 °C overnight, followed by secondary antibodies at 37 °C for 1 h. The stained slices were visualized with diaminobenzidine, counterstained with hematoxylin and observed by bright field imaging using a fluorescence microscope (DM6 B; Leica, Germany).

Immunofluorescence analysis
For lung tissues, immunofluorescence, after deparaffinization, antigen retrieval, and blockade, the slices were incubated with CD31 (1:100) and α-SMA (1:200) antibodies at 4 °C overnight, followed by incubation with Alexa 488 Goat Anti-rabbit IgG and Alexa 647 Goat Anti-mouse IgG at room temperature for 1 h. For in vitro immunofluorescence, HUVECs were seeded on glass coverslips in 24-well culture plates and treated as indicated. The cells were washed and fixed with 4% paraformaldehyde, permeabilized with 0.3% Triton X-100, and then blocked with 5% bovine serum albumin at room temperature. The fixed cells were incubated with primary antibodies against VE-Cadherin (1:200) and α-SMA (1:200), or Slug (1:200) at 4 °C overnight, followed by incubation with the appropriate secondary antibodies at room temperature for 1 h. 4′,6-diamidino-2-phenylindole (DAPI) was used to stain the nuclei. All samples were visualized and photographed using a laser scanning confocal microscope (LSM 880; Carl Zeiss, Oberkochen, Germany) or a fluorescence microscope (DM6 B; Leica).

Real-time quantitative PCR
Total RNA of HUVECs was isolated using Trizol (Invitrogen, Carlsbad, CA, USA) and then reverse transcribed to cDNA using a TaKaRa PrimeScript™RT Master Mix kit (Shiga, Japan). TaKaRa TB GREEN Premix Ex Taq (Kyoto, Japan) and Applied Biosystems QuantStudio6 Flex real-time PCR system (Thermo Fisher Scientific, Waltham, MA, USA) were used for real-time quantitative PCR (RT-qPCR). β-actin was used as an internal control. The primers were designed and synthesized by General Biosystems (Chuzhou, China), and the sequences are shown in Table 1. Gene expression levels were normalized to β-actin, and relative fold changes in mRNA expression were determined using the 2-ΔΔCt method.

Statistical analysis
The data are presented as mean ± standard deviation. Oneway analysis of variance was used to analyze the statistical differences between groups, followed by the post hoc least significant difference test. P < 0.05 was considered statistically significant. All analyses were conducted using SPSS version 22.0. GraphPad Prism 6 software was used to generate the graphs.

Linagliptin ameliorated pulmonary fibrosis in SSc mouse model
All the mice were sacrificed on day 28 and lungs were harvested for further analysis. The BLM-treated mice showed a significant body weight loss compared with the control mice, whereas the linagliptin or rapamycin interventions reversed this trend (Fig. 1a). H&E staining of lung sections showed that subcutaneous injection of BLM resulted in destruction of normal lung structures and marked thickening of alveolar septa. Masson's trichrome staining showed a significant increase in collagen deposition in the lungs of BLM-treated mice. However, both linagliptin and rapamycin significantly alleviated these pathological changes and reduced collagen production (Fig. 1d, e). Consistently, linagliptin or rapamycin treatments reduced the BLM-induced increased levels of hydroxyproline, a marker of collagen protein in lungs (Fig. 1c). In the Ashcroft Fibrosis Scale, the linagliptin and rapamycin groups had lower scores than those of the BLM group (Fig. 1b).

Linagliptin mitigated BLM-induced inflammation and oxidative stress
To evaluate the effect of linagliptin on BLM-induced pulmonary inflammatory responses, the levels of TNF-α and IL-6 in the BALF were determined. We found that BLM administration induced increased levels of TNF-α and IL-6 in the BALF, and that these levels were reduced by linagliptin or rapamycin treatments (Fig. 2a, b). Considering the important role of oxidative stress in the induction of pulmonary fibrosis in SSc [15], we measured the MDA content and SOD activity in the lungs, both of which are accepted oxidative stress markers [16]. As shown in Fig. 2c, d, the MDA content was increased and the activity of SOD was significantly decreased in the lungs of BLM-treated mice. In contrast, linagliptin or rapamycin treatment remarkably suppressed the BLM-induced alterations in SOD and MDA.

