Anti-apoptosis mechanism of triptolide based on network pharmacology in focal segmental glomerulosclerosis rats

Focal segmental glomerulosclerosis (FSGS) is a clinical pathological syndrome, and its typical pathological feature is sclerosing lesions in the focal glomeruli and in the glomerular segment. The clinical manifestations of FSGS patients are massive proteinuria, hematuria, hypertension, and progressive decrease in renal function. The condition of 3.6% of patients with end-stage renal disease developed from FSGS [1,2]. Currently, the main clinical therapies for FSGS are immunologic drugs, glucocorticoids, and blockers of the renin–angiotensin system; however, their therapeutic effects are not satisfactory. [3]. Triptolide (TPL) is the most active and effective diterpene lactone epoxide compound isolated from Tripterygium. [4]. TPL has anti-inflammatory, anti-tumor, and immunologic effects on many diseases [4]. TPL inhibits the secretion of many cytokines, adhesion molecules, and chemokines and affects the functions of various cells, including dendritic cells and renal tubular epithelial cells [5,6]. TPL has been reported to alleviate the progression of glomerulosclerosis and the excretion rate of urinary albumin to inhibit the progression of diabetic neuropathy [6]. However, the effects and mechanisms of TPL in FSGS are still unclear. We explored the mechanism of FSGS-mediated podocyte pathogenesis based on the FSGS rat model. The present study has great significance for the diagnosis, prevention, and treatment of FSGS.

In the past, research on Chinese herbal extracts focused on a particular aspect and on finding the biological characteristics explaining the pharmacological effect with respect to this aspect [7]; however, this approach is usually one-sided. It is important to explore the relation between the acquired proof and the research results. With the development of bioinformatics and network pharmacology, proposal of a theory and proving this theory through experiments has become the main method to explore the mechanism of Chinese herb compounds [8].

Network pharmacology is based on high-throughput omics data analysis, virtual computer computing and network database retrieval, and it combines systems biology with multidirectional pharmacology [9]. The mechanism of drug action was researched via the construction and analysis of biological networks. The systematic and holistic nature of network pharmacology is consistent with the characteristics of Chinese herbs, which exhibit multi-components, multi-targets, and systematic regulation. It has been widely used to explore the pharmacological basis of Chinese medicine and the drug mechanism and to interpret drug compatibility [10,11]. Network pharmacology has been recognized by many Chinese medicine researchers [12]. The multi-component and multi-target network research mode breaks the traditional research mode of a single ingredient and a single target, providing a new method for comprehensive analysis of the mechanism of the compounds [13]. In the early stage, a total of 47 target genes and the corresponding 16 active constituents of Tripterygium were used to construct the ingredient-target network. The present study mainly explores the mechanism of TPL in FSGS through bioinformatics and functional experiments.

Using the Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (http://lsp.nwu.edu.cn/tcmsp.php, TCMSP), each candidate’s drug ability was analyzed according to its oral bioavailability (OB) and drug-likeness (DL) indices recommended by the TCMSP. OB refers to the degree and rate of drug absorption into the circulatory system, which is an important indicator for objectively evaluating the intrinsic quality of drugs. The higher the OB of the compound, the more likely it is to be developed for clinical application. DL is the sum of the pharmacokinetic properties and safety, which arises from the interactions among the physicochemical properties and structural factors, including solubility, permeability, and stability. It can be used to optimize compounds, analyze the results of drug activity, predict in vivo pharmacokinetics, direct structure modifications, etc. As the TCMSP recommends, molecules with OB ≥ 30% and DL ≥ 0.18 were considered to exhibit relatively better pharmacological properties and were screened out as candidate compounds for further analysis.

