Identifying the mechanisms and molecular targets of Hongjingtian injection on treatment of TGFβ1-induced HK-2 cells: coupling network pharmacology with experimental verification

Introduction

Tubulointerstitial fibrosis (TIF) is an irreversible organic change of kidney caused by many pathogenic factors (1). TIF characterized by tubular atrophy, the activation of interstitial fibroblasts and extracellular matrix (ECM) accumulation, is the basic pathological process that chronic kidney diseases (CKD) eventually move towards chronic renal failure (2).

However, due to the complex mechanism of TIF, there are no specific preventive and therapeutic measures in clinical practice. So, it is still a hot research topic to seek effective prevention and treatment measures. Traditional Chinese medicine (TCM) therapy is a unique treatment in China, which has significant clinical efficacy and less side effects. In recent years, many studies have confirmed that TCM has become an effective method to treat TIF through a variety of experimental methods (3,4). TCM injection is a therapeutic drug combining TCM theory and modern scientific means. It is made into a finished product by extracting the pharmaceutical compounds of herb-medicine, which has the characteristics of quick onset and rapid action. The Hongjingtian injection (HJT) is a water-soluble extract of Rhodiola. In China, HJT has been used to treat vascular diseases. Clinical studies have shown that HJT can improve hemorheology, reduce intima-media thickness, and alleviate atherosclerotic inflammatory reaction (5,6). Our previous study has demonstrated that salidroside, the main component of HJT, have renoprotective effects to reduce proteinuria, inhibit renal fibrosis (7).

However, elucidating the pharmacological effects of HJT on TIF is challenging, due to the complex active compounds and unclear mechanism of action. Recently, with the growing understanding of complex diseases, the focus of drug mechanism has shifted from the well-accepted “one target, one compound” model designed toward a single target to a new “multi-target, multi-compound” model aimed at systemically modulating multiple targets. Fully understanding the complex network relationships among diseases, targets, and compounds still remains a big challenge. Fortunately, with the rapid development of systems biology, multidirectional pharmacology, computational biology, network pharmacology has become a new paradigm for clarifying the multivariable interactions of compound-gene-disease system. It provides new scientific and technological support for the mechanism of TCM at the system level, and provides new ideas and directions for experimental research of TCM.

Therefore, in this study, we combined the network pharmacology with the Gene Expression Omnibus (GEO) database to further explore and explain the possible mechanism of HJT in treating TIF. Moreover, the potential mechanism and key target were verified by cell experiments. The flowchart is presented in Figure 1. We present the following article in accordance with the MDAR reporting checklist (available at https://atm.amegroups.com/article/view/10.21037/atm-22-5035/rc).

Figure 1 The technical strategy of the current study. HJT, Hongjingtian injection; TIF, treating tubulointerstitial fibrosis; DEG, differentially expressed gene; KEGG, Kyoto Encyclopedia of Genes and Genomes; qRT-PCR, quantitative real-time polymerase chain reaction; WB, Western blot.

Methods

Screening and identification of active compounds and potential targets of HJT

A total of 49 compounds of HJT were obtained by searching the literature (8). To identify the active compounds of HJT, the ingredients conforming to the requirements of lipinski, based on the published literature and the information from the SwissADME database (http://www.swissadme.ch/). Subsequently, putative targets of these potential compounds were identified from the ChemMapper (http://lilab.ecust.edu.cn/chemmapper/), Swiss Target Prediction (http://www.swisstargetprediction.ch/) and SEA (http://sea.bkslab.org) database. Then, the UniProt database (https://www.uniprot.org/) was used to convert targets into official gene symbol. Eventually, all putative targets of HJT were retrieved after removing duplicated targets. In addition, Cytoscape 3.7.2 software was used to establish and visualize the compound-target network of HJT based on the obtained results.

Search, identification, and analysis of differentially expressed gene (DEG)

Expression profiling data from GSE20247 were down-loaded from the GEO database (http://www.ncbi.nlm.nih.gov/geo/) based on the microarray platform GPL570 (Affymetrix Human Gene Expression Array), which contained 3 samples from healthy individuals and 3 TGF-β1-induced TIF samples. Based on the annotation information in the platform, probe IDs were used to identify the corresponding genes. Genes with an adjusted P<0.01 and log2(fold change) >1 or log2(fold change) <−1 were considered significantly differentially expressed and TIF-related targets. Then, these DEGs were visualized using a volcano plot. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013).

Enrichment of the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway

Based on the above analysis, the Venn diagram that intersects the potential targets of HJT and the DEGs of TIF was drawn by Jvenn (http://jvenn.toulouse.inra.fr/app/index.html). Then, with the help of Omicshare information platform (https://www.omicshare.com/), we employed the KEGG pathway analysis the gene related signaling pathway (9).

Construction of protein-protein interaction (PPI) network and key targets analysis

The PPI network of these intersection targets was obtained by using the String online tool (https://string-db.org/) (10) and visualized using the Cytoscape V3.7.2 (https://www.cytoscape.org/). Then, core targets were selected according to the median values of Degree Centrality (DC), Closeness Centrality (CC), and Betweenness Centrality (BC), which are the three most important parameters for screening the core composite targets (11). Subsequently, the co-expression relationship of genes in the PPI network were analyzed by using the MCODE module in Cytoscape V3.7.2. Ultimately, the key genes enriched in the selected pathway were selected and “key target-signal pathway map” were made using the Cytoscape V3.7.2.

Experimental validation

Reagents and antibodies

TGF‑β1 was obtained from novoprotein (CA59, Suzhou, China). HJT was obtained from Tonghua Yusheng Pharmaceutical Co., Ltd. (YBZ11852006, Tonghua, China). Anti-cytokeratin18 was obtained from Boster (A01357-1, Wuhan, China). Anti-Vimentin was obtained from proteintech (10366-1-AP, Wuhan, China). Anti-SPHK1 was obtained from proteintech (10670-1-AP, Wuhan, China).

