Salvianolic Acid C Inhibits the Epithelial-Mesenchymal Transition and Ameliorates Renal Tubulointerstitial Fibrosis

Background : Salvianolic acid C (SAC) is a natural compound derived from Salvia miltiorrhiza that can protect against renal diseases. The aims of this work were to explore the effect of SAC on kidney tubulointerstitial fibrosis and study the associated mechanism. Methods : Models for unilateral ureteral obstruction (UUO) and aristolochic acid I (AAI) were established in mice to study renal tubulointerstitial fibrosis. Rat kidney fibroblasts (NRK-49F) and human kidney epithelial cells (HK2) were used as cellular models to evaluate the effects of SAC on kidney fibrosis. Results : Treatment with SAC for two weeks reduced the level of renal tubulointerstitial fibrosis in UUO-and AAI-induced fibrotic kidneys, as demonstrated by Masson’s staining and Western blot. SAC inhibited extracellular matrix protein expression in NRK-49F cells and TGF-β -stimulated HK2 cells in dose-dependent fashion. Moreover, SAC inhibited the expression of epithelial-mesenchymal transition (EMT) factors in animal and cellular models of kidney fibrosis, as well as the EMT-related transcription factor snail. Furthermore, SAC inhibited the fibrosis-related signaling pathway Smad3 in the fibrotic kidneys of two mouse models and in renal cells. Conclusions : We conclude that SAC inhibits EMT and ameliorates tubulointerstitial fibrosis through involvement of the signaling pathway for transforming growth factor-β (TGF-β )/Smad.


Introduction
Chronic kidney disease (CKD) is a major health issue worldwide because of its high prevalence, bad prognosis, and high medical burden [1]. CKD prevalence has been increasing recently and has now reached 14.3% worldwide [2]. Renal tubulointerstitial fibrosis occurs in all CKDs that progress to late-stage kidney disease [3]. This condition is characterized by pathological accumulation of extracellular matrix (ECM) proteins, including fibronectin and type 1 collagen [4]. Signaling through the TGF-β/Smad pathway is an important factor for renal fibrosis, whereby the Smad3 transcriptional factor regulates expression of fibrogenic genes [5,6].
During epithelial-mesenchymal transition (EMT), epithelial cells transform into a mesenchymal phenotype. Upregulation of EMT genes, including N-cadherin and vimentin, represent hallmarks of renal tubulointerstitial fibrosis [7]. A major inducer of EMT is Snail, which plays a major part in organ fibrosis, including renal tubulointerstitial fibrosis [8,9].
Salvianolic acid C (SAC) is a phenolic acid extracted from Salvia miltiorrhiza [10]. The effect of different components of salvianolic acids on renal fibrosis has recently been studied. Salvianolic acid B exerts an anti-fibrotic ef-fect on kidney diseases by regulating the HPSE/SDC1 axis and activating sirt1-mediated autophagy [11,12]. Salvianolic acid A ameliorates kidney fibrosis in 5/6 nephrectomized rats through inhibition of nuclear factor kappa-B (NF-κB) and p38 mitogen-activated protein kinase (MAPK) signaling pathways [13]. Salvianolic acid A has a similar structure to SAC, and can be converted into SAC by heating and other conditions [14]. SAC inhibits inflammatory responses in several disease models [10,15,16]. Moreover, salvianolic acid A in combination with SAC can inhibit fibrosis in obstructive rat kidneys [17]. The action of SAC alone on kidney tubulointerstitial fibrosis remains unknown, however.
The goal of the current work was to study the effect of SAC on kidney tubulointerstitial fibrosis, as well as the associated mechanism.

Mice
SPF grade C57BL/6 male mice were obtained from Shanghai Laboratory Animal Company (Shanghai, China) and housed at Shanghai University of Traditional Chinese Medicine in line with their rules and regulations. The animal ethics committee of the Shanghai Univer- collagen I were analyzed by Western blotting and then quantified. (C) Representative immunofluorescent images of collagen I staining in sham or UUO mouse kidneys. These were also quantified. DAPI (nuclei) = blue, collagen I = green. Bar = 20 µm. N = 6 in each experimental group. One representative of at least three independent experiments is shown. Data represents the mean ± SD. **p < 0.01. ***p < 0.001. sity of Traditional Chinese Medicine approved this study (PZSHUTCM18111601).

