Compound turtle shell softening liver pills (TSP) (batch number: 20090107) were bought from the Inner Mongolia Fu Rui Medicine Company, dissolved in distilled water to produce a solvent containing 0.06 g/ml of crude drug. The materials to prepare NTSD included ground beetle, turtle shell, bupleurum root, Scutellaria root, Pinellia, Artemisia capillaris, cassia twig, peach kernel, Poria, Radix Astragali and white peony root (Chongqing Tongjunge Pharmacy, Chongqing, China). The above materials were soaked in distilled water for 1 h, decocted (turtle shell, ground beetle first), filtered and concentrated, to make the decoction. The different crude drug dosages of NTSD in distilled water for application were 2.84, 1.42 and 0.71 g/ml, respectively. All the above preparation work was completed in the Department of Pharmacy, Chongqing Southwest Hospital, China.

Healthy male specific pathogen free Sprague-Dawley (SPF SD rats) were purchased from the Experimental Animal Center of Daping Hospital, Affiliated to the Third Military Medical University (Chongqing, China) and were fed in the SPF animal laboratory at temperatures between 18 and 22°C and a relative humidity of 50–80%, with free access to a standard diet and sterile water. HSC-T6 cells were cultured in 5% CO₂ saturated humidity environment at 37°C using DMEM culture medium (containing 10% FBS, 100 U/ml penicillin and 100 U/ml streptomycin). The present study was approved by the Institutional Review Board of Southwest Hospital, the Third Military Medical University, and the methods were carried out in accordance with the approved guidelines.

From an initial 90 rats, in 80 rats HF was induced by CCl₄ (Chengdu Kelong Chemical Factory, Chengdu, Sichuan, China) mixed with olive oil (Sichuan Tianyuan Olive Oil Co., Sichuan, China) to a final CCl₄ concentration of 40%. The solution was injected as 5 ml/kg subcutaneously into the right hind leg, followed by repeated injections (3 ml/kg), twice a week for 8 weeks. Two weeks later, we randomly selected 5 HF rats for H&E staining of liver tissue in order to confirm the formation of HF. The remaining 75 rats were 1 week later randomly divided into groups treated with low (7.1 g/kg/day), medium (14.2 g/kg/day) and high (28.4 g/kg/day) doses of NTSD as well as TSP (0.6 g/kg/day, TSP control) and normal saline (10 ml/kg/day, as the no treatment control) by gavage for 6 weeks. The 10 control rats were treated the same way as the HF rats except that they received only olive oil and saline (Fig. 1).

After 6 weeks of treatment, rats in each group were fasted for 24 h and then anesthetized with a 3% sodium pentobarbital (1 ml/kg) by a single intraperitoneal injection. The same part of the right liver lobe was dissected from each rat for fixation in 100 ml/l formaldehyde solution and hematoxylin and eosin (H&E) staining. The liver cell degeneration, collagen fiber hyperplasia and tissue morphology changes were investigated with the aid of a light microscope (BX51; Olympus, Tokyo, Japan). Quantitative and semi-quantitative methods were adopted to score the HFs. Zero points indicated no pathological changes in liver cells, a normal liver or no obvious collagen fiber hyperplasia; 1 point indicated that the proportion of degenerated liver cells were <25% of the total cells, collagen fiber showed slight hyperplasia, central vein and portal area exhibited a small amount of fiber elongation but no septum formation; 2 points indicated that the proportion of degenerated liver cells was between 25 and 50% of all liver cells with obvious collagen fiber hyperplasia; the connective tissue around the central vein and portal area were thickened, which extended fiber chords to form incomplete fiber septa; 3 points indicated that the proportion of degenerated liver cells was between 50 and 75% of all liver cells with massive collagen fiber hyperplasia as well as false lobules, which were formed by individual complete or incomplete thick fiber septa; 4 points indicated that the proportion of degenerated liver cells was >75% with thicker complete fiber septa and a large formation of false lobules.

After 6 weeks of treatment, rats in each group that had been fasted for 24 h were given 3% sodium pentobarbital anesthesia (1 ml/kg) and blood samples collected from the femoral artery and subsequently the serum was isolated. The Au2700 full automatic biochemical instrument (Olympus) was used for the detection of serum ALT, AST, albumin and globulin content.

