Carthamus tinctorius L. was grown in Tacheng, Xinjiang Uygur Autonomous Region, China, and the flowers (safflower) were provided by Huahui Kaide Pharmaceutical Co., Ltd (Shanxi, China) and identified by Professor Jiashi Li (Beijing University of Traditional Chinese Medicine). HSYA was isolated and purified from the aqueous extract of C. tinctorius L. by macroporous resin-gel column chromatography, as described previously (Zang et al., 2008). The molecular weight of HSYA is 612 and the molecular structure has been previously described by Feng et al. (2013). HSYA was diluted with aseptic 0.9% NaCl for experimental use. HSYA analysis was performed using a high-performance liquid chromatography system (Wang Y. et al., 2014). The purity of the HSYA was previously determined by the area normalization method to be 95.2% (Pan et al., 2016).

Recombinant human TGF-β1 (PeproTech Company, Rocky Hill, NJ, USA) was dissolved in phosphate-buffered saline (PBS) containing 5% trehalose. The TGF-β receptor type I kinase inhibitor, SB431542 (Sigma–Aldrich, St. Louis, MO, USA), was diluted with dimethylsulfoxide (DMSO) to a final concentration of DMSO of 0.1% in the reaction system. Fluorescein isothiocyanate (FITC)-labeled TGF-β1 was synthesized and purified by dialysis against PBS pH 7.4 by Beijing Fanbo Biochemicals Co. Ltd, as described previously (Xia et al., 1996). TGF-β type II receptor (TβRII) siRNA, non-targeting control siRNA, siRNA transfection reagent, α-smooth muscle actin (α-SMA), fibronectin (FN), TβRII, and TGF-β type I receptor (TβRI) antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Collagen I α 1 (COL1A1) antibody was from Novus Biological (Littleton, CO, USA). Human matrix metalloproteinase-2 (MMP-2), tissue inhibitor of matrix metalloproteinase-1 (TIMP-1), and TIMP-2 antibodies were from R&D Systems (Minneapolis, MN, USA). Phospho-Smad2 (Ser465/467), phospho-Smad3 (Ser423/425), Smad2/3, phospho-p44/42 mitogen-activated protein kinase (MAPK) (extracellular signal-regulated kinase 1/2; ERK1/2) (Thr202/Tyr204), and p44/42 MAPK (ERK1/2) primary antibodies were from Cell Signaling Technology (Danvers, MA, USA). The 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfo-phenyl)-2H-tetrazolium inner salt (MTS) assay kit was from Promega (Madison, WI, USA), the SYBR^® Premix Ex Taq^TM (Perfect Real Time) kit was from Agilent Technologies (Santa Clara, CA, USA), and glyceraldehyde phosphate 3-dehydrogenase (GAPDH) antibody, RIPA lysis buffer, phenylmethanesulfonyl fluoride, and the bicinchoninic acid assay kit were from Beyotime Institute of Biotechnology (Jiangsu, China).

MRC-5 human fetal lung fibroblasts were obtained from the Institute of Basic Medicine, Chinese Academy of Medical Sciences, and cultured in minimum essential medium (Thermo Scientific, Walton, MA, USA) supplemented with 10% fetal bovine serum (Thermo Scientific), 1% non-essential amino acids (Keygen, China), 100 g/mL streptomycin, and 100 U/mL penicillin (Thermo Scientific) at 37°C, in a 5% CO₂ incubator.

In the first part of the study, we assessed the effects of HSYA on TGF-β1-induced cell proliferation, migration, and expression levels of MMP-2, TIMP-1, TIMP-2, TβRII, and TβRI. Cells were divided into eight groups as follows: normal control; HSYA blank control (45 μmol/L); TGF-β1; TGF-β1+HSYA (5, 15, and 45 μmol/L); TGF-β1+SB431542; and TGF-β1+SB431542+HSYA (45 μmol/L). Cells in the TGF-β1+SB431542+HSYA (45 μmol/L) group were treated with HSYA, SB431542, and TGF-β1. In the TGF-β1+SB431542 group, an equivalent volume of 0.9% NaCl was added instead of HSYA; an equivalent volume of DMSO was added instead of SB431542 in the TGF-β1+HSYA (5, 15, and 45 μmol/L) groups; and equivalent volumes of 0.9% NaCl and DMSO were added instead of HSYA and SB431542, respectively, in the TGF-β1 group. In the HSYA blank control, equivalent volumes of PBS containing 5% trehalose and DMSO were added instead of TGF-β1 and SB431542, and the normal control included equivalent volumes of 0.9% NaCl, PBS containing 5% trehalose, and DMSO instead of HSYA, TGF-β1, and SB431542, respectively.

