Cyclophilin A/CD147 Induces the Epithelial-to- mesenchymal Transition and Renal Fibrosis in Chronic Allograft Dysfunction by Regulating MAPK Signaling

Xuzhong Liu Huai’an First People’s Hospital, Nanjing Medical University Zhiwang Tang Huai’an First People’s Hospital, Nanjing Medical University Xi Jiang Huai’an First People’s Hospital, Nanjing Medical University Tianwei Wang Huai’an First People’s Hospital, Nanjing Medical University Lun Zhao Huai’an First People’s Hospital, Nanjing Medical University Zongyuan Xu Huai’an First People’s Hospital, Nanjing Medical University Kun Liu (  chenhao_nanjing@126.com ) Huai’an First People’s Hospital, Nanjing Medical University


Introduction
Chronic allograft dysfunction (CAD) remains to be the main factors in uencing the long-term allograft survival of renal transplant recipients, which was characterized by interstitial brosis, tubular atrophy and glomerulitis (1,2). While substantial studies have reported the mechanism of renal allograft brosis, there is still lack of detailed and comprehensive knowledge for the CAD. It is known that various factors, such as transforming growth factor-beta 1 (TGF-β1), tumor necrosis factor-α (TNF-α), have been veri ed to play a crucial role in the formation of allograft brosis (1,3,4). Recent years, Cyclophilin A (CyPA), an ubiquitous intracellular protein, has been recognized to be secreted into extracellular space by various cell types, such as macrophages, activated platelets (5). Moreover, extracellular CyPA exerts its in ammatory effects through binding to the Extracellular Matrix Metalloproteinase Inducer (EMMPRIN, CD147) in a celltype speci c manner (6). It is well-documented that extracellular CyPA is currently involved in the expression of adhesion molecules and induction of the in ammatory responses, including the chemotaxis of monocyte, neutrophile granulocytes, T lymphocytes, and eosinophile granulocytes (7,8).
Importantly, CyPA/CD147 is believed to be involved in the pathogenesis of brosis in several organs. The Bsg, the name of mouse CD147 gene, promotes the transition of broblasts to myo broblasts by increasing the expression of α-smooth muscle actin (α-SMA) in corneal broblasts (9). In cardiac model in mice, knockout of CyPA leads to the attenuation of macrophages and T lymphocytes in ltration, and also retards the progression of myocardiac brosis, indicating that CyPA may represent a pharmacological target to modulate the myocardial remodeling in myocarditis (10). Furthermore, knockout of Bsg is signi cantly related with the reduced production of matrix metalloproteinase (MMP) and induction of TGF-β1, thus decreasing the formation of renal brosis (11). Therefore, interactions between CyPA and CD147 could be strongly suggested to be an important regulator of organ brosis, including renal brosis.
In this study, we hypothesized that CyPA/CD147 could protmote the progression of renal allograft brosis by targeting the epithelial-to-mesenchymal transition (EMT), which has been identi ed as a crucial origin of the renal brosis, consequently contribute to the CAD pathogenesis.

Ethic statement
The study protocol was in compliance with the principle of the Declaration of Helsinki and Istanbul.
Whole study design, including animal and cell experiments, was approved by the ethics committee of Huai'an First People's Hospital.

Animals and experimental design
Adult male Lewis and F344 rats weighted ranging from 200 to 250 g were purchased from Charles River Laboratories (Beijing, China) and raised in the Animal Feeding Center of Nanjing Medical University with clean tap water and standard rat chow.
Orthotopic renal transplant surgery was conducted between the donor of Lewis rats and the recipient of F344 rats in CAD group. The mean warm ischemia time and cold ischemia time for the corresponding surgeries were 34.1 minutes (mins) and 15.9 mins. Nephrectomy surgery of right kidney in recipient rats was performed after the 3 days post-transplant. Cyclosporine A with a concentration of 5 mg/kg bodyweight (Novartis, Switzerland) was administrated once a day intraperitoneally, which lasted for 14 days post-transplant. The sham surgery was performed on the F344 rats from the same batch, which was considered as sham group. The recipients were executed by head-breaking after 20 post-transplant weeks, and renal allograft tissues were harvested. Then, samples were divided into two parts, which were conserved in -80°C liquid nitrogen and xed in the para n, respectively. In addition, the blood samples of recipient rats in both groups were collected in the 4, 12 and 20 post-transplant weeks, and stored in -80°C liquid nitrogen. Concentrations of serum creatinine and blood usea nitrogen (BUN) were assessed according to the manufacturer's instructions (Jiancheng, Beijing, China).

