Network Pharmacology Study of Bioactive Components and Molecular Mechanisms of the Glycoside Fraction from Picrorhiza scrophulariiflora Against Experimental Colitis

Introduction

As one of the modern refractory diseases listed by the World Health Organization, the occurrence of inflammatory bowel disease (IBD) seriously affects the quality of life of patients.^1,2 To date, IBD is incurable, but its remission can be achieved and maintained with aminosalicylate, corticosteroids, immunosuppressants, and biologics.^3,4 However, the abovementioned IBD treatments have obvious defects, such as high side effects, high cost, repeated recurrence, and even aggravation.^5,6 Developing novel therapeutic strategies aimed at achieving durable clinical remission in the absence of serious adverse events has been a major topic in IBD research. Recently, many case studies have shown that traditional herbal medicine (THM) has great potential to induce IBD remission with low toxicity, suggesting that the development of new drugs from THM is a promising strategy for the treatment of IBD.^7–9

Picrorhiza scrophulariiflora Pennell is a perennial herb of genus Picrorhiza (Scrophulariaceae), mainly distributed in southeastern Tibet and northwestern Yunnan province in China.¹⁰ The rhizome of P. scrophulariiflora is a well-known bitter herbal medicine in traditional Chinese medicine (TCM) to treat infantile malnutrition, dysentery, jaundice, and hemorrhoids, which has been officially listed in Chinese Pharmacopoeia since 1977.¹¹ P. scrophulariiflora and its contained formulas are clinically used to treat colitis symptoms in China, and their clinical application has been documented in many ancient Chinese medical classics.¹² In addition, Picrorhiza kurroa Royle ex Benth, the sister plant of P. scrophulariiflora, is a famous traditional medicine in Ayurvedic medicine, which is extensively used to treat various immunity-related diseases, such as asthma and arthritis, in India and Nepal.¹³ Previous studies have revealed that P. scrophulariiflora mainly contains flavonoids, triterpenoids, iridoid glycosides, phenylethanol glycosides, and phenolic glycosides, which are highly similar to the chemical compositions of P. kurroa.¹⁴ Pharmacological studies have revealed that P. scrophulariiflora has multiple biological properties, including hepatoprotection, neuroprotection, cardiovascular protection, immunomodulation, antiasthma, and oxidation,^15,16 and that glycosides (glycosidic fraction) are the key components responsible for its efficacy.¹⁷ Based on all of above information, P. scrophulariiflora is a potentially effective regimen for the treatment of colitis. However, to date, no evidence-based study has been conducted to evaluate the efficacy of P. scrophulariiflora in colitis.

Therefore, we evaluated the protective effect of GPS, an effective fraction, from P. scrophulariiflora on dextran sulfate sodium (DSS)-induced colitis in C57BL/6 mice and LPS-stimulated RAW264.7 macrophages, and we investigated its anti-inflammatory bioactive components and mechanisms using network pharmacology.

Materials and Methods

Regents

DSS salt (colitis grade, molecular weight: 36,000–50,000 Daltons) was purchased from MP Biomedicals (Ohio, USA). The hematoxylin-eosin (H&E) staining solution was purchased from Sigma (St. Louis, MO, USA). Trizol reagent was obtained from Invitrogen (Carlsbad, USA), and Primescript™ RT reagent kits and SYBR^® Premix Ex Taq™ II kits were acquired from Takara Bio Inc. (Kusatsu, Japan). Antibodies including p44/42 MAPK (Erk1/2), phospho-p44/42 MAPK (Thr202/Tyr204), p38 MAPK, phospho-p38 MAPK (Thr180/Tyr182), SAPK/JNK, phospho-SAPK/JNK (Thr183/Tyr185), Akt, phospho-Akt (Thr450), and beta-Actin were purchased from Cell Signaling Technology (MA, USA).