To comprehensively understand the molecular mechanisms, disease–compound–target networks were constructed using the Cytoscape visualization software 3.7.1. All target genes related to FSGC were obtained from the GeneCards database (https://www.genecards.org/). All the candidate compounds of Tripterygium were retrieved from the TCMSP to obtain the associated targets. Next, disease, compounds, and targets were inputted into the software, and a disease–compound–target interaction network was constructed. In the process of constructing the network, the layout algorithm (attribute circle layout) was applied. We can set the geometric position of every node and visually display the network topology using color, graphics, and symbols, making reasonable arrangements for every node and creating a clear visual effect. Degree and betweenness centrality are two important parameters of the topology structure, which were used to evaluate the essentiality of each target and compound.

Search Tool for the Retrieval of Interacting Genes (STRING, https://string-db.org) is an online tool and used to construct the PPI network with confidence network edges and a medium confidence of 0.400 as the product criteria. Cytoscape 7.1.0 was used to perform the visualization of PPI network. The Molecular Complex Detection (MCODE) plug-in was used to screen the significant modules in the PPI network with a degree cut-off = 2, node score cut-off = 0.2, k-core = 2, and maximum depth = 100. The corresponding proteins in the central nodes and highly degree were potential core proteins encoded by key candidate genes that have important physiological regulatory functions.

The Database for Annotation, Visualization, and Integrated Discovery (DAVID) database was used to perform GO enrichment analysis and KEGG pathway enrichment analysis. The GO terms were classified into three categories: biological process (BP), cellular component (CC); and molecular function (MF). P<0.01 was considered to indicate a statistically significant difference.

A total of 40 Sprague Dawley (SD) rats (male, weighing 160–180 g) were provided by Yison Bio co.LTD (Shanghai, China). Animals were housed individually in polycarbonate cages with wood chip bedding and were maintained in an air-conditioned animal room (temperature: 24°C, relative humidity: 55 ± 5%) on a 12-h light/dark cycle. Each animal experiment was carried out following the local Care for Laboratory Animals guidelines formulated by the Animal Experimental Center. The Ethics Committee had approved the studies using laboratory animals at the Guangxing Affiliated Hospital of Zhejiang Chinese Medical University.

All rats were randomly divided into a sham operation group (Sham), model group (FSGS), a group administered 80 mg/(kg·d) of Tripterygium by gavage (TPL(80)+FSGS), and a group administered 160 mg/(kg · d) of Tripterygium by gavage (TPL(160)+FSGS) (n=10 rats/group). One day before the operation, the TPL (80 or 160)+FSGS groups were administered TPL (Purifa Technology Development Co. Ltd., Chengdu, Sichuan, China) 80 or 160 mg/(kg d) by gavage; the Sham and FSGS groups were given isovolumic normal saline till the end of the experiment. The animals were intraperitoneally anesthetized with pentobarbital sodium (60 mg/kg body weight) and then placed on a homeothermic pad to maintain a core body temperature of 37°C to establish the FSGS model. The rats were first subjected to unilateral nephrectomy (left side) on day 1 and then injected in the caudal vein with adriamycin 5 mg/kg (on day 7) and adriamycin 3 mg/kg (on day 28) dissolved in 0.9% saline at a dilution of 2 mg/ml. Meanwhile, the kidneys of the control rats were exposed without dissecting the kidney tissue, followed by layer-by-layer suturing. These rats were then injected with saline on days 7 and 28 through the tail vein after the sham operation. Eight weeks post-surgery, blood samples were obtained from the tail veins, and the animals were killed. Following adequate anesthesia with pentobarbital sodium (180 mg/kg body weight), the organs were removed, frozen, or fixed in 4% paraformaldehyde. The serum and whole kidneys were harvested for biochemical, histological, and molecular analyses. The urinary protein levels of the rats were quantified before the end of the experiment. Animals with >100 mg/24 h urinary protein indicated successful establishment of the model, and they were included in subsequent experiments.

An apoptosis detection kit (Promega, Madison, WI) was used to detect apoptosis according to a previously described method [14]. In brief, renal sections were subjected to TUNEL staining in accordance with the manufacturer’s instructions. Later, IF microscopy was used to analyze the samples using a Zeiss Axiovert 200 M fluorescent microscope equipped with an AxioCamMR3 camera. Six fields (magnification 400×) were randomly selected from every section from 10 different rats, and cells with positive TUNEL staining were analyzed.

