Flavanomarein inhibits high glucose-stimulated epithelial-mesenchymal transition in HK-2 cells via targeting spleen tyrosine kinase

Flavanomarein (FM) is a major natural compound of Coreopsis tinctoria Nutt with protective effects against diabetic nephropathy (DN). In this study, we investigated the effects of FM on epithelial-mesenchymal transition (EMT) in high glucose (HG)-stimulated human proximal tubular epithelial cells (HK-2) and the underlying mechanisms, including both direct targets and downstream signal-related proteins. The influence of FM on EMT marker proteins was evaluated via western blot. Potential target proteins of FM were searched using Discovery Studio 2017 R2. Gene Ontology (GO) analysis was conducted to enrich the proteins within the protein-protein interaction (PPI) network for biological processes. Specific binding of FM to target proteins was examined via molecular dynamics and surface plasmon resonance analyses (SPR). FM promoted the proliferation of HK-2 cells stimulated with HG and inhibited EMT through the Syk/TGF-β1/Smad signaling pathway. Spleen tyrosine kinase (Syk) was predicted to be the most likely directly interacting protein with FM. Combined therapy with a Syk inhibitor and FM presents significant potential as an effective novel therapeutic strategy for DN.

Prediction of potential targets of FM via a computational reverse docking approach. In total, 43 proteins were identified, including 15 human proteins. In order to establish the key proteins contributing to the effects of FM on cell migration, we constructed a PPI network of the 15 human proteins based on the results of reverse target screening incorporating 1254 nodes and 1463 interactions (Fig. 3a). Subsequently, GO analysis was performed to categorize all proteins within the PPI network based on the "biological process" they are involved in (Fig. 3b). We focused on the cell motility-relevant GO terms "cell migration" and "cell motility" and their mother terms. Intriguingly, RET and Syk appeared in all the cell migration and motility-relevant GO terms (Table 1). Notably, Syk mediates upregulation of high glucose-induced TGF-β1 in HK-2 cells 22 and is involved  Data are presented as means ± SD, n = 3 (three separate experiments). * P < 0.05, ** P < 0.01, *** P < 0.001, compared with control, and # P < 0.05, ## P < 0.01, ### P < 0.001, compared with HG (HG, high glucose; FM, flavanomarein). mol in the optimal binding pose, showing better binding energy than the endogenous ligand LASW836 (−57.4404 kcal/mol). As shown in Fig. 4c, Lys458, Asn499, Asp512, Leu453 and Glu452 play decisive roles in hydrogen bond formation, in particular, Lys458, which contributes to stabilizing the complex of Syk and FM. A www.nature.com/scientificreports www.nature.com/scientificreports/ model of the complex of Syk bound to FM in solvent is presented in Fig. 4d. The RMSD reference of FM, plotted in Fig. 4e, showed that interactions of the receptor-ligand complex reach the equilibrium state after 12 pescs. A similar situation was observed in the analysis of interactions between O of FM and HN in the amino residue of Lys458 in Syk (Fig. 4f-h), suggesting that these two residues of the catalytic site stabilize the interactions between FM and Syk. A hydrogen bond heat map of the Syk-FM complex is presented in Supplemental Fig. 1. The ordinate represents all possible hydrogen bonds in the protein and the vertical coordinates are the steps in the simulation, indicating activation of hydrogen bonds in each step. We additionally investigated the binding affinity of FM for Syk based on SPR. The response unit (RU) values increased significantly with incremental FM doses from 6.25 to 200 μM (Fig. 4i), indicating that FM directly binds Syk in a concentration-dependent manner. The equilibrium dissociation constant of FM binding to immobilized Syk on a CM5 chip (K D = kd/ka) was 3.064 × 10 −5 M, supporting the theory that Syk is a direct target of FM.

A Syk inhibitor inhibits α-SMA, FN, and Vimentin and increases E-cadherin expression in HG-treated HK-2 cells.
To validate whether Syk is a direct target of FM, HG-exposed HK-2 cells were treated with BAY61-3606, a potent, ATP-competitive, and highly selective inhibitor of Syk tyrosine kinase with no suppressive effects on Lyn, Btk, Fyn, Itk and Src. Protein expression of E-cadherin, Vimentin, α-SMA, and FN in a diabetic kidney model was detected via western blot, as shown in Fig. 5. Compared with the control group, the HG group showed a significant decrease in E-cadherin, and conversely, a significant increase in α-SMA, Vimentin, and FN levels. Relative to the HG group, E-cadherin expression was markedly increased in the group co-treated with FM (80 µM) and the Syk inhibitor, BAY61-3606 (1 µM). The FM + BAY61-3606 treatment group displayed the highest increase in E-cadherin overall. Moreover, FM, BAY61-3606, and FM + BAY61-3606 treatment caused a marked decrease in the levels of α-SMA, Vimentin, and FN, compared with the HG group. Our results suggest that Syk is implicated in the anti-EMT effect of FM.
The Syk inhibitor ameliorates renal EMT through suppression of the Syk/TGF-β1/Smad signaling pathway. To elucidate the mechanism underlying the anti-EMT effects of FM, components of the Syk/ TGF-β1/Smad signaling pathway, considered to play a key role in EMT, were investigated via western blot. As shown in Fig. 6, p-Syk levels in HK-2 cells were significantly increased, followed by increase in TGF-β1, p-Smad2, and Smad4 levels. p-Syk, TGF-β1, p-Smad2, and Smad4 expression was markedly decreased upon treatment with 80 µM FM and/or 1 µM BAY61-3606, supporting the theory that the anti-EMT effects of FM are associated with the Syk/TGF-β1/Smad signaling pathway.
Syk-siRNA ameliorates renal EMT by suppressing the Syk/TGF-β1/Smad signaling pathway. To elucidate the mechanism underlying the anti-EMT effects of FM, Syk -siRNA was transfected into HK-2 cells. As shown in Fig. 7a, the gene silencing effect of 1# siRNA was greatest, which was therefore selected for follow-up experiments. Compared with levels in the control group, TGF-β1, p-Smad2, and Smad4 expression in the Syk siRNA group was significantly decreased (Fig. 7b-f), further indicating that the inhibitory effects of FM on EMT are associated with Syk /TGF-β1/Smad signaling.