Mouse Unilateral Ureteral Obstruction (UUO) Model
Male mice weighing from 18 to 22 g were anesthetized intraperitoneally with 8 mg/kg of pentobarbital before performing an abdominal incision on the left side. This was followed by either a sham operation or left ureteral ligation. Animals were assigned into 4 groups at random: (a) Sham + dimethyl sulfoxide (DMSO) (n = 6), (b) Sham + SAC (n = 6), (c) UUO + DMSO (n = 6), (d) UUO + SAC (n = 6). Intraperitoneal injections of DMSO or SAC (10 mg/kg body weight, TopScience, T3149, Shanghai, China) were administered after the surgery and for 14 consecutive days. Mice were sacrificed at day 14, with kidney tissue harvested for protein extraction and tissue embedding.

Mouse Model Induced by Aristolochic Acid I (AAI)
Mice were injected intraperitoneally with 5 mg/kg AAI (A9451, Sigma, St. Louis, MO, USA) in normal saline (NS) or with NS only once per week for 2 weeks (2 injections in total). One week after the second injection of AAI, the DMSO or SAC (10 mg/kg body weight, TopScience, T3149, Shanghai, China) was injected intraperitoneally and daily for two weeks (n = 6). Mice were sacrificed after the treatment, and kidneys were collected for protein extraction and tissue embedding.

Cell Culture
Cells were grown as previously described by our group [18]. Normal rat kidney interstitial fibroblasts (NRK-49F) rat kidney interstitial fibroblast cells were obtained from the National Infrastructure Cell Line Resource, Chinese Academy of Medical Sciences (cat. 3111C0001CCC000413), which has been identified for the species by PCR and was negative for mycoplasma contamination. The cell line has been tested for the mycoplasma and been authenticated from the Chinese Academy of Medical Sciences. The cell line is orginated from American Type Culture Collection, VA, USA (Cat. No. CRL1570). The culture medium for NRK-49F cells was Dulbecco's Modified Eagle Medium/Nutrient Mixture F-12 (DMEM/F12) with 10% fetal bovine serum (FBS) and 0.5% penicillin/streptomycin. NRK-49F cells were grown in 6-well plates until 60-70% confluence was reached, then starved in medium containing just 0.5% FBS for 12 h. After overnight starvation, fresh medium with 0.5% FBS was added and the cells treated with DMSO or with vari-

Fig. 2. SAC attenuates the development of renal tubulointerstitial fibrosis in mice with aristolochic acid nephropathy (AAN).
Mice (n = 6) were injected intraperitoneally with aristolochic acid I (AAI) or normal saline (NS) once a week for two weeks. One week after the second injection of AAI or NS, mice were treated with DMSO or SAC (10 mg/kg) for two weeks. (A) Renal fibrosis was assessed by Masson's trichrome staining and then quantified. Bar =100 µm. (B) The expression of fibronectin and collagen I were analyzed by Western blotting and then quantified. (C) Representative immunofluorescent images of collagen I staining in control or AAN mouse kidneys. These were also quantified. DAPI (nuclei) = blue, collagen I = green. Bar =20 µm. N = 6 in each experimental group. One representative of at least three independent experiments is shown. Data represents the mean ± SD. *p < 0.05. **p < 0.01. ***p < 0.001. ous concentrations of SAC (TopScience, T3149, Shanghai, China). The kidney proximal tubular epithelial cell line Human Kidney 2 (HK2, cat. SCSP-511) was purchased from the Cell Bank, Shanghai Institute of Biological Sciences (Chinese Academy of Science), which has been validated by short tandem repeat profiling and was negative for mycoplasma contamination. The cells were grown in 6well plates to 40-50% confluence, then starved overnight with DMEM/F12 containing just 0.5% FBS. The following day, fresh DMEM/F12 with 0.5% FBS was added, together with 2.5 ng/mL TGF-β (PeproTech, 100-21, Rocky Hill, NJ, USA) and different SAC concentrations for 48 h. The SAC concentrations used here were the same as those used in two previous publications [15,19].

Masson's Trichrome Staining
Staining with Masson's trichrome was carried out as described earlier [18]. Briefly, paraffin-embedded renal tissue sections (4 µm) were cut and then stained using hematoxylin and ponceau, followed by incubation with phosphomolybdic acid. Kidney tissue was then stained using ani-line blue and acetic acid and photographed using a Nikon 80i microscope (Nikon Eclipse 80i, Tokyo, Kanto Plain, Japan).