Cell culture and treatments

The cell lines HK-2 were obtained from Procell (PC-H2022062701, Wuhan, China). The cells were cultured in modified Eagle’s medium (SH30024, MEM, HyClone) with 12% fetal bovine serum (Gibco, Thermo Fisher Scientific, USA) and 1% penicillin-streptomycin (J160016, Hyclone, USA). All cells were maintained in 5% CO₂ at 37 ℃ under a humidified atmosphere. For inducing EMT, cells were treated with different concentrations of TGF‑β1 (0, 2, 5 and 10 ng/mL) at 37 ℃ for 48 h.

Cell Counting Kit‑8 (CCK8) assay

A total of 3×10³ HK-2 cells were plated in 96‑well plates and, after a 24 h incubation period, cells were treated with different concentrations of HJT (0, 1.25, 2.5, 5, 10, 20, 40, 80, 200, 1,000 mL/L) at 37 ℃ for 48 h. Subsequently, 10 µL CCK8 solution (A311-02-AA, Vazyme, China) was added to each well and incubated at 37 ℃ for a further 2 h, and cell viability was determined by measuring the absorbance at 450 nm on a spectrophotometer.

Quantitative real-time polymerase chain reaction (qRT-PCR) analysis

Total RNA was extracted using a SPARK easy Cell RNA Kit (AC0205, Sparkjade, China,) and cDNA was synthesized using a SPARK script II RT Plus Kit (AG0304, Sparkjade, China) according to the manufacturer’s protocol. Subsequently, the quality of the extracted RNA was detected. The primer sequences were as follows: SPHK1 (forward) 5'-CTGTCACCCATGAACCTGCTGTC-3' and (reverse) 5'-ACGCCGATACTTCTCACTCTCTAGG-3', P4HA2 (forward) 5'-CCAGGAACCAAGTACCAGGCAATG-3' and (reverse) 5'-CTGCTCCATCCACAACACCGTATG-3', GAD1 (forward) 5'-TCCTCCTGGAAGTGGTGGACATAC-3' and (reverse) 5'-AGCAACTGGTGTGGGTGATGAAAG-3', AKR1B1 (forward) 5'-TGACACCAGAACGCATTGCTGAG-3' and (reverse) 5'-AGACCCTCCAGTTCCTGTTGTAGC-3' and PTGES (forward) 5'-TCCTGGGCTTCGTCTACTCCTTTC-3', and (reverse) 5'-TGTAGGTCACGGAGCGGATGG-3' and GAPDH (forward) 5'-CCTTCCGTGTCCCCACT-3', and (reverse) 5'-GCCTGCTTCACCACCTTC-3'. Real-time PCR was performed using a SYBR qPCR SuperMix Plus (R311-02, Vazyme, China). The thermocycling conditions were as follows: Initial denaturation at 95 ℃ for 2 min, followed by 40 amplification cycles at 95 ℃ for 15 sec and 60 ℃ for 10 sec. GAPDH was used as the endogenous control for normalization, and the expression was analyzed using the 2‑^ΔΔCT method. The data were extracted from three independent biological experiments with three technical replicates in each experiment.

Western blot (WB) analysis

The total protein in HK-2 cell was collected by RIPA lysis method and the protein concentration was detected by using BCA kit (PC0020, Solarbio, China). Then the proteins were separated using 12% SDS-PAGE (G2037, Servicebio, China) electrophoresis and transferred to PVDF membrane via the electro-blotting at 210 mA for 75 minutes. After taking out the PVDF membrane, wash it three times with TBST for 15 minutes each time. A 5% skimmed milk powder was added and sealed for 2 h. Primary antibody (1:1,000) was put in and cultivated in a refrigerator at 4 ℃ overnight. The next day, TBST was used for washing three times to remove primary antibody, HRP labeled secondary antibody (1:5,000) was added. After incubation for 2 h, TBST was used for washing three times, with 15 min each time. Finally, ECL luminescence kit (E412-02, Vazyme, China) was used for luminescence and the results of WB analysis were performed by using the software ImageJ. The data were extracted from three independent biological experiments.

Wound healing assay

HK-2 cells at 10⁶ cells/mL were cultured in 6-well plate. A 1,000 µL pipette was used to create a scratch in the cells. The non-adherent cells were removed by two washes of PBS, and cells in experimental group were intervened with HJT (2.5 mL/L), TGF-β1 (10 ng/mL) and the TGF-β1 and HJT (10 ng/mL TGF-β1+ 2.5 mL/L HJT). Subsequently, the cells were maintained at 37 ℃ for 48 h. The scratches were photographed at 0, 24 and 48 h respectively by an inverted microscope (XD202, Jiangnan, China). Finally, calculate the relative distance according to the equation (W₀–W_t)/W₀ ×100%. The data were extracted from three technical replicates.

Transwell assay

10⁵ cells/mL of HK-2 after the intervention were collected and inoculated in the upper chamber, and the cells were cultured in serum-free medium for 12 h. The non-invading cells on the upper chamber were eliminated using a cotton swab. After fixing with 4% paraformaldehyde, the cells on the lower side of the chamber were incubated with crystal violet. The results were counted in three fields randomly selected by a forward microscope (Ni-V, Nikon, Japan). The data were extracted from three technical replicates.

Immunofluorometric assay

First, the treated cells on the round coverslip were fixed with 4% paraformaldehyde for 20 min, penetrated with 0.3% Triton X-100 for 15 min, incubated in a 5% BSA blocking buffer for 2 h at room temperature and then incubated with Vimentin and Cytokeratin-18 antibody (1:200 dilution) overnight at 4 ℃. The next day, secondary antibodies labeled with fluorescence (Alexa Fluor 488, Beyotime, 1:1,000) were applied for 1h at room temperature protected from light, followed by incubation with 0.1% DAPI for 10 min. Addition, cells were washed with PBS between each step. Finally, the cells were observed using a forward fluorescence microscope (Ni-V, Nikon, Japan), and photographs were recorded.

Statistical analysis

Statistical analysis was performed using SPSS 21.0 software. All the experimental results were expressed as mean ± standard deviation (SD). The Student’s t-test was used to analyze the statistical difference between the experimental and control groups. Moreover, the difference among multiple groups was determined by the one-way analysis of variance (ANOVA) test followed Turkey’s posttest. P value <0.05 was considered significant.