Thirty healthy SPF SD rats were randomly divided into 6 groups (5 rats in each group), which were administered by gavage low (7.1 g/kg/day), medium (14.2 g/kg/day) or a high (28.4 g/kg/day) NTSD dose as well as TSP (0.6 g/kg/day) and normal saline as the control. All types of gastric perfusion medicine and liquids were delivered at a rate of 10 ml/kg/day. Gastric perfusion was performed twice a day for 3 consecutive days, with the rats having been fasted for 12 h before each perfusion. Two hours after the last gastric perfusion, Lumianning II (Jilin Huamu Animal Health Product, Co., Ltd., Changchun, Jilin, China) was injected intramuscularly for anesthesia and blood samples collected from the inferior vena cava. These samples were stored at 4°C overnight, then centrifuged for 15 min at 1000 × g to isolate the serum. The sera from the same groups were mixed, filtered through a 0.22-μm filter then inactivated at 56°C for 30 min, in order to remove possible biologically active substances, and samples stored at −70°C until required for subsequent analysis.

HSC-T6 cells were seeded onto 96-well plates (the volume of each well was 150 μl and contained 3×10³ cells), and the culture medium (containing 20% of different doses of NTSD containing serum) was changed after 24 h, with equivalent volume TSP-containing serum treated cells as the positive control, an equivalent volume of normal rat serum-treated cells as the normal control, and saline-treated HF cells as the negative control. The 96-well plates 24, 48, 72 and 96 h, respectively, after incubation started were removed and the cell growth status observed using an inverted microscope (IX71; Olympus) and digital images were taken for a permanent record. Subsequently, we replaced each well with fresh medium, added 10 μl CCK-8 kit solution and removed the plates after 2.5-h incubation, slightly oscillated them for 25 sec to mix well, and then detected the light density at 450 nm using a plate reader (Victor X2 Multilabel plate reader; Perkin-Elmer, Waltham, MA, USA). The cell proliferation curve was plotted with the light density value as the y-axis and the processing time as the x-axis. Each experiment was repeated 3 times.

HSC-T6 cells with 1×10⁶/ml density were routinely cultured for 24 h, and replaced by fresh medium (containing 20% of different doses of normal serum, NTSD or TSP medicated serum). After 72 h, cells from each group were collected and the cell number counted to ensure each group contained cell numbers from 1×10⁴–2×10⁶. Twenty-four to 72 h after 80% ethanol fixation, cells were stained with propidium iodide (PI, 4 mg/ml; Sigma-Aldrich), and flow cytometry (BD FACSCalibur; BD Biosciences, Franklin Lakes, NJ, USA) used to detect and analyze cell cycle changes. Each experiment was repeated 3 times.

For the detection of cell apoptosis, we took 100 μl cell suspensions (5–10 million cells), added 195 μl binding buffer to re-suspend the cells and mixed them thoroughly with 5 μl Annexin V-FITC (Beyotime Institute of Biotechnology, Beijing, China). Then the samples were incubated for 10 min in the dark at room temperature (20–25°C), centrifuged for 5 min at 1000 × g, re-suspended in 195 μl binding buffer, mixed with 10 μl PI, and cell apoptosis was detected by flow cytometry. Each experiment was repeated 3 times.

The 4 μm paraffin or 6 μm frozen sections from liver tissue of rats in all groups were immunostained with anti-TGF-β1 antibody (1:400; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA), anti-Smad3 antibody (1:400; Santa Cruz Biotechnology) and anti-Smad7 antibody (1:400; Santa Cruz Biotechnology) and detection kit (ZSGB-BIO, Beijing, China) reagents. Pepsin (ZSGB-BIO) digestion was used for antigen retrieval and a DAB color reagent kit (ZSGB-BIO) was used to develop the color. Under the microscope, 5 non-overlapping fields in each section were selected under a high power field (×400). Image-Pro Plus software (Media Cybernetics, Bethesda, MD, USA) was used to analyze the mean optical density value.

