DanLou Tablet (Batch Number: 20140101038, 0.3g/tablet), a commercial preparation of DLP, was obtained from Jilin Kangnaier Pharmaceutical Group Co. Ltd. (Jilin, China). The chromatograms with characterization of the dominating compound (s) of DanLou Tablet was determined as same as those shown in our previous studies (Dong J. et al., 2013a; Wu et al., 2014). DanHong injection (Batch number: 12081024077, 10 ml/ampulla), a commercial preparation of DHP, comprising 750 g Salvia miltiorrhiza, 250 g Safflower, and 7 g Sodium chloride was supplied by Heze Buchang Pharmaceutical Co., Ltd (Shandong, China). The chromatograms with characterization of the dominating compound (s) of DanHong injection was determined as same as those shown in our previous studies (Liu et al., 2013). Bleomycin (BLM) hydrochloride was purchased from Nippon Kayaku Co. (Tokyo, Japan). Rosiglitazone (ROS) (Batch number: 122320-73-4), which has been well established by invasive techniques to reduce lung fibrosis (Burgess et al., 2005; Lee et al., 2008; Jin et al., 2012), was purchased from Alexis Biochemicals (San Diego, CA, United States). Mouse anti-α-SMA antibody was purchased from Boster Biotec (Wuhan, China). Rabbit anti-TβRII antibody was purchased from Santa Cruz Biotechnology (Santa Cruz, CA, United States). Chloral hydrate (Batch number: Q/12HB 4218-2009) was purchased from Tianjin Kermel Chemical Reagent Co. Ltd. (Tianjin, China), freshly prepared to 3.5% solution with saline before experiment. HE Staining Kit was purchased from Wuhan Boster Biological Co., Ltd. (Wuhan, China). Formalin was purchased from Shanghai Weiao Biological Co., Ltd. (Shanghai, China). Sodium chloride injection was purchased from Cisen Pharmaceutical Co., Ltd. (Shandong, China).

Seventy male C57BL/6 mice, weighing 20–25 g, were purchased from HFK Bioscience Co, Ltd. (Beijing, China) and housed in a 12 h light/dark cycled facility with free access to food and water. All experiments were reviewed and approved by the Institutional Animal Care and Use Committee at the Tianjin International Joint Academy of Biotechnology and Medicine (TJAB-JY-2011-002) and were carried out under the Guidelines for Animal Experiments of Tianjin University of TCM.

IPF was established in mice model by intratracheal instillation of BLM hydrochloride (5 mg/kg dissolved in 0.9% saline). After the bleomycin treatment, mice were frequently examined for 2 weeks by μCT. Mice presenting fibrosis lesions involving more than 25% of the lung (Model group) were divided into three experiment groups, treating once daily with DLP, DHP and Rosiglitazone (ROS, the positive control drug), respectively, for another 2 weeks. DLP and ROS were administered via oral gavage and DHP was administered via intraperitoneal injection. According to the normal clinical analysis (Guo et al., 2014; Yao et al., 2017), the mouse equivalent dose of DLP and DHP were calculated on the basis of body surface area by multiplying the human dose (43 mg/kg and 0.20 ml/kg body weight) by the correction factor ratio (12.3) (Nair and Jacob, 2016). This calculation resulted in a mouse equivalent dose for DLP and DHP of 530 mg/kg and 2.5 ml/kg body weight, which set to be the medium drug concentration in our study. Thus, we set up three gradient concentrations for DLP (177, 530, 1590 mg/kg body weight) and DHP (0.8, 2.5, 7.5 ml/kg body weight), and ROS (5 mg/kg bodyweight) to treat the BLM-induced pulmonary fibrosis mice. Micro-CT scans were performed at baseline, 2 and 4 weeks to evaluate the effect of these treatments. Mice without treatments (Normal group), treated with BLM alone but no fibrosis or inflammation findings (Sham group) were given an equal volume of saline in the same schedule.