Results

Screening of bioactive ingredients and putative targets from HJT

Forty-nine compounds of HJT were found in the literature, and all compounds were identified in the SwissADME database after applying the criteria of lipinski. The molecular names, structures and ADME-related parameters of these compounds are listed in Table 1. Finally, 36 bioactive compounds were screened from HJT, including 10 in Organic acids, 4 in Phenylethanoids, 5 in Phenylpropanoids, 1 in Flavonoid glycosides, 5 in Monoterpene glycosides, 7 in Octyl glycosides and 4 in other compounds.

Table 1

Information for candidate bioactive compounds of HJT

Category	Code name	Structure	Molecule name	Molecular formula	Lipinski	Molecular weight	Num. rotatable bonds	Num. H-bond acceptors	Num. H-bond donors
Organic acids	A-C1		4-(β-D-Glucopyranosyloxy)-3,5-dihydro xybenzoic acid or its isomer	C13H16O10	Yes; 1 violation: NHorOH >5	332.26 g/moL	4	10	7
	A-C2		4-(β-D-Glucopyranosyloxy)-3,5-dihydro xybenzoic acid or its isomer	C13H16O10	Yes; 1 violation: NHorOH >5	332.26 g/moL	4	10	7
	A-C3		Vanillic acid β-glucoside	C14H18O9	Yes; 0 violation	330.29 g/moL	5	9	5
	A-C4		Gallic acid	C7H6O5	Yes; 0 violation	170.12 g/moL	1	5	4
	A-C5		Protocatechuic acid	C7H6O4	Yes; 0 violation	154.12 g/moL	1	4	3
	A-C6		4-Hydroxybenzoic acid	C7H6O3	Yes; 0 violation	138.12 g/moL	1	3	2
	A-C7		4-Hydroxyphenylacetic acid	C8H8O3	Yes; 0 violation	152.15 g/moL	2	3	2
	A-C8		p-Hydroxybenzoic acid β-glucoside	C13H16O8	Yes; 0 violation	300.26 g/moL	4	8	5
	A-C9		Ethyl gallate	C9H10O5	Yes; 0 violation	198.17 g/moL	3	5	3
	A-C10		Creoside III	C18H24O9	Yes; 0 violation	384.38 g/moL	8	9	5
Phenylethanoids	B-C1		Salidroside	C14H20O7	Yes; 0 violation	300.30 g/moL	5	7	5
	B-C2		p-Tyrosol	C8H10O2	Yes; 0 violation	138.16 g/moL	2	2	2
	B-C3		2-(β-D-Glucopyranosyloxy)-1-(4-hydrox yphenyl) ethanone	C14H18O8	Yes; 0 violation	314.29 g/moL	5	8	5
	B-C4		2-(4-Hydroxyphenyl)-ethyl-O-β-D-gluco pyranosyl-6-O-β-D-glucopyranoside or its isomer	C20H30O12	No; 2 violations: NorO >10, NHorOH >5	462.45 g/moL	8	12	8
	B-C5		2-(4-Hydroxyphenyl)-ethyl-O-β-D-gluco pyranosyl-6-O-β-D-glucopyranoside or its isomer	C20H30O12	No; 2 violations: NorO >10, NHorOH >5	462.45 g/moL	8	12	8
	B-C6		2-(4-Hydroxyphenyl)-ethyl-O-β-D-gluco pyranosyl-6-O-β-D-glucopyranoside	C20H30O12	No; 2 violations: NorO >10, NHorOH >5	462.45 g/moL	8	12	8
	B-C7		2-(4-Hydroxyphenyl)-ethyl-O-β-D-gluco pyranosyl-6-O-β-D-glucopyranoside or its isomer	C20H30O12	No; 2 violations: NorO >10, NHorOH >5	462.45 g/moL	8	12	8
	B-C8		6’’-O-Galloyl-salidroside	C21H24O11	No; 2 violations: NorO >10, NHorOH >5	452.41 g/moL	8	11	7
	B-C9		2-Phenylethyl-6-O-α-L-arabinopyranosyl-β-D-glucopyranoside	C19H28O10	Yes; 1 violation: NHorOH >5	416.42 g/moL	6	10	7
Phenylpropanoids	C-C1		1-O-p-Coumaroyl-β-D-glucopyranose	C15H18O8	Yes; 0 violation	326.30 g/moL	5	8	5
	C-C2		Lariciresinol-4-O-β-D-glucopyranoside	C26H34O11	No; 3 violations: MW >500, NorO >10, NHorOH >5	522.54 g/moL	8	11	7
	C-C3		(7R,8S)-Dihydrodehydrodiconiferyl-alco hol-3'-O-β-D-glucopyranoside or (7R,8S)-Dihydrodehydrodiconiferyl-alco	C25H32O11	No; 3 violations: MW >500, NorO >10, NHorOH >5	508.51 g/moL	9	11	7
	C-C4		(7R,8R)-3-Methoxy-8'-carboxy-7'-en-3',8-epoxy-7,4'-oxyneolignan-4,9-diol	C19H18O7	Yes; 0 violation	358.34 g/moL	5	7	3
	C-C5		(7R,8S)-Dihydrodehydrodiconiferyl-alco hol-4-O-β-D-glucopyranoside	C25H32O11	No; 3 violations: MW >500, NorO >10, NHorOH >5	508.51 g/moL	9	11	7
	C-C6		Lariciresinol-4-O-β-D-glucopyranoside	C26H34O11	No; 3 violations: MW >500, NorO >10, NHorOH >5	522.54 g/moL	8	11	7
	C-C7		(+)-Isolarisiresinol-4-O-β-D-glucopyranoside	C26H34O11	No; 3 violations: MW >500, NorO >10, NHorOH >5	522.54 g/moL	8	11	7
	C-C8		p-Coumaric acid	C9H8O3	Yes; 0 violation	164.16 g/moL	2	3	2
	C-C9		Cinnamylalcohol-9-O-(6'-O-α-L-arabinopyranosyl)-β-D-glucopyranoside	C20H28O10	Yes; 1 violation: NHorOH >5	428.43 g/moL	7	10	6
	C-C10		6-Feruloyloxyhexanoic acid	C16H20O6	Yes; 0 violation	308.33 g/moL	10	6	2
Flavonoid glycosides	D-C1		Rhodionidin	C27H30O16	No; 3 violations: MW >500, NorO >10, NHorOH >5	610.52 g/moL	6	16	10
	D-C2		Kaempferol-7-rhamnoside	C21H20O10	Yes; 1 violation: NHorOH >5	432.38 g/moL	3	10	6
Monoterpene glycosides	E-C1		Rosiridin	C16H28O7	Yes; 0 violation	332.39 g/moL	7	7	5
	E-C2		Sachaloside VII	C21H38O11	No; 2 violations: NorO >10, NHorOH >5	466.52 g/moL	10	11	7
	E-C3		Sachaloside VIII	C21H38O11	No; 2 violations: NorO >10, NHorOH >5	466.52 g/moL	10	11	7
	E-C4		Sacranoside A	C21H34O10	Yes; 1 violation: NHorOH >5	446.49 g/moL	6	10	6
	E-C5		Sachaloside II	C21H34O10	Yes; 1 violation: NHorOH >5	446.49 g/moL	5	10	6
	E-C6		Creoside V	C21H38O10	Yes; 1 violation: NHorOH >5	450.52 g/moL	10	10	6
	E-C7		Kenposide A	C21H36O10	Yes; 1 violation: NHorOH >5	448.50 g/moL	9	10	6
Octylglycosides	F-C1		Creoside I	C14H24O7	Yes; 0 violation	304.34 g/moL	7	7	4
	F-C2		Creoside I	C14H24O7	Yes; 0 violation	304.34 g/moL	7	7	4
	F-C3		Creoside I	C14H24O7	Yes; 0 violation	304.34 g/moL	7	7	4
	F-C4		Creoside II	C14H26O7	Yes; 0 violation	306.35 g/moL	7	7	5
	F-C5		Creoside II	C14H26O7	Yes; 0 violation	306.35 g/moL	7	7	5
	F-C6		Rhodiooctanoside	C19H36O10	Yes; 1 violation: NHorOH >5	424.48 g/moL	11	10	6
	F-C7		Rhodiooctanoside	C19H36O10	Yes; 1 violation: NHorOH >5	424.48 g/moL	11	10	6
Other compounds	G-C1		4-Hydroxybenzyl-β-D-glucopyranoside	C13H18O7	Yes; 0 violation	286.28 g/moL	4	7	5
	G-C2		Picein	C14H18O7	Yes; 0 violation	298.29 g/moL	4	7	4
	G-C3		Creoside IV	C17H32O10	Yes; 1 violation: NHorOH >5	396.43 g/moL	9	10	6
	G-C4		Pyrogallol	C6H6O3	Yes; 0 violation	126.11 g/moL	0	3	3