Total protein was extracted by western and IP cell lysis solution (Beyotime Institute of Biotechnology, Shanghai, China) according to the manufacturer's instructions from liver tissues of rats in all groups, and the protein concentration measured with a BCA protein concentration kit (Beyotime Institute of Biotechnology). The primary antibodies included anti-TGF-β1 antibody (1:100; Santa Cruz Biotechnology), anti-Smad3 antibody (1:200; Santa Cruz Biotechnology), anti-Smad7 antibody (1:200; Santa Cruz Biotechnology) as well as proliferating cell nuclear antigen (PCNA) antibody (1:800; Millipore, Billerica, MA, USA).

Both liver tissue and HSC-T6 cell samples were extracted with TRIzol reagent (Invitrogen, Carlsbad, CA, USA) to obtain their total RNA, according to the manufacturer's instructions. Then, we used a PCR kit (Takara, Dalian, China) with the total RNA as a template to amplify the cDNA of TGF-β1, Smad3, Smad7 and PCNA genes, and with β-actin gene added as the internal reference. Reverse transcription reaction conditions were: 30°C for 10 min; 42°C for 30 min, 99°C for 5 min, 5°C for 5 min, 1 cycle. PCR reaction conditions: 94°C for 1 min; 94°C for 30 sec, 62.8°C for 30 sec, 72°C for 1 min, 35 cycles; 72°C extension for 5 min. After the reaction, the PCR reaction product was ran on a 1% agarose gel electrophoresis and Goldview™ (G8140; Beijing SBS Genetech, Co., Ltd., Beijing, China) nucleic acid dye staining used to confirm the product. The sequences of primers of PCNA, TGF-β1, Smad3 and Smad7 are shown in Table I.

For statistical analyses SPSS for Windows (version 13.0; SPSS, Inc., Chicago, IL, USA) was used and data shown as the mean ± SD. Student's t-test was used for comparison between groups with a normal distribution. P<0.05 was considered to be a statistically significant difference.

We first added and omitted several traditional Chinese medicines from the original TSP formula to prepare NTSD. The TSP formula contains 8 kinds of traditional Chinese medicines for resolving hard masses and recovering blood stasis, which have been maintained in the NTSD compounds, whereas 15 kinds of herbs inducing loss of vital Qi and pathogenic stagnation factors were omitted. Instead we added Radix Astragali, Poria, as well as Artemisia capillaris to the novel NTSD composition.

Hence, the complete formula of NTSD was as follows (in 142 g): ground beetle 10 g, turtle shell 15 g, bupleurum root 10 g, scutellaria root 12 g, pinellia 15 g, Artemisia capillaris 15 g, cassia twig 10 g, peach kernel 10 g, Poria 15 g, Radix Astragali 15 g and white peony root 15 g. The ingredients were treated with soaking, decocting, filtering and concentration to derive the final NTSD formula.

As shown in Table II, compared with healthy control rats, the ALT and AST serum concentrations in the non-treated HF rats were significantly increased (P<0.05) and albumin concentrations significantly decreased (P<0.05) compared to the healthy control rats, which indicated that the liver function of HF rats was affected. However, TSP treatment significantly reduced the levels of ALT and AST levels increased in HF rats together with an increased albumin content, which proved its anti-HF effect. More importantly, NTSD also showed a significant improvement in liver function regarding ALT, AST and albumin serum concentrations, which supported its equal anti-HF efficacy. There was no significant difference between the effects of high, medium and low doses of NTSD on liver function suggesting that the effect of NTSD in low dose (7.1 g/kg/day) might have been saturating already (Table II).

We next investigated the role of NTSD in the pathological morphology changes in liver tissue in HF rats. As shown in Fig. 1 and Table III, most of the normal structure of hepatic lobes in saline treated HF rats (the non-treatment group) was destroyed or disappeared, large amounts of fat tissue degenerated and underwent necrosis, which turned into an empty net with a few liver cells remaining, massive fiber tissue proliferated around the portal and hepatic necrosis area to form a wide, uneven thickness and connected fibrous septa, and the HF scores were >3 points, which proved the success of the HF model. Compared with the non-treatment HF group, NTSD and TSP treatment groups had less hepatic lobe destruction, reduced fat degeneration, decreased inflammatory cell infiltrations, mild hyperplasia of collagen fiber, only thin fiber bundles, no obvious false lobules, and significantly decreased HF scores (Fig. 2 and Table III). Therefore, the above results clearly illustrated that both NTSD and TSP could improve the pathological changes of liver tissue in HF rats, and the results were consistent with the improvement of liver function.