Mice were sacrificed at 4 week and the lower lobes of left lung tissues were taken and fixed by 10% formalin. The mice lung imaging was performed using a μCT scanner (IvIS Lumina K series III, PerkinElmer) for 5 min with help of computer software (Living Image version 4.3.1, PerkinElmer). The X-ray system uses a microfocus tube with a spot size of 5 μm and produces X-rays in a cone-beam geometric formation. The images were acquired at 90 kV, 160 mA and the wide field of view (FOV) scanning at 40 mm. The image field width was up to 68 mm and the voxel size was 35 × 35 × 35 μm. The pathological grade of inflammation and fibrosis in the whole lung was evaluated under 25–100 magnifications and determined according to the criteria. Moreover, the Wilcoxon signed-rank test using statistical methods was conducted to compare the μCT images of different groups after 2 and 4 weeks (Natarajan et al., 2012). Abnormal areas on μCT imaging were evaluated before and after ROS administration. Observes gave a point (from 0 to 5 scores) at each level for consolidation, GGA, reticular opacity and honeycombing according to method proposed by Lee et al. (2008).

The lower lobes of left lung tissues were fixed in 10% formalin, following embedded in paraffin wax. After that 5 μm sections of lung edge were cut and selected for hematoxylin and eosin (H&E) stain, hematoxylin colors nuclei of cells (and a few other objects, such as keratohyalin granules and calcified material) blue. The nuclear staining was followed by counterstaining with an aqueous solution of eosin, which colors eosinophilic structures in various shades of pink.

In the process of Masson’s trichrome stain, it produces green collagen, pink cytoplasm and dark brown of cell nuclei. The fibrosis results of Masson’s trichrome stain were graded according to Modified Ashcroft Scale (Hubner et al., 2008), which were quantified from grade 0 to 8 designed for a standardized fibrosis evaluation in small animals, with higher score corresponding to a more serious level of fibrosis. As an indirect parameter of fibrosis, the alveolar air area was quantified by Image-Pro Plus 6.0 software (Media Cybernetics; Rockville, MD, United States). The Air Area is expressed as a percentage of area (μm²) occupied by air referred to the lung total area (μm²) within the region of interest (ROI). Bronchi and blood vessels have been removed from the ROI area.

Five micrometer lung sections were incubated with the primary antibodies in 10% bovine serum in PBS overnight at 4°C. After washing with PBS buffer for 3 times, it was followed by the incubation with secondary and third antibodies for 1 h each at 37°C, respectively. Primary antibody dilutions used were as follows: α-SMA 1:500 and TβRII 1:200. The stained slides were mounted and visualized in bright-field with Olympus CX41 microscope. At least nine random microscopic fields were selected and the integral optical density (IOD) was calculated using Image-Pro Plus 6.0 software to determine the intensity of IHC staining.

Networks analysis of the targeted IPF related genes regulated by DLP and DHP were performed as described in our previous study (Lyu et al., 2017, 2018). The main source of disease targets for IPF was obtained from IPA¹ database. According to our previous study (Dong J. et al., 2013), 15 compounds used as marker substances of DLP were uploaded into the IPA system to enable the discovery visualization, including gallic acid, puerarin, daidzin, peoniflorin, naringin, rosmarinic acid, salvianolic acid A, formononetin, calycosin, ethyl gallate, lipoteichoic acid, chlorogenic acid, ferulic acid, cryptotanshinone and tanshinone IIA. Similarly, 14 compound were chosen as major ingredients of DHP for IPA analysis, such as L-proline, L-phenylalanine, caffeic acid, danshensu, rutin, hydroxysafflor yellow A, safflor yellow A, salvianolic acid B, uridine, syringin, chlorogenic acid, ferulic acid, rosmarinic acid, salvianolic acid A (Liu et al., 2013). “Build-Path Explorer” module was applied to discover IPF-related targets, and the relationship between the targets and DLP or DHP ingredients, respectively.

All analyses were performed using SPSS 9.0 software (SPSS Inc., Chicago, IL, United States) or GraphPad (GraphPad Prism 6, San Diego, CA, United States). The results are expressed as the mean ± standard deviation. Multigroup comparisons of the means were carried out by one-way analysis of variance (ANOVA) test with post hoc Tukey’s test. The statistical significance for all tests was set at P-values of less than 0.05.