HJT, Hongjingtian injection;

According to the target screening of the bioactive compounds in the Chemmaper, Sea and Swiss database, 1,044 potential target genes were selected for the 36 compounds of HJT after removing duplicate targets (Table S1). Moreover, the UniProt database was used to translate the official names of potential targets so that they could be used in network construction for further biological characterization.

Construction of a compound-putative target network of HJT

Chinese herbal compounds can interfere with diseases by regulating a network through binding multiple targets. Therefore, a network of “compound-putative target” was established to predict these targets through the acquisition of detailed information on the bioactive ingredients of HJT. This network consisted of 969 nodes and 3,188 edges (Figure 2A), indicating the interactions of chemical compounds and putative targets. As can be seen from the figure, different active compounds corresponded to different targets, which reflected the characteristics of multi-compound, multi-target of HJT. Among them, the degree values of C-C8 (p-Coumaric acid), A-C9 (Ethyl gallate), C-C4((7R,8R)-3-Methoxy-8'-carboxy-7'-en-3',8-epoxy-7,4'-oxyneolignan-4,9-diol), A-C7 (4-Hydroxyphenylacetic acid), C-C10 (6-Feruloyloxyhexanoic acid), A-C3 (Vanillic acid β-glucoside) are the highest, 245, 223, 173, 166, 140 and 121 respectively, which indicated that they were the most important active compounds in the network (Figure 2B).

Figure 2 The information of HJT and GSE20247. (A) The compound-target network. The green nodes represent active compounds and the yellow nodes represent potential protein targets. The edges represent the interactions between them and nodes size are proportional to their degree. (B) Top 20 active compounds with degree value. (C) A volcano plot of the distribution of DEGs. HJT, Hongjingtian injection; DEG, differentially expressed gene.

Identification of TIF-related DEGs

Differential genetic analysis between TIF and healthy individuals was performed with |log2 FC| >1 and P<0.01. Ultimately, 1,283 DEGs were identified. A volcano plot of the distribution of DEGs is shown in Figure 2C. Among them, 580 up-regulated genes are represented by red dots, and 703 down-regulated genes are represented by green dots.

Analyses of enrichment of the KEGG pathway

As shown in Figure 3A, HJT has 79 genes that act on TIF. Table 2 lists the specific details. The 79 cross gene were mapped to 218 pathways by KEGG enrichment analysis, the main signaling pathways involved in the treatment of TIF were identified and the first 25 pathways related to TIF and significantly enriched screened were screened (FDR <0.05) (Table 3), including Metabolic pathways (ko01100), Pathways in cancer (ko05200), MicroRNAs in cancer (ko05206), Relaxin signaling pathway (ko04926) among others. As the Figure 3B shown, the y-axis represents the KEGG pathway and the x-axis indicates the number of genes enriched in that pathway. The redder the color, the smaller the value of P_adjust; it also denotes greater credibility and greater importance. In contrast, the bluer the color, the greater the value of P_adjust. Addition, as shown in Figure 3C, enrichment of the KEGG pathway could be divided approximately into energy metabolism, environmental information processing, cellular process, organismal systems and human diseases. It is worth noting that the most genes are enriched in metabolic pathways (ko01100). This pathway contains a total of 1,554 genes, of which 24 of the above-mentioned 79 cross genes are included. And it contains 7 up-regulated genes and 17 down-regulated genes.

Figure 3 KEGG analysis. (A) 79 overlapping gene symbols between the disease and drug. (B) Top 25 pathways of KEGG enrichment. The y-axis represents the KEGG pathway and the x-axis indicates the number of genes enriched in that pathway. (C) Classification of 25 pathways. KEGG, Kyoto Encyclopedia of Genes and Genomes.