Discussion
Coreopsis tinctoria Nutt, also known as Kunlun snow chrysanthemum, is predominantly distributed in the South Xinjiang region of China 23 . Earlier studies by our group confirmed that the ethyl acetate extract exerts a renal protective effect in high-glucose-fat diet and streptozotocin-induced diabetic rats 24 . Its component eriodictyol 7-O-β-D glucopyranoside could ameliorate lipid disorders via suppressing lipogenesis and protecting mitochondrial function 25 . In streptozotocin-induced glucose-intolerant rats, the flavonoid-rich fraction exerted a cytoprotective effect on tBHP and cytokine-induced cell injury in MIN6 cells 26 . Extracts of Coreopsis tinctoria Nutt. flower exert antidiabetic effects via inhibiting α-glucosidase activity 27 . As the main active component of Coreopsis tinctoria Nutt, FM has attracted significant research attention. However, the direct target proteins of FM and downstream signaling network are yet to be established. In the present study, 43 potential direct target proteins (including 15 human-related proteins) of FM were predicted using a ligand-protein inverse-docking algorithm, and a PPI network constructed based on the data obtained with the 15 proteins. GO analysis was conducted to enrich the proteins of the PPI network for biological processes. Based on the findings, we hypothesized that Syk plays a critical role in facilitating the effect of FM. Examination of specific interactions via molecular dynamics and SPR analyses confirmed that FM binds Syk in a dose-dependent manner.
Syk, a non-receptor tyrosine kinase, is a member of the protein tyrosine kinase family 28 characterized by intracellular signal cascades, such as activated Fc and B cell receptors 29 . Syk is critical for mediating Fc receptor responses www.nature.com/scientificreports www.nature.com/scientificreports/ in numerous cell types, including mast cells 30 and dendritic cells 31 , and an effective target for the treatment of autoimmune diseases and inflammation. Moreover, Syk is reported to mediate upregulation of HG-induced TGF-β1 in HK-2 cells 22 . Here, we examined the possibility that Syk inhibitors exert an anti-EMT effect in concert with FM. www.nature.com/scientificreports www.nature.com/scientificreports/ FM could reduce the expression of p-Syk caused by high glucose and inhibit the TGF-β1/Smad pathway downstream of Syk. These effects were consistent and synergistic with those of the Syk inhibitor, BAY61-3606. In the CDOCKER program, scores above 20 are considered significant. As shown in Fig. 4, the small molecule inhibitors fit right into the gaps, confirming a high -CDOCKER_INTERACTION_ENERGY score. In view of the finding that FM had a lower binding energy score (−75.1069 kcal/mol) than the Syk inhibitor, LASW836 (−57.4404 kcal/mol), we expect broader prospects for application of FM in the future. Using molecular docking and SPR, we showed that FM could directly bind Syk and thereby exert a major inhibitory effect on its function. www.nature.com/scientificreports www.nature.com/scientificreports/ Phosphorylation of Syk can be activated under high glucose conditions, leading to stimulation of the downstream TGF-β1/Smad pathway. Our experiments additionally demonstrated that FM inhibits activation of the TGF-β1/ Smad pathway induced by high glucose through suppression of Syk activation, leading to an inhibitory effect on EMT in HK-2 cells. The decrease in activity may thus be caused via direct binding and inhibition by FM.
Earlier, Yang et al. 22 explored the role of Syk in HG-induced TGF-β1 upregulation in HK-2 cells. Their experiments showed that Syk serves as a mediator of intracellular signal transduction for TGF-β1 upregulation and is activated by high glucose. As one of the key fibrogenic factors, TGF-β1 is a classical regulatory cytokine and an important inducer of renal tubular epithelial cell transdifferentiation 21,32,33 . TGF-β1 is reported to exert biological effects by activating downstream mediators known as Smads. Smad proteins, the main downstream effector molecules in the TGF-β1 signaling pathway, transduce the TGF-β1 signal from the cell membrane to the nucleus. In the TGF-β1/Smad pathway, Smad proteins are phosphorylated by TGF-β1 and its receptors to regulate the expression of related genes and promote EMT development. Activated Smad3 and Smad2 subsequently bind to Smad4 to form a Smad complex, which is transported to the nucleus and works with other cofactors to regulate www.nature.com/scientificreports www.nature.com/scientificreports/ the transcription of target genes [34][35][36][37] . The signaling cascade from Syk to TGF-β1 and Smads thus appears to be a critical pathway. The findings that the Syk inhibitor, BAY61-3606, and Syk siRNA abrogate the influence of high glucose on EMT expression support the involvement of the signal cascade from Syk to the TGF-β1/Smad axis in the anti-EMT effect.
In conclusion, we propose that a HG-induced signal transduction pathway promotes TGF-β1 expression in HK-2 cells in which Syk plays an important role. Syk activation triggers stimulation of TGF-β1/Smad activity, implicating its involvement in diabetic kidney disease. Overall, a signaling network of FM including its direct target, Syk, was established in this study based on inverse docking and bioinformatics data. Binding of FM to Syk and activation of the downstream signaling cascade may be a critical determinant of the anti-EMT effects of FM. Based on the collective findings, we propose that the renal protective effect of FM in high glucose-exposed HK-2 cells is partially mediated via the Syk /TGF-β1/Smad signaling pathway. Further in vivo studies are necessary to clarify the mechanisms underlying the inhibitory effects of FM on EMT. (c-f) Statistical analysis of western blot data on TGF-β1, p-Smad2 and Smad4 in HK-2 cells. Data are presented as means ± SD, n = 3 (three separate experiments). * P < 0.05, ** P < 0.01, *** P < 0.001, compared with control, and # P < 0.05, ### P < 0.001, compared with control siRNA (control siRNA, negative scramble control siRNA). www.nature.com/scientificreports www.nature.com/scientificreports/