The second part of the experiment was performed to confirm the inhibitory effects of HSYA on TGF-β1-induced α-SMA, COL1A1, and FN expression. For this, six treatment groups were established as follows: normal control; HSYA blank control (45 μmol/L); TGF-β1; TGF-β1+HSYA (45 μmol/L); and TGF-β1+SB431542. Cell treatments were as above.

The third part of the experiment was conducted to detect the effects of HSYA on expression levels of α-SMA, COL1A1, and FN, and on phosphorylation of Smad2, Smad3, and ERK following TβRII knockdown. Cells were divided into nine groups as follows: normal-untransfected; normal-non-targeting control siRNA; normal-TβRII siRNA; TGF-β1-untransfected; TGF-β1-non-targeting control siRNA; TGF-β1-TβRII siRNA; (TGF-β1+HSYA)-untransfected; (TGF-β1+HSYA)-non-targeting control siRNA; and (TGF-β1+HSYA)-TβRII siRNA. The normal-untransfected, TGF-β1-untransfected, and (TGF-β1+HSYA)-untransfected groups were treated the same as the normal, TGF-β1, and TGF-β1+HSYA (45 μmol/L) groups in the first part of the experiment, respectively. The normal-TβRII siRNA, TGF-β1-TβRII siRNA, and (TGF-β1+HSYA)-TβRII siRNA groups first underwent TβRII knockdown, and were then treated like the normal, TGF-β1, and TGF-β1+HSYA (45 μmol/L) groups in the first part of the experiment, respectively. The normal-non-targeting control siRNA, TGF-β1-non-targeting control siRNA, and (TGF-β1+HSYA)-non-targeting control siRNA groups received non-targeting control siRNA instead of TβRII siRNA.

Cells were seeded in 96-well plates at a density of 3 × 10³ cells/well and cultured for 24 h. After 12 h of serum starvation, the cells were pretreated with various doses of HSYA or 2 μmol/L SB431542 before stimulation with 1 ng/mL TGF-β1. A blank well with the same volume of medium but without cells was treated simultaneously, as a control well. After incubation for 24, 48, or 72 h, 20 μL CellTiter 96 Aqueous One Solution Reagent was added for 1 h at 37°C, and the optical density (OD) at 490 nm was then measured using a microplate reader (BioTek, Winooski, VT, USA).

To establish a grid for orientation, in each well of a six-well plate, we first drew five lines parallel to the long edges of the plate. Then cells were cultured in these labeled six-well plates for 24 h to reach 80–90% confluency. After serum starving for 12 h, the cell layer in the middle of the well was scratched perpendicularly to the five lines using a pipette tip and a wounded region formed. The wounded region and five transverse lines crossed each other to form a grid which was used to follow cell migration in the same field over time. The size of the regions within each grid was approximately the same as the microscopic field of view at 50× magnification. Subsequently, cells were washed twice with PBS and then treated with various doses of HSYA or 2 μmol/L SB431542 for 30 min before stimulation with TGF-β1. Images of the wounded region from the same sector of the grid were acquired at 0, 24, and 48 h following treatment using an inverted microscope (Leica, Germany). The area and height of the rectangular wounded region were calculated using Image J software. The area divided by height gave the width. Cell migration rates were calculated as follows:

