Tangshen Formula is made up of seven kinds of herbs comprising astragalus root, burning bush twig, rehmannia root, bitter orange, cornus fruit, rhubarb root and rhizome, and notoginseng root. It was prepared by Beijing Yadong Biopharmaceutical (Beijing, China), and the standardization of the formula was in accordance with the Pharmacopoeia of The People's Republic of China 2015 (11).

Eighteen diabetic male KKAy mice (8 weeks old, 20–25 g) and six male C57BL/6J mice were purchased from Beijing Vital River Laboratory Animal Technology (Beijing, China). All animals were housed at 23 ± 3°C and 55 ± 15% humidity with 12:12 h light/dark cycle and were allowed access to chow and water ad libitum. KKAy mice were treated with a high-fat diet, and their blood glucose was measured once in 2 weeks by One Touch Ultra blood glucose monitoring system (LifeScan, Milpitas, CA, USA). The KKAy mice were randomly divided into the following three groups: (1) KKAy group: KKAy mice-administered distilled water (n = 6); (2) KKAy+TSF group: KKAy mice-administered TSF (3.57 g/kg per day) by oral gavage (n = 6) (12); and (3) KKAy+Irbesartan group: KKAy mice-administered irbesartan by oral gavage (22.5 mg/kg per day) (n = 6). The control group consisted of healthy C57BL/6J mice-administered distilled water (n = 6).

All the animals were weighed once a week. After 16 weeks, all mice were sacrificed under anesthesia. Serum was collected for assay. The apex of the heart from each animal was used for a pathologic analysis, and the rest of the heart tissue was used for a molecular biologic analysis. Animal care and treatments were in accordance with the NIH Guiding Principles for the Care and Use of Laboratory Animals, and the protocol was approved by the Beijing National Proteome Science Center Animal Management Ethics Committee.

The apex of the heart tissue was fixed with 4% phosphate buffered saline (PBS) buffered paraformaldehyde, embedded in paraffin, and then sectioned into 3 μm thicknesses and stained with Masson trichrome according to the standard procedure. Masson trichrome staining causes muscle fibers to turn red and collagen fibers to turn blue. Immunohistochemistry (IHC) staining was used to detect the distribution and expression of biomarkers for fibrosis. We used a microwave-based heating antigen retrieval and 3% H₂O₂ to block non-specific staining. The antibodies used included: collagen I (1310-01, SouthernBiotech, Birmingham, AL, USA; 1:200 dilution), collagen III (GB111629, Servicebio, Wuhan, China; 1:200 dilution), α-SMA (GM085102, Gene Tech, Shanghai, China; 1:200 dilution), phosphorylated Smad2/3 (sc-11769, Santa Cruz Biotechnology, Dallas, TX, USA; 1:200 dilution), and anti-β-catenin (610154, BD Transduction Laboratories, Shanghai, China; 1:50 dilution). Images were quantitatively analyzed using Image-Pro Plus v 6.0 (Media Cybernetics, Rockville, MD, USA). Briefly, the accumulation of collagens I and III and α-SMA in 10 random areas, except staining around blood vessels, was analyzed under 200× for the percentage of positive areas in the examined field. Numbers of p-Smad2/3+ cells were counted in 10 random areas for each sample under 400×.

Heart ultrastructure was detected using transmission electron microscopy (JEOL-100CXII; JEOL, Tokyo, Japan). The heart was sectioned into 1 mm cubes and fixed with 2.5% glutaraldehyde at 4°C for 24 h, and then embedded in an epoxy resin. Ultrathin sections were sliced and stained with uranyl acetate and lead citrate at room temperature, for 30 and 15 min, respectively. Twenty images per sample were randomly selected and observed, and the ultramicrostructure of the myocardium was observed under the magnification ×12,000.

Cardiac fibroblasts were derived from the hearts of neonatal mice. The hearts were removed from the C57BL/6J newborn mice (Beijing Vital River Laboratory Animal Technology, Beijing, China) and, under aseptic conditions, were sectioned into tissue blocks of about 1 mm³. The tissue blocks were washed with Hanks Balanced Salt solution without Ca²⁺ and Mg²⁺ and digested at 37°C with 0.08% trypsin and 0.06% collagenase. The digested supernatant was collected and neutralized with low glucose (5 mmoL/L) Dulbecco Modified Eagle medium/fatty acid (DMEM/FA) containing 20% serum until the tissue mass became transparent and digestion was terminated. The resulting suspension was filtered with a 200-μm BD Falcon cell strainer (Corning Life Sciences, Corning, NY, USA) and centrifuged from 4 to 10 min at 25–300 × g. The filtered solution was then plated onto culture dish.