To evaluate the therapeutic effects of DLP and DHP in the bleomycin-induced lung fibrosis, we preformed serial μCT scans on mice before and after DLP, DHP, or ROS administration. Micro-CT imaging has been used as a diagnostic standard to classify the severity of lung fibrosis in patients in correlation to other pathologic findings (Lee et al., 2009; Jin et al., 2012).

Normal and sham groups (each group n = 6) did not reveal any lung abnormality from μCT imaging (Figure 1Aa,b). After the bleomycin treatment for 2 weeks, we chose 36 mice performing inflammation and fibrosis to test the effects of drugs. Different concentrations of DLP, DHP and ROS were administered to these 36 mice for 2 weeks at 24 h intervals, the μCT were analyzed before and after the treatments. No animals died during the entire treatment period. In contrast, in the model group without any drug treatments (n = 6), 4 mice died after 4 weeks and 2 mice were found with an increased attenuation with obscuration of the pulmonary vessels, which indicated the lung fibrosis (Figure 1Ac). The BLM-treated mice show great tolerance with higher concentration of DLP or DHP after 4 weeks treatments, even up three times of clinical dosage, 1590 mg/kg and 7.5 ml/kg separately. Drug treatment had no marked influence on the body weight of mice. Figure 1Ad indicated DLP with low concentration affect the inflammatory area, while high concentrations of DLP (≥530 mg/kg) sharply reduced the degree of BLM-induced fibrosis and reversed pathological damage to the high extent (Figure 1Ae,f). In contrast, mice treated with different concentrations of DHP (Figure 1Ag–i) show marginally effect on the pulmonary fibrosis. Positive control experiment shows that level of lung fibrosis treated with ROS were greatly reduced, as shown in Figure 1Aj.

HE stain further demonstrated histopathological changes in the form of severely damaged architecture and thickened interstitium in the lung tissues from BLM-treated mice, compared with those from control and sham mice (Figure 1Ba–c). These pathological changes were significantly alleviated by subsequent treatment with DLP (Figure 1Bd–f) or ROS (Figure 1Bj). In contrast, the DHP only marginally alleviated BLM-induced lung injury (Figure 1Bg–i). Thus, the tissue pathology findings correlated well with those of lung μCT imaging.

Next, we quantified the degree of pulmonary fibrosis before (at 2 weeks) and after (at 4 weeks) drug-administration using a scale adapted from clinical practice (Lee et al., 2008; Figure 2). After DLP and ROS treatments, the average degree of fibrosis were reduced, where bronchial wall became thinner and reticular opacity was greatly reduced. Specifically, statistical analysis of the μCT lung fibrosis data (Table 1) indicated that the mean total score before and after ROS administration were 4.3 ± 0.5 and 2.0 ± 0, respectively (p = 0.0001). DLP treatment yielded a marked decrease from 3.7 ± 0.8 to 1.8 ± 0.4 (DLP low dose), from 3.8 ± 0.8 to 1.7 ± 0.5 (DLP medium dose) and from3.7 ± 0.8 to 1.7 ± 0.5 (DLP high dose), respectively, with p-value less than 0.003. In contrast, DHP treatment did not lead to any reduction in μCT scores (p > 0.05). Therefore, DLP but not DHP is effective in reducing fibrosis in the lung of BLM-induced mice.

Next, we investigated whether these drugs could also mediate mitigation of pulmonary fibrosis by using Masson’s trichrome staining. Lung sections were selected and stained by Masson. BLM-treated mice have been revealed severe pulmonary fibrosis comparing with normal slides. Followed series of concentrations of DLP, DHP and ROS were added to the mice model by intragastric administration, we observed the change of collagen deposition areas.