Preparation of Different Fractions of P. scrophulariiflora

The rhizomes of P. scrophulariiflora were collected from the Yunnan Province of China in August 2021. The samples were authenticated by Prof. Yongxian Cheng (Health Science Center, Shenzhen University), and voucher specimens (No. Ps20210901) were stored at the Institute for Inheritance-Based Innovation of Chinese Medicine, Shenzhen University. The air-dried and powdered P. scrophulariiflora rhizomes were extracted twice with 8-fold 70% ethanol under reflux for 2h each time. The extracting solutions were combined and concentrated under reduced pressure to obtain an ethanol extract of P. scrophulariiflora (EPS). Previous studies have reported that glycosides are the key components responsible for their biological properties, such as immunomodulation, hepatoprotection, and cardiovascular protection.^{11,13,18–21} Subsequently, glycoside and non-glycoside fractions of P. scrophulariiflora were prepared using a macroporous adsorption resin D-101 column (Solarbio, Cat#M0041) and eluted successively with water and ethanol (water, 40%, and 100% ethanol), as previously reported.²² The eluate of water, 40% ethanol, and 100% ethanol were concentrated and freeze-dried to yield the water fraction (WPS), 40% ethanol (glycoside fraction, GPS), and 100% ethanol fraction (non-glycoside fraction, NGPS) of P. scrophulariiflora, respectively.

UPLC-QTOF-MS Analysis of the GPS Extract

We identified the components in the GPS extract using the UPLC-QTOF-MS method, as previously reported, but with minor changes.¹ The mobile phase was (A) 0.1% formic acid in water and (B) 0.1% formic acid acetonitrile. The optimized linear gradient was as follows: flow rate 0.40 mL/min, 0–6 min, 5–75% B; 6–9 min, 75–100% B; 9–12 min, 100% B; 12–12.10 min, 100–5% B; and 12.10–15 min, 5%B.

Cell Culture and Viability Assay

RAW 264.7 cells were obtained from ATCC (Manassas, VA, United States) and cultured routinely, as in our previous studies.^23,24 Cell viability was detected with a CCK-8 kit according to the manufacturer’s instructions.

NO Release Assay

In total, 5×10⁴ cells/well were seeded in 48-well plates overnight and treated with GPS extract (50 to 150 µg/mL) for 1 h, and then challenged with 1 µg/mL LPS for 24 h. Nitric oxide (NO) inhibition (%) was determined by a NO release assay using Griess reagent according to the manufacturer’s instructions.

Colitis Induction and Evaluation

Male C57BL/6 mice (6–8 weeks old, 20–22 g) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). The animal studies were approved by the Ethics Committee and performed according to Institutional Guidelines and Animal Ordinance (Health Science Center, Shenzhen University, No. 2018020). Before the experiment, the animals were acclimated to the feeding environment for a week.

Colitis was copied in mice with 2% DSS drinking water, as in our previous study,² and then colitis mice were randomly divided into DSS model, cyclosporin A (CSA), and two GPS extract treated groups. At the same time, the vehicle control group without DSS was set up. During the experiment, the doses of GPS extract were set at 200 and 400 mg/Kg, respectively, according to the preliminary experiment, and the mice were intragastric for 7 consecutive days. In parallel, the dose of CSA was set at 25 mg/Kg as in a previous study.²⁵ The vehicle control and DSS model groups were given the same volume of vehicle. Body weight, fecal consistency, and fecal occult blood were recorded daily. The disease activity index (DAI) was determined based on the above three items, as previously described.^26–28 At the end of the experiment, all mice were euthanized, and the colonic segment was excised. After measuring the length of the colon, colon tissues were collected for further experiments.