Linagliptin attenuated BLM-induced EndMT in vitro
HUVECs are commonly used in in vitro studies related to endothelial cell function, including EndMT in pulmonary fibrosis [17,18]. In the present study, BLM treatment for 48 h induced HUVECs to transform from a cobblestonelike shape to an elongated, spindle shape, but this was attenuated by linagliptin or rapamycin treatments (Fig. 3a). At the molecular level, Western blotting results revealed that BLM induced decreased protein levels of endothelial markers (CD31 and VE-cadherin) and increased levels of mesenchymal marker (α-SMA and collagen I) in HUVECs. However, these alterations were suppressed by linagliptin or rapamycin treatments ( Fig. 3b-f). Immunofluorescence analysis further confirmed these results (Fig. 3k).
In accordance with the Western blotting and immunofluorescence analysis, RT-qPCR analysis showed that BLM upregulated the mRNA expression of CD31 and VE-Cadherin and downregulated the mRNA expression of α-SMA and Collagen I, and that these effects were inhibited by linagliptin or rapamycin treatment (Fig. 3g-j).

Linagliptin inhibited BLM-induced enhanced migration ability
HUVECs undergoing EndMT lose cell-cell adhesion and acquired an enhanced migration ability [14]. Scratch and migration assay were performed to evaluate the effects of linagliptin on migration ability of HUVECs. The results showed that BLM treatment significantly enhanced the migration ability of HUVECs, which was attenuated by linagliptin or rapamycin treatment (Fig. 4a, b).

Linagliptin attenuated BLM-induced EndMT in vivo
CD31/α-SMA double-labeling immunofluorescence was used to evaluate EndMT in vivo. Co-localization of CD31/α-SMA in the endothelium indicates the intermediate stages of EndMT. The BLM group had more double-positive cells compared with the control group; however, this elevated percentage of double-positive cells in the BLM-treated group was reversed in the linagliptin or rapamycin-treated mice (Fig. 5a). Immunohistochemistry analysis showed that the expression of CD31, which is mainly localized at the intima of the capillary vessels, was decreased in the BLM group compared with that in the control group, whereas elevated levels of α-SMA-positive cells were observed in the BLM group compared with the control group (Fig. 5b, c). These expression changes in EndMT-related markers were suppressed by linagliptin or rapamycin treatment.

Linagliptin inhibited EndMT-related transcription factors in vitro and in vivo
Snail, Slug, and Twist are key transcription factors in EndMT induction. In vitro experiments revealed that the protein and mRNA levels of these factors were significantly elevated in the BLM group compared with those in the control group; however, linagliptin or rapamycin treatment suppressed these BLM effects ( Fig. 6a-g). Immunofluorescence analysis showed that linagliptin or rapamycin attenuated the increased amount of Slug located in the nucleus induced by BLM (Fig. 6h). Next, we investigated the effect of linagliptin on EndMT-related transcription factors in vivo. Consistently, all the above mentioned transcription factors' protein in lungs in different groups were measured by respective kits. The results are presented as means ± SD (n = 6). *P < 0.05 vs control; #P < 0.05 vs BLM 1 3 levels were increased in the BLM group, compared with the control group. However, linagliptin or rapamycin treatment significantly reversed the BLM-induced increased protein levels of these factors (Fig. 7a-d). Furthermore, immunohistochemistry analysis showed an increased number of slug-positive cells in the vessel inner wall in the BLM group compared with the control group, and this upregulation was significantly reduced in the linagliptin and rapamycin groups (Fig. 7e).

Linagliptin inhibited the Akt/mTOR pathway in vivo
Activation of the Akt/mTOR pathway is an important inducer of EndMT [14]. It has also been demonstrated that inhibition of this pathway reversed BLM-induced EndMT [13,19]. Hence, we explored the effect of linagliptin on activation of the Akt/mTOR pathway. Western blot analysis revealed that BLM significantly upregulated the phosphorylation level of Akt and mTOR in SSc mice, and that this g Immunofluorescence double staining showed the expression change and location of VE-Cadherin and α-SMA in HUVECs in different groups. Nuclei were stained by DAPI. The results are presented as means ± SD from three independent experiments. *P < 0.05 vs control; #P < 0.05 vs BLM effect was reversed by linagliptin or rapamycin treatment (Fig. 8a-e). Taken together, the results indicated that the Akt/mTOR pathway is involved in the inhibition effect of linagliptin on BLM-induced EndMT.