Primary mice myocardial fibroblasts were cultured in DMEM with 10% fetal bovine serum (GIBCO, Carlsbad, CA, USA) at 37°C with 5% CO₂. To investigate the role of TGF-β1, the fibroblasts were exposed to TGF-β1 for 6, 16, 24, 36, and 72 h sequentially at 0, 2.5, 5, 10, and 15 μg/ml, respectively. After the TGF-β1-stimulated myocardial fibroblast injury model was successfully obtained, the cells were treated with TSF for 72 h at the concentrations of 100 and 250 μg/ml.

An MTT assay was used to detect whether TSF had an effect on cell viability. Being starved overnight with DMEM/F12 medium without fetal bovine serum (FBS), C57BL/6J mouse primary myocardial fibroblasts were then grown in 96-well plates and incubated with TSF at the concentrations of 0, 100, 250, 500, 750, and 1,000 μg/ml for 72 h. Each well was added with 1/10 volume of MTT solution and incubated at 37°C with 5% CO₂ for 4 h. After adding formazan lysate, the assay was performed at 490 nm wavelength with a microplate analyzer (Thermo Fisher Scientific, Waltham, MA, USA). The experimental methods were completed according to the kit instructions, and cell viability was calculated.

Total RNA was isolated from mouse hearts using the TriZol method. RNA concentration was then quantified using a NanoDrop-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA). Total RNA was reverse transcriptased to the complementary DNA (cDNA) template using the M5 SuperFast qPCR RT kit (Ju Hemei, Beijing, China) quantitative PCR (qPCR) was performed with the QuantStudio 5 Real-Time PCR system (Applied Biosystems, Waltham, MA, USA) using the IQ SYBR Green Supermix reagent (CWBIO, Beijing, China), as previously described (13). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as the internal control to normalize the gene expression in the same cDNA. This was followed by applying the ΔΔC(t) threshold cycle method, to analyze the qPCR data (14). The specific primer sequences used are listed in Table 1.

Mouse heart tissues were homogenized in a mixture of radioimmunoprecipitation assay (RIPA) buffer and 1% cocktail (Bimake, Houston, TX, USA) to extract the protein. Protein concentration was measured using the Bicinchoninic acid assay (BCA) Protein Assay kit (Epizyme Biotech, Shanghai, China) and then was denatured. The proteins were separated by 12% sodium dodecyl sulfate (SDS) and transferred to polyvinylidene difluoride (PVDF) membranes. After blocking in 5% skim milk for 1 h, the membranes were incubated with primary antibody at 4°C overnight. The following day, the membranes were washed three times with tris-buffered saline and Tween-20 (TBST) and incubated with secondary antibody. Membrane immunostaining was observed and analyzed using the ChemiDoc XRS system (Bio-Rad, Hercules, CA, USA). Signals were quantified by the Image J program (National Institutes of Health, Bethesda, MD, USA). β-actin normalization detected the protein ratio and was expressed as mean ± SME.

The primary and secondary antibodies were: TGF-β1 (1:1,000 dilution; BS1361, Bioworld Technology, Nanjing, China), smad7 (1:500 dilution; sc-9183, Santa Cruz Biotechnology), anti-rabbit (1:5,000 dilution; S8002, Beijing Guanxingyun Sci and Tech, Beijing, China), anti-mouse (1:5,000 dilution; S8001, Beijing Guanxingyun Sci and Tech, Beijing, China), anti-goat (1:2,000 dilution; ab205723, abcam, Cambridge, UK), FN (1:500 dilution; sc-8422, Santa Cruz Biotechnology), wnt1 (1:1,000 dilution; AF5315, Affinity Biosciences, Cincinnati, OH, USA), β-catenin (1:500 dilution; sc-7963, Santa Cruz Biotechnology), active-β-catenin (1:500 dilution; 05-665, Merck Millipore, Darmstadt, Germany), and MMP7 (1:1,000 dilution; ER1913-08, Huabio, Woburn, MA, USA).

All data collected from this study were expressed as means ± SEM. The statistical analysis was performed and viewed with the GraphPad Prism software (LaJolla, CA, USA), and differences among the means were assessed using a one-way ANOVA.

During the 16 weeks of treatment with TSF or irbesartan, there were no deaths observed in KKAy mice. Body weights of all animals were measured once a week, and weight gain was recorded (Figure 1A). Weight was significantly increased in the KKAy group compared to the C57BL/6J group. However, in the treatment groups, weights were significantly decreased due to TSF or irbesartan therapy. After 16 weeks of treatment, blood glucose levels in KKAy+TSF and KKAy+irbesartan groups were lower than the levels in the KKAy group, but without a significant difference (Figure 1B).

Irbesartan ameliorates myocardial fibrosis in diabetic cardiomyopathy (15). It is an angiotensin domain receptor inhibitor and is considered to be a common medication for improving the structure and function of cardiomyocytes and reducing the risk of death from heart failure (16). Therefore, irbesartan was used as a positive control group to investigate the effects of drug therapy.