PPI network and co-expression network construction and analysis

The 79 cross genes were imported into STRING and the PPI network was obtained of which contains 65 nodes and 132 edges (Figure 4A). Subsequently, the topological properties of the afore-mentioned merged PPI network were analyzed according to three key parameters: BC, CC, and DC, screened targets above median values of BC, CC, and DC, thereby selecting the core genes of the TIF-treated effect of HJT. The cutoff value of the screening was BC >0.015, CC >0.304 and DC >3. These 25 core genes are marked in pink. Besides, the heat map of these 25 core genes were listed in Figures 4B. Finally, as shown in the Figure 4C, co-expression relationships in the PPI network were analyzed and further higher-scoring cluster containing 25 genes were shown in purple. Through the Venn diagram we found that there are 5 key genes enriched in the Metabolic pathway (ko01100), which were AKR1B1 (log2fc =−1.341623), GAD1 (log2fc =−1.563923), P4HA2 (log2fc =1.069761), SPHK1 (log2fc =2.22411) and PTGES (log2fc =−2.268903). And there were three intersections between these five genes and the sub-network genes of the co-expression network, SPHK1, AKR1B and PTGES, respectively. These results illustrate that AKR1B1, GAD1, P4HA2, SPHK1 and PTGES may be key genes in HJT to regulate TIF through metabolic pathway, and there may be a co-expression relationship among SPHK1, AKR1B1 and PTGES.

Figure 4 Analysis of PPI and key genes. (A) Protein-protein interaction network of HJT acting on TIF. Network nodes represent proteins, edges represent protein-protein associations. The pink nodes represent 25 core genes. (B) The heatmap of core genes. (C) Co-expression relationships in the PPI network were analyzed and further higher-scoring cluster containing 25 genes were shown in purple. There are 5 key genes enriched in the Metabolic pathway, three of which have co-expression relationship, SPHK1, AKR1B1 and PTGES, respectively. PPI, protein-protein interaction; HJT, Hongjingtian injection; TIF, Tubulointerstitial fibrosis.

TGF affects the expression of AKR1B1, GAD1, P4HA2, SPHK1 and PTGES

We chose TGF-β1 to represent the severity of EMT in vitro cell models. In the above network pharmacological analysis, we predicted that AKR1B1, GAD1, P4HA2, SPHK1 and PTGES in the metabolic pathway were key targets of HJT anti TIF. Thus, we chose these five key genes for further experimental verification. HK-2 cells were pretreated with TGF-β1 (0, 2, 5, 10 ng/mL) for 48 hours. Real-time polymerase chain reaction analysis results showed that compared with the control group (no TGF-β1 pre-treatment), AKR1B1, GAD1, PTGES were down-regulated and SPHK1 and P4HA2 were up-regulated by treatment of 10 ng/mL TGF-β1 in HK-2 cells (Figure 5).

Figure 5 The effect of different concentrations of TGF-β1 on five key genes was determined by qPCR. *, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001; ns, P>0.05. CON, control group; qPCR, quantitative polymerase chain reaction.

These results are consistent with the expression trend of the microarray we explored. Finally, we selected TGF-β1 at 10 ng/mL to further investigate its effects on HJT to the expression of predicted five major gene targets based on the fact that none of the three concentrations of TGF-β1inhibited cell activity.

Optimal concentration of HJT

It cannot be ignored that although HJT has a protective effect on HK-2 cells, high dose may cause cell death. In order to explore protective effects of HJT on the HK-2 cells, CCK-8 assay was used to evaluate the survival rate of HK-2 at the doses of 0, 1.25, 2.5, 5, 10, 20, 40, 80, 200, 1,000 mL/L of HJT. As the results showed in Figure 6 that the HJT at the dose within the range of 0–2.5 mL/L has little significant toxicity in HK-2 cell. The half-maximal inhibitory concentration (IC50) of HJT for HK-2 cells was 99.81 mL/L. Therefore, 2.5 mL/L HJT was selected for subsequent experiments.

Figure 6 Evaluate the survival rate of HK-2 at the different doses of HJT using CCK-8 assay. *, P<0.05; **, P<0.01; ***, P<0.001; ns, P>0.05. HJT, Hongjingtian injection.

Effect of HJT on the expression of key targets in TGF-β1-induced HK-2

qRT-PCR was used to further validate SPHK1, GAD1, P4HA2, AKR1B1 and PTGES for HJT treatment of TIF (Figure 7A). The results showed that HJT could reverse the expression of SPHK1 with statistical significance (P<0.05). However, the improvement of other genes was not significant (P>0.05). The WB results demonstrated that the expression level of SPHK1 was promoted in HK-2 (P<0.01). After treatment of HJT, the expression of SPHK1 was markedly suppressed (P<0.05) (Figure 7B,7C).

Figure 7 Effect of HJT on five key genes in HK-2 interfered by TGF-β1. (A) Effect of HJT on the expression of key targets in TGF-β1-induced HK-2. *, P<0.05. **, P<0.01. (B,C) The expression level of SPHK1 was promoted in HK-2 (**, P<0.01). After the treatment of HJT, the expression of SPHK1 was markedly suppressed (*, P<0.05). HJT, Hongjingtian injection.

The progression of EMT was promoted by TGF-β1 but restored by HJT

The immunofluorescence assay showed that TGF-β1 upregulated Vimentin and downregulated cytokeratin 18 in HK-2 cells. On the contrary, the expression of them was reversed after intervention with HJT (Figure 8A,8B). Besides, HK-2 migration capacity was measured through wound-healing assay and Transwell assay. As Figure 8C,8D shown, the first row was the scratch size of control, TGF-β1, HJT and TGF-β1 and HJT group at 0 h, which was used as a baseline comparison, while the second and third lines were the scratch size at 24 and 48 h in the corresponding groups, respectively. After 24 hours of intervention, TGF-β1 showed a 16.7% increase compared with control group (P<0.05), and 5.6% decrease of TGF-β1 + HJT to TGF-β1 (P>0.05). After 48 hours of intervention, TGF-β1 showed a 21.1% increase compared with control group (P<0.05), and 16.7% decrease of TGF-β1 + HJT to TGF-β1 (P<0.05). To conform the effects on cells migration, Transwell assay was carried out and TGF-β1 showed a 550.2% increase in comparison with control group (P<0.0001). TGF-β1 + HJT showed a 44.9% decrease in comparison with TGF-β1 (P<0.0001, Figure 8E,8F). Taken together, TGF-β1 could regulate vimentin and cytokeratin 18, and promote the migration of HK-2 cells, consequently accelerated the EMT progression. However, HJT can reverse the phenomenon.