Materials and Methods
Materials and reagents. FM was purchased from Extrasynthese (Z.I Lyon Nord, Genay, France; purity ≧99.0%). BAY61-3606 was purchased from Sigma Chemical Co.(St. Louis, MO, USA; purity ≥98.0%). Human proximal tubular epithelial cells (HK-2; China Center for Type Culture Collection, CCTCC) were maintained in MEM (5.5 mM D-glucose) supplemented with 10% (v/v) fetal bovine serum (FBS), 100 μg/mL streptomycin and 100 U/ml penicillin at 37 °C with 5% CO 2 . To induce EMT, HK-2 cells were cultured for 48 h in high-glucose medium (60 mM D-glucose was selected based on preliminary experiments 38 and previously published findings 39 ). High mannitol as an osmotic control exerted minimal or no effects on EMT factors and TGF-β1 39,40 . Thus, no osmotic control was required. Upon reaching 60-70% confluence, HK-2 cells were cultured in serum-free medium for 24 h prior to experimentation. All primary antibodies were obtained from Abcam (Cambridge, MA, USA).

Determination of the effects of FM on viability and proliferation of HK-2 cells under normal
and high glucose conditions. The MTS cell proliferation assay (Promega, Madison, WI, USA) was used to determine cell viability. In total, 5000 HK-2 cells were seeded on 96-well plates in complete medium for 24 h and incubated with various concentrations of FM for 24 h and 48 h in serum-free culture medium (normal conditions) or FM in 60 mM D-glucose for 48 h in serum-free culture medium (high-glucose conditions). MTS solution (20 μL) was added to each well, followed by incubation for 2 h at 37 °C in a humidified atmosphere containing 5% CO 2 . The resulting color was assayed at 490 nm using a microplate reader (Bio-Rad Laboratories, Inc., Hercules, CA, USA). Network construction and analysis. To determine the key proteins contributing to the anti-EMT function of FM, a PPI network was constructed based on the results of reverse target screening. In total, 15 human proteins were identified in reverse target screening and visualized using Cytoscape 3.6.1 platform BisoGenet Version 3.0.0 (http://www.cytoscape.org), as described previously. Subsequently, GO analysis was conducted to enrich proteins of the PPI network for biological processes.

Confirmation of the FM target. Molecular docking and dynamic simulation. The CDOCKER program
in Discovery Studio ™ 2.5 (DS; Accelrys Software Inc., San Diego, CA, USA) based on the CHARMm force field was utilized to achieve molecular docking results. The Syk protein was obtained from Protein Data Bank (PDB ID: 3vf9, http://www.rcsb.org/) and the FM molecular preparation created according to the Zink program (http:// zinc15.docking.org/). Before the docking procedure, a force field was applied and binding energy minimized. Next, a ligand-based similarity search scheme was employed and the docking protocol performed as a default setting to avoid a potential reduction in docking accuracy. The complex was solvated with water in a cubic box with an explicit periodic boundary model to stimulate the environment within the cell. After creating a harmonic restraint, a standard dynamics cascade was performed, including heating, minimization, equilibration and production dynamics 41,42 .