Histology
The collected rat renal allograft tissues were used for the HE, Masson trichome and immunohistochemistry (IHC) staining as described previously (12). Brie y, 2-μm sections were stained with Haematoxylin and Eosin during the HE staining. The morphological changes were observed and evaluated by interstitial brosis and tubular atrophy, and brosis area was calculated from ve elds of view in each section. Furthermore, sections were de-para nized and rehydrated in a series of alcohol with graded concentrations. The primary antibodies (rabbit polyclonal anti-E-Cadherin, 1: 100; rabbit

Real-time PCR
Rat kidney tissues were snap-frozen in liquid nitrogen for total mRNA isolation, whereas total mRNA of HK-2 cells from each group was also extracted by the RNA extraction kits (TIANGEN, Beijing, China). Brie y, a PrimeScriptTMRT reagent kit (TaKaRa Biotechnology, Japan) was used to synthesize the cDNA.

Western Blotting
The detailed procedures of Western Blot assay were described as previously reported (12). Brie y, total proteins were extracted from the rat kidney tissues and HK-2 cells derived from each group. Primary antibodies used in this study were shown as below: anti-α-SMA antibody (1: Enzyme-linked immunosorbent assay (ELISA) ELISA assays were performed to assess the concentration of Collagen-and Fibronectin in the supernatant of cell culture treated with siRNA-CD147 and/or CyPA. The whole procedures were performed as described by the manufacturer's instructions. The assays were performed for at leat three times.

Statistical analysis
All data was shown as mean ± standard deviations (SD). Statistical analysis was conducted with the unpaired, two-tailed Student's t-test for single comparisons. A P value less than 0.05 was considered as statistical signi cant. All statistical analysis was performed by SPSS v18.0 (SPSS Inc., Chicago, IL, USA).

CyPA and CD147 are highly expressed in chronic allograft dysfunction
We have established a 20-weeks rat renal transplant model with CAD. As shown in Figure 1, remarkable interstitial brosis and tubular atrophy (IF/TA), which was considered as the crucial pathological characteristics for CAD, was observed in the HE and Masson staining from allograft in the CAD group ( Figure 1A-B). Moreover, signi cant in ltration of mononuclear cells among the IF/TA was also found. The quantitative analysis of brosis area between two groups supported the identi cation of rat renal transplant CAD model ( Figure 1C). Consequently, we analyzed the allograft function by serum creatinine and blood urea nitrogen (BUN), and results showed that renal allograft function was signi cantly reduced in a time-dependent manner in CAD group when compared with the sham subjects ( Figure 1D-E).
We tested protein expression of E-Cadherin, as well as α-SMA, the biomarkers of epithelial-tomesenchymal transition (EMT), in the allograft tissues. The results showed that the expression of α-SMA was up-regulated and expression of E-Cadherin was decreased, suggesting the existing presence of EMT progression in CAD (Figure 2A-D). Then, we designed to explore the protein distribution and expression of CyPA and CD147 in the pathogenesis of CAD. Enhanced expression of CyPA and CD147 were identi ed in 20-weeks renal allograft tissues and mainly distributed in the allograft tubular epithelial regions ( Figure   2E-H). Furthermore, we examined the total protein and mRNA expression of CyPA and CD147 in allograft tissue. The results were consistent with those in IHC staining ( Figure 2I-K). Thus, total results suggested the potential association of increased CyPA and CD147 with the progression of EMT and renal allograft brosis.