Discussion
In the present study, we further confirmed that linagliptin improved pulmonary fibrosis in SSc mouse model by inhibiting EndMT via suppression of AKT/mTOR signaling pathway. Its suppressive effects on oxidative stress and inflammation also contributed to the protective effects.
Intratracheal administration of BLM caused severe acute lung injury followed by pulmonary fibrosis and inflammation in mice [20], which is the most commonly used animal model for studying idiopathic pulmonary fibrosis. Subcutaneous injection of BLM also induced pulmonary fibrosis with a chronic lung injury and more insidiously progressive course, which was more similar to pulmonary fibrosis in SSc patients, including patchy interstitial fibrosis and mononuclear cell infiltration. Furthermore, Yoshizaki A et al. have reported that SSc-related antibodies such as anti-Scl-70, anti-U1RNP and anti-CENP B were positive in the sera of mice treated with subcutaneous injection of BLM [21]. Therefore, subcutaneous injection of BLM may better mimic SSc-related fibrosis. And it has been commonly used to study skin and pulmonary fibrosis associated with SSc [21][22][23]. So we employed this SSc mouse model in the present study. The results showed that subcutaneous BLM administration caused loss of normal lung structure, thickened alveolar walls, and increased collagen fiber content, accompanied by significant weight loss.
DPP4 inhibitors have shown anti-fibrotic activity in various animal models of fibrosis. Mk-0626 and evogliptin repress the expression of profibrotic genes in unilateral ureteral obstruction-induced renal fibrosis model and obesity-induced myocardial fibrosis model, respectively [24,25]. Saxagliptin can improve renal fibrosis in high-fat diet-induced diabetic mice via inhibition of epithelial-to-mesenchymal transition and oxidative stress [26]. Linagliptin is able to stop the progression of liver fibrosis induced by carbon tetrachloride in mice and the possible mechanism may be via suppression of oxidative stress and mTOR [27]. Inhibition of fibroblasts activation is one of the mechanisms by which vildagliptin and saxagliptin improve bleomycin and CGVHD-induced skin fibrosis [10]. Here, our results reveal that linagliptin, the most potent DPP4 inhibitor, is also effective in improving pulmonary fibrosis in SSc mouse model.
Myofibroblasts are the main orchestrators of fibrosis by synthesizing and releasing a large amount of extracellular matrix proteins, leading to structure destruction and function loss of normal tissues and organs [28]. Myofibroblasts emerge from several sources, including expansion and activation of quiescent resident tissue fibroblasts [29], migration, and tissue accumulation of bone marrow-derived circulating CD34 + fibrocytes [30] and epithelial cells or perivascular cells (pericytes) that have undergone a phenotypic transition into mesenchymal cells [31]. EndMT is another important source of myofibroblasts newly discovered in recent years [5]. In this complex biological process, endothelial cells lose their endothelial characteristics and gain mesenchymal cell characteristics, including a spindle-shaped cell morphology, loss of polarity and intercellular junctions, and enhanced migration ability [14]. EndMT also results in the initiation of expression of mesenchymal cell specific proteins, such as α-SMA, vimentin,N-cadherin. Subsequently, the involvement of EndMT in SSc-related lung diseases was also confirmed [7]. Therefore, EndMT has become an interesting target for SSc treatment. Inhibition of EndMT also plays important roles in the improvement of pulmonary fibrosis in lipopolysaccharide-induced lung injury and renal fibrosis in streptozotocin-induced diabetic mice by vildagliptin and linagliptin, respectively [32,33]. Therefore, inhibition of EndMT is also likely to be a potential mechanism by which linagliptin ameliorates BLM-induced pulmonary fibrosis. In the present study, we treated HUVECs with BLM for 48 h and induced EndMT in vitro, similarly to a previous study [14]. Our results showed that linagliptin treatment attenuated BLM-induced EndMT in vitro. BLM-induced enhanced migration of cells was also alleviated. BLM can also induce EndMT in vivo [34]. In the study, the lungs of BLM-treated mice developed EndMT, as evidenced by the increased number of co-localized CD31/α-SMA cells Protein levels of Akt, P-Akt, mTOR, and P-mTOR were determined by Western blotting. The results are presented as means ± SD (n = 6). *P < 0.05 vs control; #P < 0.05 vs BLM observed in the lungs of the BLM group, compared with the control group. As expected, the BLM-induced EndMT was inhibited by linagliptin in vivo.