Masson trichrome staining revealed a large number of blue collagen fibers in the KKAy group compared with the C57BL/6J group. In contrast, in TSF and irbesartan groups, the blue collagen area was significantly reduced (Figure 1C). Moreover, quantitative data demonstrated that the treatment with TSF for 16 weeks significantly attenuates these histologic injuries.

Under electron microscopy, the ultrastructure of myocardial cells was observed and the mitochondria remained intact in the C57BL/6J group. In the KKAy group, the myocardial structure was diminished and mitochondrial vacuolization and crest dissolution were evident. In the KKAy+TSF group, the myocardial ultrastructural injury was alleviated though mitochondria did exhibit slight vacuolization. There were no obvious pathologic changes in the KKAy+TSF and KKAy+ irbesartan group compared with the C57BL/6J group (Figure 1D).

One of the main symptoms of myocardial fibrosis is the accumulation of collagen in the ECM. Collagen I account for about 80% of the collagen in the heart muscle and increases in myocardial fibrosis. Collagen III is also an important component of the cardiac ECM (17). α-SMA can be assembled into stress fibers to reshape the surrounding ECM. Thus, in our study, we examined the therapeutic effect of TSF on myocardial fibrosis in KKAy diabetic mice. In the heart samples, the immune-positive areas of collagens I and III and α-SMA were yellow-brown (Figure 2A). IHC revealed that KKAy mice developed myocardial fibrosis, including a significant upregulation of collagens I and III as well as α-SMA, which were attenuated by the treatment with TSF or irbesartan (Figure 2B). Similarly, western blotting revealed that the upregulation of collagens III and α-SMA were significantly decreased after TSF treatment (Figure 2C).

Real-time qPCR and western blot revealed that KKAy mice developed myocardial fibrosis, including a significant upregulation of TGF-β1 at the messenger RNA (mRNA) and protein levels as well. This upregulation was inhibited after TSF treatment (Figure 3A). IHC results of p-Smad2/3 and western blot analyses of p-Smad2/3 showed that the activation of TGF-β/Smad2/3 signaling in the KKAy group and its effect were reduced after TSF treatment (Figure 3B). The protein expression level of smad7 was markedly decreased in KKAy mice compared with the C57BL/6J group (p < 0.01), whereas the protein expression was significantly increased in the TSF group compared with the KKAy group (Figure 3A). No significant difference was found between C57BL/6J and TSF groups. These results indicated that TSF is effective in inhibiting the TGF-β/Smad pathway in KKAy mice.

To confirm the effect of TSF on Wnt/β-catenin signaling, the western blot analysis was used to detect Wnt1 and β-catenin and their downstream target MMP7, FN. Immunohistochemical analysis of anti-β-catenin showed that the relative expression of anti-β-catenin was upregulated in the KKAy group and was significantly reduced after the treatment with TSF, which demonstrated the activation of Wnt/β-catenin (Figure 4A). The presence of the bands only at the target protein sites indicates that the staining is specific. The expression of Wnt1, β-catenin, and active-β-catenin was increased in the KKAy mice compared with the C57BL/6J group, demonstrating that the Wnt/β-catenin pathway was active in the KKAy animal model. TSF decreased the expression of Wnt1, β-catenin, and active-β-catenin proteins in the TSF treatment group compared with the KKAy group. Irbesartan also showed an inhibitory effect on Wnt1 and active-β-catenin expressions. A significant difference in the levels of MMP7 and FN expressions was found between the C57BL/6J group and KKAy group (p < 0.01) as well as between the KKAy group and TSF group. In the KKAy group, the expressions of MMP7 and FN significantly increased compared with the C57BL/6J group. In contrast, TSF treatment decreased the expression levels of MMP7 and FN (Figure 4B), suggesting that the treatment with TSF attenuates diabetic myocardial fibrosis via the Wnt/β-catenin mechanism. Taken together, this study highly suggests that TSF has an inhibitory effect on the Wnt/β-catenin signaling pathway.

Combined with protein and mRNA expressions of α-SMA and collagen I, the optimal myocardial fibroblast fibrotic injury model was obtained under the condition of 5 ng/ml TGF-β1 induced for 72 h (Figures 5A–C). To determine whether TSF affected cell proliferation, MTT assays were used to investigate the cytotoxicity of TSF (Figure 5D). The results showed that the cell survival rate decreased gradually as TSF concentration increased. Thus, the concentrations of 100 and 250 μg/ml TSF were chosen for the in vitro experiments.

Cardiac fibroblasts induced by TGF-β1 showed an upregulation of mesenchymal markers, including collagens I and III, and α-SMA (Figure 5E), which were all significantly downregulated after 100- and 250-μg/ml TSF treatment.