Rat glomerular podocytes were provided by Yubo Bio-Technique Co. Ltd (Shanghai, China), which were then cultivated according to a previously described method. Rat podocytes were cultivated in RPMI 1640 (Sigma-Aldrich, U.S.A.) containing streptomycin (100 μg/ml), penicillin (100 U/ml) (Solarbio, Beijing, China), and 10% fetal bovine serum (FBS, Gibco, NY, Grand Island). Subsequently, the cells were cultivated in a 5% CO₂ incubator (Heraeus, Japan) at 33°C with interferon-γ (IFN-γ, 40 units/ml, Sigma, St Louis, MO, U.S.A.). Later, to induce differentiation, the podocytes were maintained at 37°C for 2 weeks in the absence of interferon. Podocytes (3 × 10⁵ cells/ml) were plated into 6-well plates in the presence of complete medium. After 24 h of standing, the podocytes were subjected to 24 and 48 h of TPL treatment at different concentrations (0, 5, 10, 20, 40, and 80 µmol/ml) before they were collected for subsequent analysis.

The previously described human IL4 plasmid DNA at full length [15] was utilized to increase IL4 expression in cells via using transient transfection. pcDNA3.1-Myc/His EV plasmid (Life technologies) and On-Target Plus scramble RNA (Dharmacon) were used as transient transfection controls. Sequences for IL4 overexpression was ACAUUACUGCCUGAAGGGUGAAUUAACGC.

Cells were grown into the 96-well plates at the density of 1 × 10⁵ cells/well, followed by 24 and 48 h of culture. Afterwards, cell viability was detected via using the CCK-8 kit (Dojindo Molecular Technologies, Gaithersburg, MD, U.S.A.). Then, cells in each group were cultivated for additional 24 and 48 h, respectively. Next, the CCK8 solution (10 μl) was added into cell at 37°C for 4 h. The absorbance was determined at 450 nm for obtaining the cell growth curve by the iMark microplate absorbance reader (Bio-Rad Laboratories, Inc., Hercules, CA, U.S.A.). Each experiment was carried out in triplicate.

Cell apoptosis was examined using the Annexin V-FITC/PI kit. Briefly, the cells were subjected to 0.25% trypsin digestion (Thermo Fisher Scientific, Waltham, MA, U.S.A.), followed by two washes with cold PBS; resuspension with 5 μl of PI, 5 μl of annexin V-FITC, and 500 μl of binding buffer; and incubation under 15 min of ambient temperature in the dark. Typically, Annexin V-FITC can bind to phosphatidylserine located on the outer apoptotic cell membrane, whereas PI can penetrate and stain cells with impaired membranes before binding to and labeling DNA. Data were collected using a flow cytometer (BD FACSCalibur; BD Biosciences, Franklin Lakes, NJ, U.S.A.) and analyzed by FlowJo. Clumped cells were excluded from the FSC-H/FSC-A dot plot for selecting the single cells. Cells in the annexin V-FITC-/PI+, annexin V-FITC+/PI+, and annexin V-FITC+/PI− quadrants were regarded as apoptotic cells.