CyPA/CD147 interaction induces the progression of EMT
To clarify the mechanism of CyPA/CD147 during the EMT progression and renal brosis, in vitro experiments were performed. Western blotting results showed the protein expression of CD147, as well as α-SMA and Fibronectin (FN), were remarkably higher after the stimulation of CyPA in a dose-dependent manner in HK-2 cells, whereas the expression of E-Cadherin was conversely reduced ( Figure 3A). Similar results were indicated by the mRNA expression ( Figure 3B-D). Next, we silenced the expression of CD147 in HK-2 cells to further con rm the effect of CyPA on the EMT progression. The siRNA-CD147 intervention signi cantly reduced the expression of CD147 in HK-2 cells, whereas the positive regulation of α-SMA and negative regulation of FN were also reversed ( Figure 4A-D). In addition, the concentration of Collagenand FN in the supernatant were assessed by ELISA assay. The results showed that secretion of Collagenand FN from the HK-2 cells induced by CyPA were signi cantly inhibited by siRNA-CD147. Therefore, the in vitro study indicated that the CyPA/CD147 interaction could signi cantly induce the progression of EMT in HK-2 cells.

MAPK signaling pathway contributes to the progression of EMT and renal brosis induced by CyPA/CD147 interaction
It's reported that MAPK signaling pathway participates in the regulation of CyPA and CD147 (13). Hence, we further explored the potential mechanism involved in the effect of CyPA/CD147 interaction on the progression of EMT and renal brosis. Firstly, we tested the activation of MAPK signaling in the CyPA/CD147 interactions. Figure 5A showed signi cant activation of phosphorylated p38 MAPK, the crucial mediator of MAPK signaling, of which could be signi cantly inhibited by the treatment of siRNA-CD147 ( Figure 5B). Then, we used the speci c inhibitor for MAPK signaling (SB203580), to con rm the activation of MAPK signaling. We have observed that upon the intervention of SB203580, the expression of CD147 showed no signi cance ( Figure 5C-D), whereas the expression of α-SMA was remarkably inhibited ( Figure 5C and 5E), suggesting the downstream role of MAPK signaling among the interaction of CyPA/CD147. Importantly, the activation of phosphorylated p38 MAPK was also observed in the renal allograft tissue from rat renal transplant CAD model ( Figure 5F-G).

Discussion
Our in vivo and vitro study showed that CyPA/CD147 interactions were signi cantly activated in the progression of renal allograft brosis and CAD. Moreover, CyPA/CD147 could remarkably induce the process of EMT by MAPK signaling and then contribute to the renal allograft brosis, indicating the CyPA/CD147 could serve to be an important target for the renal allograft brosis. To the best of our knowledge, this is the rst study to investigate the role of CyPA/CD147 interactions in the renal allograft brosis.
Numerous reports were designed to investigate the cellular mechanism of CAD in renal transplant, factors of which could be divided into immunological and non-immunological factors (14). Among these, EMT was identi ed to play a crucial role in the pathogenesis of renal allograft brosis by inducing the expression of α-SMA and extracellular matrix, including collagen-and FN (15). In this study, we have identi ed that EMT induced by CyPA/CD147 contributes to the renal allograft brosis, which was consistent with the speci c role of EMT in renal brosis. Also, blockade of CD147 could also attenuate the progression of EMT and renal brosis, suggesting the downstream effect of EMT in the regulation of CD147. Similarly, one study focusing on the renal brosis in the unilateral ureteral obstruction (UUO) mice model provides the convincing outcomes of Bsg/CD147 in the promoting the development of renal brosis by knocking out the Bsg in mice, which was consistent with our ndings (11). Compared to this study, we also noticed that EMT could serve as a key downstream process involved in the regulation of CyPA/CD147 in renal brosis. Interestingly, we also observed the signi cant in ltration of mononuclear cells and highly expressed CyPA/CD147 distributed in the IF/TA of rat model with CAD. With regard to the effect of CyPA/CD147 in in ammatory responses, we tended to consider that there was a potential link of CyPA/CD147 and immunological cells in ltration during the CAD process, which still remained further analysis. Therefore, our ndings supported the CD147 as a candidate target molecule for the treatment of renal brosis, especially in renal allograft brosis.
We have identi ed that CyPA may have a direct pro-brotic effect on the tubular epithelial cells and renal brosis. The canonical stimulation for the renal brosis has been recognized as TGF-β1, TNF-α, MCP-1 and CXCL-10 (16, 17). Considering the important role of CD147 in the regulation of renal brosis, the CyPA binding to the CD147 was also noticed to be involved in the stimulation of renal brosis. Moreover, extracellular CyPA has been described to induce the chemotaxis of T cells, monocytes, and macrophages, which leads to the massive recruitment and in ltration of in ammatory cells in the renal allograft tissues (10,18). It is well known that certain pro-brotic cytokines could be produced and secreted during the in ammatory response, triggering the pathogenesis of organ brosis. Therefore, the indirect pro-brotic effect of CyPA should also draw great attention even though the crucial role of CyPA/CD147 interaction in renal brosis.
Recent study has reported the positive relation of CyPA/CD147 signaling and osteoclast-related MMP-9 expression in mice in ammatory periapical lesions progression (18). Additionally, CyPA inhibitor NIM811 signi cantly reduced MMP-9 secretion during the differentiation process of foam cells, which also suggested the potential association of the CyPA/CD147 and MMP-9 expression. Importantly, suppressed activities of MMP-2 and MMP-9 proteins in the Bsg-/-UUO mice was observed, providing the direct evidence for the CyPA in the regulation of MMP-9 expression (11). Thus, we believed that except for the EMT process, the MMP-9-related signaling induced by CyPA/CD147 interactions also plays a key role in the extracellular matrix deposition and degradation, and consequently renal brosis. In the in vitro study, we used the CD147 mAb to selectively block the interactions between CyPA and CD147, and observed the attenuation of EMT and renal brosis, which was consistent with other studies and our in vivo study. With great regret, we could not perform the intervention of CD147 mAb in the rat CAD model or the Bsg-/-CAD model to explore and con rm the unique effect of CD147 in the renal allograft brosis and CAD.