MRC-5 cells were grown on sterile glass slides for 24 h in 6-well plates to reach 80–90% confluency and then serum starved for 12 h. At that time, cells were treated with 45 μmol/L HSYA or 2 μmol/L SB431542 for 30 min before stimulation with TGF-β1 for 72 h. After washing three times with cold PBS, the cells were immobilized with 4% paraformaldehyde (Thermo Scientific) for 15 min and permeabilized by 0.1% Triton X-100 in PBS for 10 min. The cells were then incubated for 30 min in blocking buffer (3% bovine serum albumin/PBS) at 37°C and stained with primary antibodies recognizing α-SMA, COL1A1, and FN diluted in blocking buffer at 4°C overnight, followed by incubation with FITC-labeled goat anti-rabbit or anti-mouse antibody (Beyotime Institute of Biotechnology) at 37°C for 2 h. Cell nuclei were stained with DAPI (Beijing, China) at room temperature for 5 min. Immunofluorescent images were captured using a fluorescence microscope (Nikon 80i, Japan).

Cells were plated in six-well plates for 24 h to reach 80–90% confluency and then incubated with serum-free medium for 12 h. Five microliters of siRNA (TβRII or non-targeting control) solution were mixed with 100 μL siRNA transfection medium to create reagent A, and 5 μL transfection reagent was mixed with 100 μL siRNA transfection medium to create reagent B. Reagents A and B were then mixed to give reagent C and kept at room temperature for 30 min. After washing with transfection medium, 200 μL reagent C and 800 μL transfection medium were added to the cells and incubated at 37°C for 6 h. One milliliter of medium containing 20% fetal bovine serum was then added to each well and the cells were incubated for 24 h, followed by the addition of 45 μmol/L HSYA, incubation for 30 min, and stimulation with TGF-β1.

Total RNA was isolated from MRC-5 cells using TRIzol reagent (Invitrogen, Carlsbad, CA, USA), according to the manufacturer’s instructions. The RNA concentration and quality were measured using a NanoDrop 2000 (Thermo Scientific, Wilmington, DE, USA). Two micrograms of RNA were reverse transcribed to cDNA using a cDNA reverse transcriptase kit (Promega, Madison, WI, USA). Real-time PCR was performed using a SYBR^® Premix Ex Taq^TM Kit with a Bio-Rad iCycler iQ5 Detection System (Bio-Rad, Hercules, CA, USA). The PCR amplification conditions were as described previously (Pan et al., 2016). The primer sequences and lengths (in bp) were as follows (5′–3′): MMP-2: TGAGCTATGGACCTTGGGAGAA (forward) and CCATCGGCGTTCCCATAC (reverse), 60 bp; TIMP-1: CTGTTGTTGCTGTGGCTGATAG (forward) and AAGGTGGTCTGGTTGACTTCTG (reverse), 137 bp; TIMP-2: GGGAAGGATTTTGGAGGTAGG (forward) and GGAGGCTGAGAAAGAAGTGAGTG (reverse), 58 bp. The threshold cycle (Ct) was automatically calculated and displayed in the PCR instrument after each cycle. Relative quantities of mRNA were calculated using the 2^-ΔΔCt method, normalized to the GAPDH gene.

MRC-5 cells were lysed with RIPA lysis buffer mixed with 1 μM phenylmethanesulfonyl fluoride and 1 μM phosphatase inhibitor cocktail 3 (Sigma–Aldrich). Cell protein concentrations were measured using a bicinchoninic acid assay kit. Protein samples were isolated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and then transferred to nitrocellulose membranes. After blocking with 5% non-fat milk, the membranes were incubated at 4°C overnight with antibodies recognizing the following antigens: α-SMA, COL1A1, FN, MMP-2, TIMP-1, TIMP-2, TβRII, TβRI, p-Smad2, p-Smad3, p-ERK, Smad2/3, ERK, and GAPDH (1:500–1000 dilution). After three washes with Tris-buffered saline-Tween, the membranes were incubated with IR Dye^®-conjugated goat anti-mouse, goat anti-rabbit, or donkey anti-goat antibodies (1:5000 dilution, LI-COR Biosciences, Lincoln, NE, USA) for 1 h. The band intensities were scanned and quantified using an Odyssey infrared imaging system (Gene Company, Beijing, China).