Transcription factors including snail, slug and twist are key mediators of EndMT. It has been confirmed that siRNA knockdown of snail blocked transforming growth factor (TGF)-β-induced EndMT [35], slug silencing reversed BLM-induced EndMT [14] and twist overexpression induced EndMT in cultured human pulmonary arterial endothelial cells [36]. The increased expression of these transcription factors also predicts the occurrence of EndMT. Thus, we examined the effect of linagliptin on the expression of these transcription factors. BLM upregulated the expression of snail, slug and twist in vitro and in vivo, while linagliptin treatment abrogated this effect. So we concluded that inhibition of EndMT contributes to the protective effects of linagliptin on pulmonary fibrosis in SSc mice.
The protective effects of DPP4 inhibitors on UUOinduced renal fibrosis and obesity-induced myocardial fibrosis mouse model are based on their inhibition of TGFβ-mediated Smad-dependent signaling pathway [24,25], which is also the most important and studied signaling pathway in EndMT [6]. An in vitro study reported that the interaction of DPP4 with integrin β1 promotes the TGFβ-induced TGF-β receptor heterodimer formation, initiates the Smad-dependent pathway, and ultimately leads to EndMT. Linagliptin inhibits EndMT by suppressing the expression of DPP4 and integrin β1, which may promote the activation of TGF-β/Smad signaling pathway [37]. The Akt/mTOR signaling pathway, a Smad-independent pathway, also plays an important role in EndMT [6]. BLM induces EndMT in HUVECs via activation of this pathway [14]. Studies have shown that Salvia miltiorrhiza and geniposide suppress BLM-induced EndMT by inhibiting this pathway [13,19]. Li et al. have also reported that linagliptin inhibited high glucose-induced transdifferentiation of hypertrophic scar-derived fibroblasts to myofibroblasts via inactivation of the insulin-like growth factor (IGF)/Akt/ mTOR pathway [38]. Inhibition of mTOR also contributes to the protective effects of linagliptin on liver fibrosis [27]. Our results showed that BLM induced the activation of the Akt/mTOR pathway, which was repressed by linagliptin. Therefore, linagliptin may suppress EndMT by inhibiting the AKT/mTOR signaling pathway in the SSc model.
Oxidative stress refers to an imbalance between oxidants and antioxidants. Although the relationship between oxidative stress and EndMT in SSc remains controversial [39], it has been validated that oxidative stress plays a pivotal role in the induction and progression of SSc [15]. Oxidants may directly activate fibroblasts or contribute to the maintenance of fibrosis by altering the balance between protease and anti-protease activity [40]. Inflammation is another important player in SSc. IL-6 and TNF-α levels have been demonstrated to be increased in the serum and skin of patients with SSc [41,42]. Moreover, blocking IL-6 decreased collagen production in the BLM model of lung fibrosis [43]. Linagliptin has also been reported to possess anti-inflammatory and antioxidant properties [44,45] and its protective effect on liver fibrosis is based on the antioxidant properties [27]. In our study, BLM-induced oxidative stress and inflammation were inhibited by linagliptin in vivo, indicating that the therapeutic effect of linagliptin against BLM-induced pulmonary fibrosis may also be attributed to its anti-inflammatory and antioxidant properties.
The study also has some limitations. First, the inhibitory effect of linagliptin on EndMT in vitro was not repeated in human pulmonary endothelial cells in this study limited by objective conditions. Second, alternative models were not recruited to further validate these effects of linagliptin. These are issues that need to be addressed in future research.

Conclusion
We further confirmed that linagliptin attenuated pulmonary fibrosis SSc in mice, and the potential mechanisms might be due to inhibiting EndMT via the suppression of Akt/mTOR signaling pathway, inflammation, and oxidative stress (Fig. 9). Thus, linagliptin may be a promising candidate for the treatment of SSc-related pulmonary fibrosis. Fig. 9 The potential mechanisms of linagliptin on pulmonary fibrosis in SSc mice