Figure 8 The progression of EMT was promoted by TGF-β1 but restored by HJT. (A,B) The immunofluorescence assay showed that TGF-β1 upregulated Vimentin and downregulated cytokeratin18 in HK-2 cells. On the contrary, the expression of them was reversed after intervention with HJT (40X). (C,D) Wound healing assay. The three lines were the scratch size at 0, 24 and 48 h in the corresponding groups, respectively (40×). (E,F) Transwell assay (Crystal violet stain, 40×). *, P<0.05. **, P<0.01. ****, P<0.0001. HJT, Hongjingtian injection. DAPI, 4',6-diamidino-2-phenylindole, a fluorescent dye with strong binding to DNA. VIM, Vimentin.

Discussion

TIF is a key and irreversible process in the development of CKD. Existing treatment strategies are insufficient to prevent disease development. Therefore, more effective treatments are needed to delay the progression of TIF. In recent years, traditional Chinese medicine has become popular in western countries due to its reliable efficacy. In TCM theory, TIF is called edema, strangury, consumptive disease and so on. Its pathological mechanism can be roughly summarized as deficiency, blood stasis, phlegm dampness and turbid toxin. Rhodiola, as one of the traditional Tibetan medicines in China, has long existed in the history of medicine. It has the effect of invigorating Qi and promoting blood circulation, clearing the pulse and relieving asthma. In modern pharmacology, Rhodiola has anti-diabetic, anti-inflammatory, anti-cancer, anti-aging, cardio-protective, and neuro-protective effects (12). However, as a preparation extracted from Rhodiola, the mechanism of HJT in treating TIF is still unclear due to the complex compounds of traditional Chinese medicine. Thus, this study aimed to clarify the potential mechanism of HJT in the treatment of TIF based on network pharmacology.

We screened out 36 effective active compounds and 1,044 potential targets of HJT. Among the main active compounds of HJT, p-Coumaric acid (p-CA) is a phenolic acid of the hydroxycinnamic acid family (13). Extensive studies have shown that p-CA is related to various bioactivities, including antioxidant, anti-inflammatory, anti-cancer and antidiabetic (14). Mani et al. has revealed that p-CA was found to offer renal protection from oxidative stress by decreasing lipid peroxidation and increasing the activities of antioxidant enzymes in treated diabetic rats (15). Zabad et al. also reported that P-CA are able to decrease the fibrotic cytokines and protect against the progression of DKD (16). Ethyl gallate (EG) is a hydroxylated, ethyl ester of benzoic acid. EG has antioxidant, anti-inflammatory and anti-proliferative effects. Cui et al. (17) demonstrated that EG suppresses proliferation and invasion in human breast cancer cells by modulating the PI3K/Akt pathway, and inhibit downstream targets such as NF-κB p-65, Bcl-2/Bax. Ahn et al. (18) proposed that EG targeting PTPN6 and PPARγ exerts an anti-diabetes effect. Crispo et al. (19) suggested that EG protects PC12 cells against oxidative stress induction through the Nrf2 pathway. As such, it has been proved that EG can protect against diabetes, neuro-degenerative disease, endothelial inflammation and some types of cancer (20). In addition, both 4-Hydroxyphenylacetic acid 4 and Vanillic acid β-glucoside showed potential antioxidant effects (21,22).

Meanwhile, the GSE20247 dataset was downloaded from the GEO database to obtain the gene expression data of TGF intervention HK-2 cells, and then analyzed the DEGs. In addition, PPI network and functional enrichment analysis were performed on common genes to understand the meaningful pathway between TGF-β1-induced HK-2 and normal control. Of which, Metabolic pathways is the pathway enriched for the most genes compared to other pathways. We speculated that HJT might participate in the treatment of TIF by regulating metabolic pathways. Therefore, five target genes related to metabolic pathways were selected for further research (AKR1B1, GAD1, P4HA2, SPHK1 and PTGES).

Our result shows that although it was found that the five key genes in HK-2 interfered by TGF-β1 were consistent with the trend shown by the GSE20247, HJT only had a significant reversal effect on SPHK1. The metabolites of sphingolipid include ceramides, sphingosine and sphingosine-1-phosphate (S1P). S1P has a function of cell proliferation and differentiation, while ceramide and sphingosine induced cell growth arrest and apoptosis and migration (23). SPHK1 is a key metabolic enzyme that catalyzes the synthesis of S1P by sphingosine (24). Fibronectin (FN) is one of ECM components and its accumulation will eventually lead to TIF (25). Chen et al. demonstrated that AGEs-RAGE could upregulate FN through regulating SPHK1 in vitro cell experiments (26). Huang et al. (27) confirmed the regulatory effect of SPHK1 on NF-κB-mediated diabetic renal fibrosis in vivo and in vitro experiments. Compared with diabetic mice, fewer renal fibrotic lesions, FN accumulation and NF-κB nuclear accumulation exhibited in glomeruli of kidneys of SPHK1^-/- diabetic mice. He et al. (28) indicated that the inhibition of SPHK1 contribute to decrease the EMT by blocking the NF-κB signaling, thereby protecting HK-2 against TIF. Liu et al. (29) confirmed that SPHK1/S1P signaling upregulates microRNA-21 in renal TECs interfered by TGF-β, thus promoting the overexpression of ECM proteins and TIF. Huang et al. (30) reported that the expression of SPHK1 and S1P were significantly down-regulated by curcumin in diabetic rat kidneys and glomerular mesangial cells (GMCs) with high glucose intervention. Meanwhile, SPHK1-S1P-mediated FN and TGF-β1 were inhibited. In addition, Lan et al. (31) also found that berberine exerts renoprotective effects on DKD through SPHK1–S1P signaling pathway. S1P is catalytically synthesized by Sphingosine kinases and catabolized by S1P phosphatases and S1P lyase (S1PL). Huang et al. (32) suggested that the expressions of SPHK1 were increased in bleomycin challenged mice. Genetic knockdown of SPHK1 ameliorated pulmonary fibrosis in mice while deletion of S1PL potentiated fibrosis. Taken together, the present study in vitro and in vivo demonstrated that SPHK1 is closely related to the pathological process of EMT and TIF. Therefore, inhibiting SPHK1 signaling may be a promising therapeutic strategy in TIF.