MRC-5 cells at a density of 1 × 10⁶ cells/well were divided into six groups: normal; FITC-TGF-β1 (20 ng/μL); FITC-TGF-β1 (20 ng/μL)+HSYA (5, 15, or 45 μmol/L); and FITC-TGF-β1 (20 ng/μL)+TGF-β1 (2 μg/μL). Cells in the FITC-TGF-β1+HSYA group were treated with HSYA and FITC-TGF-β1. In the FITC-TGF-β1+TGF-β1 group, HSYA was replaced with TGF-β1 solution (final concentration 2 μg/μL); in the FITC-TGF-β1 group, HSYA was replaced with 0.9% NaCl; and in the normal group, HSYA and FITC-TGF-β1 were replaced with 0.9% NaCl and PBS, respectively. After the reagents were added, the cell suspensions were incubated at 37°C for 40 min with shaking at 300 rpm using an orbital shaker (Eppendorf, Germany). The cells were subsequently cooled at 4°C for 10 min, centrifuged at 1000 rpm for 5 min, and washed twice with PBS. The precipitate was then suspended in 300 μL PBS and detected by flow cytometry (Beckman Coulter, Brea, CA, USA). The mean fluorescence intensity (gate %) showed the specific binding intensity of FITC-TGF-β1 to MRC-5 cell membrane receptors.

Data analysis was performed using SPSS 13.0 (SPSS Inc., Chicago, IL, USA) and GraphPad Prism 5.0 (GraphPad Software, Inc., USA) software. Multiple comparisons were made by one-way analysis of variance (ANOVA) with Student–Newman–Keuls post hoc tests. Data were presented as mean ± SD. A p-value < 0.05 was defined as statistically significant.

Previously, we showed that MRC-5 cells treated with HSYA for 72 h could inhibit TGF-β1-induced cell proliferation in a dose-dependent manner (Pan et al., 2016). To determine how early this effect appeared, we monitored the effect of HSYA on TGF-β1-induced cell proliferation after 24, 48, or 72 h. As depicted in Supplementary Figure S1, the OD measure of proliferation of MRC-5 cells was lower in the TGF-β1+HSYA treated group as compared to the group treated with TGF-β1 alone after 48 or 72 h. This effect was dose dependent. However, after treatment with 5, 15, and 45 μmol/L HSYA for 24 h, we could not find any significant change in the OD value. Further, the OD values for the TGF-β1+SB431542 treated cells were comparable to those of the TGF-β1+SB431542+HSYA treated group, suggesting that the effect of HSYA was dependent on the TGF-β receptor type I kinase pathway.

Because HSYA affected the TGF-β1 induced proliferation of MRC-5 cells, we hypothesized that it might also influence other TGF-β1-mediated processes like cell migration. To that end, we assessed the effects of HSYA on TGF-β1-induced cell migration by wound-healing assay (Figure 1A). Consistent with our cell proliferation data, there was no significant difference in cell migration between the TGF-β1+HSYA and TGF-β1 groups at 24 h (Figure 1B). However, when the wound was examined after 48 h, we found that the TGF-β1-induced migration of MRC-5 cells was significantly inhibited when treated with HSYA as compared to those cells treated with TGF-β1 alone (Figure 1C). Further, cell migration was similar in the TGF-β1+SB431542 and TGF-β1+SB431542+HSYA groups, suggesting that the effect of HSYA was dependent on the TGF-β receptor type I kinase pathway. Taken together, our data show that treatment of MRC-5 cells with HSYA for 48 h can attenuate both the TGF-β1-induced proliferation and migration response, although this effect is not apparent at 24 h.

The pathology of IPF is thought to arise, in part, from the conversion of fibroblasts into myofibroblasts, which express α-SMA and secrete ECM proteins. Increased deposition of ECM and reduced matrix degradation are the underlying causes of IPF. We have shown that HSYA decreased the elevated mRNA and protein expression of α-SMA, COL1A1, and FN induced by TGF-β1 (Pan et al., 2016). Building upon this observation, we performed immunofluorescent staining of MRC-5 cells treated with HSYA, with TGF-β1, with TGF-β1+HSYA, or with TGF-β1+SB431542 and compared the expression of α-SMA, COL1A1, and FN to untreated cells. As seen in Supplementary Figure S2, TGF-β1 treatment increased the intracellular incorporation of α-SMA-positive filaments into MRC-5 cells and strongly upregulated the expression of COL1A1 and FN staining along the cell axis as compared to untreated cells. This change was abrogated with either TGF-β1 inhibition (via SB431542 treatment) or with HSYA treatment, indicating that HSYA can modulate the TGF-β1-induced differentiation of MRC-5 cells into myofibroblast-like cells as well as the resulting increase in synthesis of ECM proteins.