Histological Examination

Colonic tissues were fixed in 4% paraformaldehyde, embedded in paraffin, stained with hematoxylin and eosin (H&E), and histopathological damage was evaluated blinded and scored as described previously.²³

Real-Time PCR Analysis

The total RNA of RAW264.7/colon tissue was extracted using Trizol reagent, following the manufacturer’s instructions. Reverse transcriptase was used to synthesize the cDNA. Amplification was performed using the SYBR Green master, and the amount was quantitatively monitored in real time. The transcription of the target gene was normalized by β-actin and data were analyzed using the 2^−ΔΔCT method. The primer sequences are shown in Table S1.

Western Blot, ELISA Analysis, and Myeloperoxidase (MPO) Determination

Proteins from RAW264.7 cells/colon tissues were extracted with ice cold RIPA buffer containing protease and phosphatase inhibitors, and analyzed by immunoblotting as previously reported.^23,24 The primary antibodies used were Akt, p-Akt, p38, p-p38, ERK, p-ERK, JNK, p-JNK, and β-actin. The intensity of the bands was measured using ImageJ. In parallel, total protein was extracted from colon tissue for cytokine assays, as described in our previous studies.^23,24 Then, the levels of IL-6 and IL-1β in colonic homogenates and supernatants of LPS-treated RAW 264.7 cells were measured using ELISA kits following the manufacturer’s protocol. The levels of MPO in supernatants extracted from colon tissues were also determined according to the manufacturer’s protocols.

Network Pharmacology Analysis

The construction of the GPS prediction targets database was based on LC-MS-MS identification results (Table S2) and network databases, including Pubchem (https://pubchem.ncbi.nlm.nih.gov/), SwissTargetPrediction (http://swisstargetprediction.ch/), and TargetNet (http://targetnet.scbdd.com/). And the construction of IBD and anti-inflammatory related predicted targets database was based on network database including GEO (https://www.ncbi.nlm.nih.gov/gds), OMIM (https://omim.org/), GeneCards (https://www.genecards.org/), DisGeNET (https://www.disgenet.org/), CTD (http://ctdbase.org/) database with “inflammatory bowel disease” and “anti-inflammatory” as keywords, respectively. The selection of validated human species-associated target genes in the STRING (https://cn.string-db.org/) database and the drug-, disease-, and anti-inflammation-related genes were intersected using a Venn diagram. Then, the intersection targets were imported into the STRING database for an analysis of the protein–protein interactions of the predicted targets and visualized using Cytoscape 3.8.2. Network topology analysis and screening were performed to derive the hub genes using Cytoscape’s analysis tool. Subsequently, in the R software, “clusterProfiler”, “org.Hs.eg.db”, “enrichplot”, “pathview”, “ggnewscale”, and “DOSE” were imported, and the q-value was set to 0.05. Then, we targeted hub genes with the filtered core intersection, ran the script, and drew a KEGG enrichment analysis bubble plot.

Molecular Docking

AutoDock Vina- 1.2.0 was used for all molecular docking. The crystallin structures of ERK (PDB: 6GDM), JNK (PDB: 1UKI), P38 (PDB: 3GCP), and Akt (PDB: 4GV1) from the PDB (https://www1.rcsb.org/) database were used as docking targets. The validated ERK inhibitor SCH772984, JNK inhibitor SP600125, P38 inhibitor SB203580, and Akt inhibitor AZD5363 were used as reference ligands. Prior to docking, the classical MM2 force field was used to optimize the structure of the small molecules. These models determine the free energy of binding between the ligand and the receptor, with lower binding free energies indicating stronger ligand receptor interactions. It is generally accepted that binding energies between ligands and receptors that are less than −4.25 kcal/mol, −5.0 kcal/mol, or −7.0 kcal/mol, indicate a certain, good, or strong binding activity, respectively.²⁹ Finally, PyMOL-2.5.2 was used to visualize the molecular docking results.

Statistical Analysis

Statistical analysis was executed using GraphPad Prism 8.0 with a one-way analysis of variance (ANOVA) and Duncan’s multiple range tests. KEGG and GO enrichment analysis was processed using the R package with clusterProfiler. P-values < 0.05 were considered statistically significant.