Immunofluorescence
Immunofluorescence staining of renal tissue was carried out as described previously [21]. In short, paraffinembedded tissue was cut into 4 µm sections, blocked with 3% BSA, incubated overnight with anti-collagen I antibody (1:500) at 4°C, rinsed in PBS, then incubated for 60 min with secondary antibody. Cell nuclei were stained with 4′,6-Diamidino-2-phenylindole (DAPI) solution for 10 min and photographed using a Nikon 80i microscope (Tokyo, Japan). Mean fluorescence intensity for each slide was analyzed using Image J and normalized to the control.

Fig. 4. SAC inhibits epithelial-mesenchymal transition in fibrotic kidneys and renal fibrotic cells. (A) The expression of Snail in sham or UUO kidneys (n = 6) was analyzed by Western blotting and then quantified. (B) The expression of Snail in NS or AAN kidneys
(n = 6) was analyzed by Western blotting and then quantified. Data represents the mean ± SD. (C) HK2 human renal epithelial cells were starved overnight, then stimulated with TGF-β and treated with various concentrations of SAC for 24 h. The expression of N-cadherin and of vimentin was analyzed by Western blotting and then quantified. One representative result of at least three independent experiments is shown. n.s. represents not significant. *p < 0.05. **p < 0.01. ***p < 0.001. methods and normalized with GAPDH. Results are shown as the fold-change in mRNA expression.

Statistical Analyses
Results are given as mean ± SD. Student t-test was used to compare two groups with GraphPad Prism 8.0.0 (GraphPad Software, San Diego, CA, USA). A p-value < 0.05 was considered as representing statistical significance. Each experimental group had an n = 6. Representative results from at least three independent experiments are presented. *p < 0.05. **p < 0.01. ***p < 0.001.

SAC Inhibits Renal Tubulointerstitial Fibrosis in Obstructed Mouse Kidneys
The effect of SAC on kidney tubulointerstitial fibrosis was first investigated in vivo. UUO is a classic mouse model for studying kidney fibrosis and was established here for this purpose. UUO mice received DMSO or SAC treatment for two weeks. Compared with sham operated mice, significant deposition of collagen was detected using Masson's trichrome staining in UUO kidneys. How-ever, this was attenuated by two weeks of treatment with SAC (Fig. 1A). Western blot experiments showed that expression of two ECM proteins, fibronectin and collagen I, was up-regulated in mouse kidneys after UUO surgery, but was lowered by treatment with SAC (Fig. 1B). Elevated levels of collagen I expression in the interstitial areas of UUO kidneys relative to sham kidneys was also shown by immunofluorescent staining. Again, this was significantly attenuated by SAC treatment (Fig. 1C).

SAC Attenuates Kidney Tubulointerstitial Fibrosis in the Mouse AAN Model
The effects of SAC on anti-fibrosis were further studied in mice with chronic aristolochic acid nephropathy (AAN). Masson's trichrome revealed mild staining in the interstitial areas of AAN kidneys ( Fig. 2A). Treatment with SAC for two weeks significantly reduced the positive areas shown by Masson's trichrome staining in AAN kidneys ( Fig. 2A). Fibronectin and collagen I expression were increased in AAN kidneys relative to the control kidneys. SAC treatment was observed to reduce the expression of these ECM proteins (Fig. 2B). Up-regulation of collagen I in the interstitial areas was further confirmed by immunofluorescent staining observed in AAN kidneys relative to control kidneys (Fig. 2C). SAC treatment significantly reduced positive areas of collagen I staining in AAN kidneys (Fig. 2C).

SAC Inhibits ECM Protein Expression in Renal Cells
Direct effects of SAC on kidney cells were further investigated in vitro by using the rat fibroblast cell line NRK-49F, as well as the human kidney proximal tubular epithelial cell line HK2. The mRNA levels of fibronectin, collagen I and collagen III in NRK-49F cells were reduced by SAC in dose-dependent fashion from 10 to 100 µM (Fig. 3A). Western blot analysis showed that SAC also dose-dependently reduced fibronectin and collagen I protein expression in these cells from 30 µM to 100 µM (Fig. 3B). The dosedependent (10 µM to 100 µM) inhibition of fibronectin expression by SAC was further shown in HK2 cells stimulated with TGF-β (Fig. 3C).