To determine whether HSYA could influence the degradation of the ECM, we measured the protein levels of MMP-2, TIMP-1, and TIMP-2 by western blotting. As shown in Figure 2, MMP-2 protein levels increased when MRC-5 cells were stimulated with TGF-β1 for 72 h compared with the normal group. However, there was no significant increase in TIMP-1 or TIMP-2 expression compared with the normal group. HSYA treatment had no effect on MMP-2, TIMP-1, or TIMP-2 protein expression levels. In total, our data indicate that while HSYA can attenuate the TGF-β1-induced transdifferentiation of MRC-5 fibroblasts into myofibroblast-like cells and associated synthesis of ECM proteins, it has no effect on the degradation of the ECM.

Protein levels of TβRI increased after activation by TGF-β1, while TβRII expression levels were not obviously increased. HSYA treatment had no effect on the expression of TβRII or TβRI (Figure 3).

We performed knockdown of TβRII with siRNA to evaluate if HSYA suppressed TGF-β1-induced cell transdifferentiation and ECM accumulation by targeting TβRII on the cell membrane. Protein levels of α-SMA, COL1A1, and FN were significantly increased in the TGF-β1-untransfected group compared with the normal-untransfected group, and these increases were attenuated in the (TGF-β1+HSYA)-untransfected group (Figure 4). These protein levels were also markedly elevated in the TGF-β1-non-targeting control siRNA compared with the normal-non-targeting control siRNA group, and this elevation was decreased in the (TGF-β1+HSYA)-non-targeting control siRNA group. These data suggest that HSYA can suppress α-SMA, COL1A1, and FN expression induced by TGF-β1. α-SMA, COL1A1, and FN expression were also significantly decreased in the TGF-β1-TβRII siRNA group compared with the TGF-β1-non-targeting control siRNA group, suggesting that these protein levels were decreased by TβRII knockdown. Importantly, there was no significant difference between the TGF-β1-TβRII siRNA and (TGF-β1+HSYA)-TβRII siRNA groups, suggesting that HSYA target TβRII to suppress α-SMA, COL1A1, and FN expression in TGF-β1-treated MRC-5 cells.

We further explored whether HSYA may target TβRII on the cell membrane to attenuate TGF-β1-induced activation of TGF-β1/Smad and ERK/MAPK signaling pathway using TβRII siRNA. Phosphorylation levels of Smad2, Smad3, and ERK were increased in both the TGF-β1-untransfected groups and TGF-β1-non-targeting control siRNA groups (Figure 5), and these levels were reduced in the (TGF-β1+HSYA)-untransfected groups and (TGF-β1+HSYA)-non-targeting control siRNA groups, suggesting that HSYA inhibited the phosphorylation of Smad2, Smad3, and ERK induced by TGF-β1. Meanwhile, Smad2, Smad3, and ERK phosphorylation levels were significantly decreased in both the TGF-β1-TβRII siRNA and (TGF-β1+HSYA)-TβRII siRNA groups, with no significant difference between them. This suggested that HSYA targets TβRII to suppress the phosphorylation of Smad2, Smad3, and ERK in TGF-β1-treated MRC-5 cells.

Fluorescein isothiocyanate-TGF-β1 binding to receptors on MRC-5 cells was measured by flow cytometry. The mean fluorescence intensity (gate %) was increased significantly in the FITC-TGF-β1 compared with the normal group, and this elevation was attenuated in the FITC-TGF-β1+HSYA (45 μmol/L) group (Figure 6). The gate % was also attenuated by unlabeled TGF-β1, at a concentration 100 times that of FITC-TGF-β1 in the FITC-TGF-β1+TGF-β1 group. These results suggested that HSYA inhibited FITC-TGF-β1 binding to MRC-5 cell membrane receptors.