Conclusion
In conclusion, our study reported the signi cant expressed CyPA/CD147 in the renal allograft brosis tissues, and suggested the potential association of CyPA/CD147 interactions with CAD. Moreover, in vitro study showed that CyPA could remarkably promote the EMT process by binding to the CD147 and triggering the intracellular MAPK signaling pathway, which contributes to the renal brosis. Our data provide evidence that CyPA is critically involved in the pathophysiology of renal allograft brosis, and CyPA/CD147 interactions may represent a target to modulate the renal allograft brosis and CAD.

Declarations
Ethics approval and consent to participate

Availability of data and materials
The data analysed during the current study are available from the corresponding author on reasonable request.

Competing interest statement:
No competing interest exists in this manuscript.

Funding
This study was supported by the program of "Innovative young talent project" of Huai'an First People's Hospital.
Authors' contributions X. Liu   CyPA were also tested (I-K). ***P <0.001 compared with the control group; ****P <0.0001 compared with the control group. Each experiment was repeated for at least three times.

Figure 3
The effect of CyPA on CD147 and tubular brosis in HK-2 cells. We stimulated the HK-2 cells by various concentrations of CyPA and extracted the total protein and mRNA from cells. The CD147 protein, as well as the biomarkers related to tubular brosis, was tested by Western Blot assay (A). Moreover, the mRNA expression of E-Cadherin (B), Fibronectin (FN; Figure 3C) and CD147 (D) were explored by RT-PCR. *P <0.05 compared with the control group; **P <0.01 compared with the control group; ****P <0.0001 compared with the control group. Each experiment was repeated for at least three times.

Figure 4
The identi cation of effect of CyPA on the CD147 expression in HK-2 cells. We conducted the siRNA-CD147 to slience the expression of CD147, and stimulated the HK-2 cells with or without CyPA. The quantitative analysis of CD147 (B), E-Cadherin (C), and α-SMA (D) was presented. Furthermore, the concentration of Collagen-(E) and Fibronectin (F) in the supernatant of cell culture were assess by ELISA assays. NS, no signi cant; ***P <0.001 compared with the control group; ****P <0.0001 compared with the control group. Each experiment was repeated for at least three times.