SAC Inhibits EMT in Fibrotic Kidneys and Renal Fibrotic Cells
The effect of SAC on EMT was evaluated in vivo as well as in vitro. Snail expression in UUO and AAN kidneys was increased compared to their respective controls (Fig. 4A,B). However, treatment with SAC reduced renal expression of Snail in these mouse models. Increasing doses of SAC (10 µM to 100 µM) also progressively inhibited N-cadherin and vimentin expression in TGF-βstimulated HK2 cells (Fig. 4C).

SAC Inhibits Smad3 Signaling in Fibrotic Kidneys and Renal Fibroblasts
We next evaluated the effects of SAC on Smad3 signaling in vivo and in vitro. Fig. 5A,B show that Smad3 phosphorylation was enhanced in fibrotic kidneys of UUO or AAN mice compared to their controls. SAC treatment decreased Smad3 phosphorylation (pSmad3) levels in UUO and AAN kidneys (Fig. 5A,B). Moreover, SAC inhibited pSmad3 in TGF-β-stimulated HK2 cells in a dosedependent fashion from 10 µM to 100 µM. This correlated with up-regulated expression of the anti-fibrotic protein Smad7 (Fig. 5C).

Discussion
The beneficial effects of SAC were recently demonstrated in cisplatin-induced acute kidney injury [10]. The effects of SAC on kidney fibrosis, however, are still un-known. To examine the role played by SAC in kidney tubulointerstitial fibrosis, we established the UUO and AAN mouse models. We showed that two weeks of treatment with SAC reduced collagen deposition in UUO or AAN kidneys, as observed with Masson's staining. Western blot and immunofluorescence methods revealed down-regulation of two ECM proteins (fibronectin and collagen I) following SAC treatment in these two mouse models. The antifibrotic effects of SAC on kidney cells were further studied using renal fibroblasts and TGF-β-stimulated kidney epithelial cells. Fibronectin and collagen I expression levels both decreased in a dose-dependent fashion after SAC treatment of renal fibroblasts and TGF-β-stimulated kidney epithelial cells. We conclude from these results that SAC protects against tubulointerstitial fibrosis in kidney diseases.
TGF-β/Smad3 is the classic pro-fibrotic signaling pathway [5,22]. The effect of phenolic acid extracts from salvia miltiorrhiza on the TGF-β/Smad3 signaling pathway has been studied earlier. Zhang et al. [23] found that salvianolic acid A ameliorates kidney damage in rats with chronic kidney failure through inhibition of TGF-β signaling. Tao et al. [24] reported that salvianolic acid B delayed progression of liver fibrosis by inhibiting TGF-β signaling. Furthermore, inhibition of TGF-β signaling by salvianolic acid B reduces lung fibrosis [25]. Anti-fibrotic effects of salvianolic acid B in renal diseases are thought to be due to inactivation of the TGF-β/Smad pathways [26]. The present study showed for the first time that SAC inhibits TGF-β/Smad3 signaling. Smad7 was found to negatively regulate TGF-β/Smad-mediated pro-fibrotic signaling [27,28]. In our study, down-regulation of Smad7 by TGF-β stimulation was restored by SAC treatment. Thus, we conclude that SAC inhibits signaling by TGF-β/Smad via the inhibition of Smad3 and activation of Smad7.
A characteristic of kidney fibrosis is the upregulation of EMT markers [9]. The transcriptional factor Snail is an important molecular switch for the EMT program and has been involved in renal tubulointerstitial fibrosis [9]. Two studies in gastric cancer and renal fibrosis found that salvianolic acid B inhibits EMT [12,29]. In the present study, SAC was observed to reduce the expression of Ncadherin and vimentin in a dose-dependent fashion in TGFβ-stimulated kidney epithelial cells. Furthermore, SAC inhibition of snail was observed in animal models. We conclude that SAC is an inhibitor of EMT in renal fibrosis.

Conclusions
We conclude that SAC inhibits EMT and attenuates kidney tubulointerstitial fibrosis through the signaling pathway for TGF-β/Smad.

Availability of Data and Materials
The data used to support the findings of this research are available upon request.

Author Contributions
MW, DC and CY conceived and coordinated the study. MW wrote the paper. JL conducted the in vitro experiments. MW, DH and JL performed the animal experiments. JL performed and analyzed the Western blotting. All authors reviewed the results and approved the final version of the manuscript.

Ethics Approval and Consent to Participate
The animal ethics committee of the Shanghai University of Traditional Chinese Medicine approved this study (PZSHUTCM18111601).