Cells were subjected to lysis within the RIPA buffer (Beyotime, Shanghai, China) to collect the lysates in tubes, followed by 20 min of centrifugation at 4°C at 12,000 g. Later, all supernatants were extracted, and protein content was measured by the BCA Protein Quantitative Kit (Beyotime, Shanghai, China). Then, Western blotting had been carried out in accordance with the instruction. Afterwards, 20 μg protein was subjected to 10% SDS-PAGE for separation, followed by transfer onto the PVDF membranes (Millipore, Billerica, MA, U.S.A.). Later, the membranes were blocked using 5% skimmed milk for 1 h, followed by overnight incubation with anti-IL-4 (dilution 1:8000, Abcam, Cambridge, MA, U.S.A., ab69811), nephrin (Abcam, ab227806; diluted at 1:1200), podocin (Abcam, ab50339; diluted at 1:1000), phosph (p)-Stat6 (BioVision, U.S.A., 3476-100; diluted at 1:1000), and GAPDH (Abcam, Cambridge, MA, U.S.A., ab181602; dilution 1:1000) rabbit anti-human antibodies, at 4°C, separately. Afterwards, cells were subjected to 1 h incubation with HRP-labeled secondary antibody (goat anti-rabbit antibody, Abcam, Cambridge, MA, U.S.A., ab116282; dilution 1:2000), prior to ECL detection. The Immobilon Western Chemiluminescent kit (WBKLS0100; Millipore, U.S.A.) was used to reveal the reactive bands using Roche Cobas e601 automated chemiluminescence image analysis system (Roche, U.S.A.).

The GAPDH and IL-4 mRNA expression was detected through RT-qPCR. Total cellular RNA was isolated by TRIzol (Invitrogen, Carlsbad, CA, U.S.A.) in accordance with manufacturer protocols. Afterwards, cDNA was synthesized by reverse transcription of RNA according to the reactions below: RNase-free dH2O, total RNA (500 ng), and 5×PrimeScript RT Master Mix (2 μl) were added until the final volume became 10 μl. The Prism 7500 (ABI, Foster City, CA, U.S.A.) was employed for real-time PCR following the standard protocol of SYBR green assay. Primers used in the present study were shown below: IL-4-F: 5′-GATCACAAAGTACTGGTCCTGG-3′. Notably, GAPDH served as a normal control, with the primers of 5′-CACCCTGTTGCTGTAGCCAAA-3′ (reverse) and 5′-TGACTTCAACAGCGACACCCA-3′ (forward). Later, qPCR was carried out in triplicate using 7500 Real-Time PCR ABI system (ABI, U.S.A.) at a format of the 96-well plate. The reaction volume of 20 μl was prepared for PCR, which included forward primer (0.8 μl, 10 μM), RNase-free dH2O (7.4 μl),reverse primer (0.8 μl, 10 μM), 2×FastStart Universal SYBR Green Master (10 μl, ROX; Invitrogen, Guangzhou, China), and cDNA (1 μl). Besides, the PCR conditions were as follows, 10 min at 95°C, followed by 15 s at 95°C for 40 cycles, and 1 min at 60°C. The sequence detection software (1.6.3, Applied Biosystems, ABI, U.S.A.) was used for data analysis. Relative GAPDH or IL-4 mRNA level was measured and standardized according to 2^−ΔΔCt method based on GAPDH.

SPSS 15.0 (http://spss.en.softonic.com/) was employed for all statistical analyses. Differences between two groups were analyzed through independent sample t-test, whereas those among several groups were examined by one-way analysis of variance (ANOVA). Rate was compared by chi-square test. The statistically significant level was set as P<0.01 or P<0.05.

Using TCMSP databases (http://lsp.nwsuaf.edu.cn/tcmsp.php), 144 compounds of Tripterygium were retrieved. According to the criteria of DL ≥ 0.18 and OB ≥ 30%, a total of 51 chemical ingredients were selected (Table 1). TPL was verified as an active ingredient of T. wilfordii.

The TCMSP and GeneCards databases were used to predict the potential targets for each compound in FSGS. As a result, 123 target genes from the GeneCards database were verified to be involved in FSGS (Table 2), and 695 target genes of Tripterygium from the TCMSP database were verified (Table 3). After importing data into Cytoscape, a disease–compound–target network was constructed (Figure 1A). In addition, 47 overlapping target genes from two databases (TCMSP and GeneCards) were used to construct the PPI network. IL4 obtained from the most significant module of the PPI network was verified as a key by using the MCODE plug-in of Cytoscape software, and it was found to be involved in FSGS (Figure 1B–D).