Normal, sham, model-operated mice section slides were shown in Figure 3Aa–c. Figure 3Ad–f show that all the tested dosages of DLP had markedly reduced the area of collagen deposition in pulmonary fibrosis. While DHP didn’t cause any decrease of the pulmonary lesion in BLM-treated mice (Figure 3Ag–i). In addition, the Ashcroft score was then used to determine the area of lung fibrosis. The data indicated that, following low- and medium-dose DLP administrations, the scores of mice with pulmonary fibrosis were significantly decreased, when compared with that of the model group, while DHP administrations were not statistically significant, when compared with that of the model group (Figure 3B). Fully in agreement with these results, the percentage of alveolar air area was significantly increased following low-, medium- and high-DLP administrations in bleomycin-treated mice compared with that of the model group, while DHP administrations were not statistically significant, when compared with that of the model group (Figure 3C).

These findings indicated an anti-fibrotic effect of DLP on pulmonary fibrosis in mice. One of the possible explanations could be that DLP has better expectorant effects than DHP, due to pulmonary fibrosis leading to sputum in the lung, so that DLP has stronger therapeutic effects on pulmonary fibrosis than DHP.

Myofibroblasts are the primary cell type involved in ECM synthesis and tissue remodeling, which leads to the loss of alveolar function (Scotton and Chambers, 2007). The hallmark of its differentiation is the expression of α-SMA. As shown in Figure 4Ac, mice treated with BLM exhibited severe lung fibrosis, where α-SMA expression sharply increased, as indicated by the intensity of its labeled antibody (brown). In comparison, mice in sham or normal groups showed few α-SMA expression and without any inflammation (Figure 4Aa,b). These immunohistochemistry results indicated that pulmonary α-SMA protein expression increased following BLM instillation, but significantly decreased by the following low, medium and high dose-DLP treatment (Figure 4Ad–f). In addition, Figure 4B showed the integral optical density of α-SMA positive expression in lung tissue samples, indicating only medium dose-DHP could significantly down-regulated the α-SMA protein expression, neither low-dose nor high-dose DHP could reduce the expression level of α-SMA in BLM-induced lung tissues (Figures 4Ag–j).

Since the epithelium-specific deletion of TGF-β receptor type II (TβRII) plays a critical role in protecting mice from bleomycin-induced pulmonary fibrosis (Li et al., 2011), we examined the effects of DLP and DHP on the expression levels of TβRII by immunohistochemistry staining and image analysis. As depicted in Figure 5Ac, BLM instillation resulted in significant increase of TβRII expression in lung tissues of mice. After 2 weeks of drug treatments, mice treated with low, medium and high dose-DLP all showed the expression levels of TβRII receptors reduced significantly, as Figure 5A,B P-value indicated. In contrast, only low dose-DHP administration could significantly down-regulated the TβRII expression level, while the TβRII expression levels after medium and high dose-DHP treatments showed slightly reduced while the differences were shown no statistically significant.

To explain the underlying mechanism that contributed to the differential efficacy of DLP and DHP for pulmonary fibrosis, we identified the unique and shared pulmonary fibrosis related targets of DLP and DHP by IPA. According to the qualitative and quantitative analysis of the major ingredients in DLP and DHP in our previous studies (Dong J. et al., 2013; Liu et al., 2013), 15 potential active ingredients of DLP and 14 potential active ingredients of DHP were used to establish the ingredient-target-disease network. These 25 ingredients cooperatively modulate 52 common pulmonary fibrosis related intracellular targets, among which 27 were uniquely regulated by DLP, five were uniquely regulated by DHP and 20 were shared by both DLP and DHP (Figure 6).

Consistent with our previous results of immunohistochemical staining, ACTA2 (encoding α-SMA) was shown to be regulated by both DLP and DHP. DHP mainly regulated pulmonary fibrosis-related inflammatory genes and signaling pathways, such as IL1β, IL4, IL5, IL6, TGFβ1, and TNFα. In addition to the above inflammatory related targets, IL10, IL17A, IL23A, ARG1, and AKT1 were also regulated by DLP ingredients. It’s worth noting that DLP could regulate not only inflammation related genes and signaling pathways, but also myofibroblast differentiation and collagen secretion related genes, such as PTPN11, RAC1, PPARG. It suggested that DLP inhibited pulmonary fibrosis through suppressing both pro-inflammatory and pro-fibrotic pathways.