During the process of TIF, epithelial cells undergo partial EMT. Partial EMT cells acquire the ability to produce profibrotic cytokines/growth factors while remaining attached to the basement membrane during fibrosis (33). In addition, a study (34) has found that compared with the normal structure of renal tubules in healthy subjects, renal sections of diabetes patients show thickening and hypertrophy of renal tubules. As a key mediator of TIF, TGF-β1 activates different intracellular signaling pathways during the process of EMT (35), contributing to the secretion of collagen and inducing its deposition, leading to tissue fibrosis (36). Besides, TGF-β1 is also closely related to hypertrophy of renal tubules (37). Therefore, we established an in vitro TIF cell model by intervening HK-2 cells with TGF-β1. During EMT, epithelial cells lose key phenotypic markers and acquire mesenchymal markers (38). In our results, it was clearly seen that HJT alone did not affect the migration and phenotypic changes of HK-2 without TGF-β1 intervention. The results of immunofluorescence showed that the expression of cytokeratin 18 decreased and the expression of vimentin increased after TGF-β1 intervention, and then reversed by HJT. In the wound healing assay and Transwell assay, compared with the control group, the TGF-β1 experimental group significantly improved the cell migration ability, but the ability decreased after adding HJT. To sum up, HJT has an inhibitory effect on EMT and migration ability. HJT treatment of TIF may reverse EMT caused by TGF-β1 by targeting SPHK1.

Conclusions

In conclusion, HJT treats TIF through multi-component, multi-target and multi-directional. In particular, SPHK1 may be the most important target of HJT in TIF. Our research revealed the mechanism of HJT in the process of action from a holistic and systematic perspective, which provides a theoretical basis for further research and application of HJT in the future. However, whether HJT is indeed suitable for the prevention and treatment of TIF still needs to be confirmed by future clinical trials. Therefore, in vivo experimental validation of potential active ingredients is urgently needed to help further evaluate the therapeutic potential of HJT.

Acknowledgments

Funding: This work was supported by the National Natural Science Foundation of China (No. 82105044), the Natural Science Foundation of Shandong Province (No. ZR2020QH310), and Traditional Chinese Medicine Science and Technology Project of Shandong Province (No. M-2022141).

Footnote

Reporting Checklist: The authors have completed the MDAR reporting checklist. Available at https://atm.amegroups.com/article/view/10.21037/atm-22-5035/rc

Data Sharing Statement: Available at https://atm.amegroups.com/article/view/10.21037/atm-22-5035/dss

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://atm.amegroups.com/article/view/10.21037/atm-22-5035/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013).