The activation and proliferation of HSC played a key role in the occurrence and development of HF (2,4). To investigate the protection mechanism of NTSD in HF rats, we studied the effects of NTSD on the growth of HSC-T6 cells in vitro. After 72 h of culture in NTSD serum-containing medium, HSC-T6 cells were slightly smaller and rounder, with a reduced adherent ability, and this change was dose-dependent (Fig. 3A). CCK-8 assay showed that NTSD-containing serum inhibited HSC-T6 cell proliferation in a dose-dependent manner and the inhibitory effect of high dose NTSD was more effective than TSP-containing serum (Fig. 3B).

RT-PCR and western blotting was used to test for any PCNA gene expression changes of HSC-T6 after NTSD and TSP-containing serum treatment. The expression level of PCNA is closely related to the synthesis of DNA and an indicator reflecting the state of cell proliferation (16). As shown in Fig. 3C and D, compared with control cells, both high dose NTSD and TSP-containing serum treated cells showed decreased transcription levels of mRNA and PCNA protein expression, with high dose NTSD-containing serum showing the highest effect, which is consistent with the previous cell proliferation data (refer to the references cited).

To analyze further the cell proliferation inhibitory effect of NTSD and TSP, the effects of drugs on the cell cycle of HSC-T6 was analyzed. Compared with control cells, cells treated by NTSD and TSP-containing serum for 72 h showed an obvious S-phase arrest (P<0.05). As shown in Fig. 4, the proportions of S-phase in control cells, low dose NTSD, medium dose NTSD, high dose NTSD, and TSP-containing serum were 23.26±3.25, 25.40±4.16, 32.90±5.31, 42.62±7.52 and 30.91±4.86%, respectively. Therefore, both NTSD and TSP effectively induced S-phase arrest and inhibited the proliferation of HSC-T6 cells in vitro.

We further studied if the NTSD containing serum could promote HSC-T6 cell apoptosis. As shown in Fig. 5, 72 h after drug containing serum treatment, FCM cell apoptosis detection showed that NTSD could induce HSC-T6 cell apoptosis in a dose-dependent manner, and percentages of apoptotic cells were 2.89±0.74, 5.64±1.03 and 8.98±1.82%, respectively. Similarly, TSP-containing serum also significantly promoted apoptosis of HSC-T6 cells, and the apoptosis percentage was 8.47±1.69% (Fig. 5). Hence, these data clearly showed that NTSD had the ability to induce apoptosis of HSC-T6 cells.

The TGF-β1/Smad signaling pathway is one of the most important signaling pathways that drives the development of HF (5,17). Therefore, the role of NTSD treatment on the TGF-β1/Smad signaling pathway of HF rats was investigated. Liver tissue immunohistochemical results showed that compared with normal rats, the expression of TGF-β1 and Smad3 in the liver of HF rats was significantly increased, while the expression of Smad7 was decreased, which supported a role of the TGF-β1/Smad signaling pathway in promoting the development of HF (Fig. 6). More importantly, low, medium and high dose NTSD and TSP significantly decreased the TGF-β1 and Smad3 expression levels and increased Smad7 (which is an inhibitory Smad protein against Smad3) expression, suggesting that NTSD and TSP could inhibit the TGF-β1/Smad signaling pathway in multiple ways. In order to confirm further the changes in the expression of TGF-β1, Smad3 and Smad7 in the treatment of NTSD and TSP, RT-PCR and western blotting was used. As shown in Fig. 7, at mRNA and protein levels, changes were detected in the expression of TGF-β1, Smad3 and Smad7, which were consistent with the immunohistochemical results. Therefore, the discovery that NTSD and TSP could significantly inhibit TGF-β1/Smad signaling pathway, strongly explains their anti-HF effects.