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Webster AC, Nagler EV, Morton RL, et al. Chronic Kidney Disease. Lancet 2017;389:1238-52. [Crossref] [PubMed]
2. Djudjaj S, Boor P. Cellular and molecular mechanisms of kidney fibrosis. Mol Aspects Med 2019;65:16-36. [Crossref] [PubMed]
3. Zhang Z, Ma J, Feng R, et al. Centella asiatica inhibits renal interstitial fibrosis by regulating Smad3 and Smad7 expression in the TGFbeta signaling pathway. Int J Clin Exp Pathol 2018;11:1009-17.
4. Gu LY, Yun S, Tang HT, et al. Huangkui capsule in combination with metformin ameliorates diabetic nephropathy via the Klotho/TGF-beta1/p38MAPK signaling pathway. J Ethnopharmacol 2021;281:113548. [Crossref] [PubMed]
5. Li X, Chen S, Shao W. Effect of Hongjingtian injection on serum high-sensitivity C-reactive protein and degree of atherosclerosis in patients with essential hypertension. Shaanxi Journal of Traditional Chinese Medicine 2017;38:161-2.
6. Zhao J, Liu Y, Chen S. Effect of Hongjingtian injection on hemorheology in patients with type 2 diabetes mellitus. Clinical Journal of Chinese Medicine 2016;8:29-30.
7. Guo C, Li Y, Zhang R, et al. Protective Effect of Salidroside Against Diabetic Kidney Disease Through Inhibiting BIM-Mediated Apoptosis of Proximal Renal Tubular Cells in Rats. Front Pharmacol 2018;9:1433. [Crossref] [PubMed]
8. Liu Y, Chen C, Qiu J, et al. Characterization of the chemical constituents in Hongjingtian injection by liquid chromatography quadrupole time-of-flight mass spectrometry. Biomed Chromatogr 2019;33:e4446. [Crossref] [PubMed]
9. Kanehisa M, Goto S. KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res 2000;28:27-30. [Crossref] [PubMed]
10. Szklarczyk D, Morris JH, Cook H, et al. The STRING database in 2017: quality-controlled protein-protein association networks, made broadly accessible. Nucleic Acids Res 2017;45:D362-8. [Crossref] [PubMed]
11. Tang Y, Li M, Wang J, et al. CytoNCA: a cytoscape plugin for centrality analysis and evaluation of protein interaction networks. Biosystems 2015;127:67-72. [Crossref] [PubMed]
12. Lee Y, Jung JC, Jang S, et al. Anti-Inflammatory and Neuroprotective Effects of Constituents Isolated from Rhodiola rosea. Evid Based Complement Alternat Med 2013;2013:514049. [Crossref] [PubMed]
13. Cha H, Lee S, Lee JH, et al. Protective effects of p-coumaric acid against acetaminophen-induced hepatotoxicity in mice. Food Chem Toxicol 2018;121:131-9. [Crossref] [PubMed]
14. Amalan V, Vijayakumar N, Indumathi D, et al. Antidiabetic and antihyperlipidemic activity of p-coumaric acid in diabetic rats, role of pancreatic GLUT 2: In vivo approach. Biomed Pharmacother 2016;84:230-6. [Crossref] [PubMed]
15. Mani A, Kushwaha K, Khurana N, et al. p-Coumaric acid attenuates high-fat diet-induced oxidative stress and nephropathy in diabetic rats. J Anim Physiol Anim Nutr (Berl) 2022;106:872-80. [Crossref] [PubMed]
16. Zabad OM, Samra YA, Eissa LA. P-Coumaric acid alleviates experimental diabetic nephropathy through modulation of Toll like receptor-4 in rats. Life Sci 2019;238:116965. [Crossref] [PubMed]
17. Cui H, Yuan J, Du X, et al. Ethyl gallate suppresses proliferation and invasion in human breast cancer cells via Akt-NF-kappaB signaling. Oncol Rep 2015;33:1284-90. [Crossref] [PubMed]
18. Ahn D, Kim J, Nam G, et al. Ethyl Gallate Dual-Targeting PTPN6 and PPARγ Shows Anti-Diabetic and Anti-Obese Effects. Int J Mol Sci 2022;23:5020. [Crossref] [PubMed]
19. Crispo JA, Piché M, Ansell DR, et al. Protective effects of methyl gallate on H2O2-induced apoptosis in PC12 cells. Biochem Biophys Res Commun 2010;393:773-8. [Crossref] [PubMed]
20. Wang HR, Sui HC, Ding YY, et al. Stimulation of the Production of Prostaglandin E2 by Ethyl Gallate, a Natural Phenolic Compound Richly Contained in Longan. Biomolecules 2018;8:91. [Crossref] [PubMed]
21. Zhao H, Jiang Z, Chang X, et al. 4-Hydroxyphenylacetic Acid Prevents Acute APAP-Induced Liver Injury by Increasing Phase II and Antioxidant Enzymes in Mice. Front Pharmacol 2018;9:653. [Crossref] [PubMed]
22. Kayano S, Kikuzaki H, Fukutsuka N, et al. Antioxidant activity of prune (Prunus domestica L.) constituents and a new synergist. J Agric Food Chem 2002;50:3708-12. [Crossref] [PubMed]
23. Takehara M, Bandou H, Kobayashi K, et al. Clostridium perfringens alpha-toxin specifically induces endothelial cell death by promoting ceramide-mediated apoptosis. Anaerobe 2020;65:102262. [Crossref] [PubMed]
24. Pulkoski-Gross MJ, Obeid LM. Molecular mechanisms of regulation of sphingosine kinase 1. Biochim Biophys Acta Mol Cell Biol Lipids 2018;1863:1413-22. [Crossref] [PubMed]
25. Gong W, Chen C, Xiong F, et al. CKIP-1 ameliorates high glucose-induced expression of fibronectin and intercellular cell adhesion molecule-1 by activating the Nrf2/ARE pathway in glomerular mesangial cells. Biochem Pharmacol 2016;116:140-52. [Crossref] [PubMed]
26. Chen C, Gong W, Li C, et al. Sphingosine kinase 1 mediates AGEs-induced fibronectin upregulation in diabetic nephropathy. Oncotarget 2017;8:78660-76. [Crossref] [PubMed]
27. Huang J, Li J, Chen Z, et al. Sphingosine kinase 1 mediates diabetic renal fibrosis via NF-kappaB signaling pathway: involvement of CK2alpha. Oncotarget 2017;8:88988-89004. [Crossref] [PubMed]
28. He F, Fan M, Jin Y, et al. Sphingosine kinase 1 inhibition decreases the epithelial-mesenchymal transition and ameliorates renal fibrosis via modulating NF-κB signaling. Am J Transl Res 2019;11:5879-87.
29. Liu X, Hong Q, Wang Z, et al. Transforming growth factor-β-sphingosine kinase 1/S1P signaling upregulates microRNA-21 to promote fibrosis in renal tubular epithelial cells. Exp Biol Med (Maywood) 2016;241:265-72. [Crossref] [PubMed]
30. Huang J, Huang K, Lan T, et al. Curcumin ameliorates diabetic nephropathy by inhibiting the activation of the SphK1-S1P signaling pathway. Mol Cell Endocrinol 2013;365:231-40. [Crossref] [PubMed]
31. Lan T, Shen X, Liu P, et al. Berberine ameliorates renal injury in diabetic C57BL/6 mice: Involvement of suppression of SphK-S1P signaling pathway. Arch Biochem Biophys 2010;502:112-20. [Crossref] [PubMed]
32. Huang LS, Natarajan V. Sphingolipids in pulmonary fibrosis. Adv Biol Regul 2015;57:55-63. [Crossref] [PubMed]
33. Grande MT, Sánchez-Laorden B, López-Blau C, et al. Snail1-induced partial epithelial-to-mesenchymal transition drives renal fibrosis in mice and can be targeted to reverse established disease. Nat Med 2015;21:989-97. [Crossref] [PubMed]
34. Habib SL. Alterations in tubular epithelial cells in diabetic nephropathy. J Nephrol 2013;26:865-9. [Crossref] [PubMed]
35. Sutariya B, Jhonsa D, Saraf MN. TGF-β: the connecting link between nephropathy and fibrosis. Immunopharmacol Immunotoxicol 2016;38:39-49. [Crossref] [PubMed]
36. Sekar V, Mani S, Malarvizhi R, et al. Positive interaction of mangiferin with selected oral hypoglycemic drugs: a therapeutic strategy to alleviate diabetic nephropathy in experimental rats. Mol Biol Rep 2020;47:4465-75. [Crossref] [PubMed]
37. Vallon V, Thomson SC. Renal function in diabetic disease models: the tubular system in the pathophysiology of the diabetic kidney. Annu Rev Physiol 2012;74:351-75. [Crossref] [PubMed]
38. Patel NA, Patel PS, Vora HH. Role of PRL-3, Snail, Cytokeratin and Vimentin expression in epithelial mesenchymal transition in breast carcinoma. Breast Dis 2015;35:113-27. [Crossref] [PubMed]