Results

GPS Exhibits Potent Anti-Inflammatory Activity on LPS-Induced RAW264.7 Cells

After determining the safe concentrations of EPS, WPS, GPS, and NGPS fractions used in RAW 264.7 cells, their anti-inflammatory activities were preliminarily screened by NO release assay under LPS stimulation. As shown in Figure S1, the inhibitory activity of the GPS fraction against NO release was significantly higher than that of the WPS and NGPS fractions, and verified that the GPS fraction was mainly responsible for the anti-inflammatory efficacy of P. scrophulariiflora. Subsequently, we further investigated the effects of GPS on other inflammatory markers in LPS-stimulated RAW 264.7 cells. As shown in Figure 1, LPS treatment significantly increased NO release, transcription and expression of iNOS in RAW264.7 cells, but these increases were significantly reversed after GPS treatment (Figures 1A–C). In addition, LPS significantly increased the transcription and production of inflammatory cytokines IL-1β and IL-6, but these were also suppressed by GPS treatment in a dose-dependent manner (Figures 1D and E). Collectively, these data indicate that GPS has a potent inhibitory effect against LPS-induced inflammation in vitro.

Figure 1 Anti-inflammatory effects of GPS on LPS-induced RAW246.7 cells in vitro. (A) NO release in LPS-stimulated RAW264.7 cells. (n = 4) (B) mRNA expression of iNOS, (n = 6). (C) Protein expression of iNOS, (n = 3). (D) mRNA expression of IL-6 and IL-1β, (n = 6). (E) The production of IL-6 and IL-1β, (n = 3). Data represent the mean ± SD. ##p < 0.01 or ###p < 0.001 vs control group; *p < 0.05 or ***P < 0.001 vs LPS-stimulated group.

GPS Exhibits Potent Anti-Inflammatory Activity Against DSS-Induced Colitis in Mice

The in vivo anti-inflammatory effects of GPS were determined in DSS-induced colitis mice. As shown in Figures 2A and B, mice treated with GPS or CSA showed a marked improvement in clinical symptoms of colitis, presenting lower body weight loss and DAI scores compared to the mice from the DSS model. Compared with the control group, mice treated with DSS showed shortened colons, which is a key indicator of colonic inflammation. As expected, this was significantly reversed by the treatment with both low- and high-dose GPS (Figure 2C). DSS treatment also resulted in extensive colonic tissue damage, including ulceration, inflammatory cell infiltration, and surface epithelial destruction, with a high histological score. However, mice treated with low and high doses of GPS exhibited less colonic damage and lower histological scores (Figures 2D and E). In addition, colonic MPO activity, a marker of neutrophil infiltration, was greatly increased in the DSS model group, whereas the MPO activities in the GPS-treated groups were also significantly suppressed (Figure 2F). Finally, we detected the transcription and production of inflammatory mediators in the colon tissues of colitis mice. As shown in Figure 3, compared to the mice in the control group, the transcription of IL-1β, IL-6 and iNOS, as well as the production of IL-1β and IL-6, were significantly elevated, whereas the transcription of IL-10 was dramatically decreased in the colon tissues of colitis mice. However, these inflammatory response changes were also significantly reversed by both GPS and CSA treatment. Taken together, these data indicate that GPS has potent anti-inflammatory properties against experimental colitis.

Figure 2 The effects of GPS on DSS-induced colitis in mice. (A) Schematic diagram of animal experiment. (B) body weight change, (C) disease activity index (DAI) score, (D) representative images of the colon and colon length, (E) histological score and H&E staining of the colon (magnification, ×100). (F) myeloperoxidase (MPO) activity. Data are expressed as the mean± SEM (n = 6). #p < 0.05 or ##p < 0.01 or ###p < 0.001 vs control group; *p < 0.05 or **p < 0.01 or ***P < 0.001 vs DSS group.

Figure 3 The effects of GPS on inflammatory mediators in the colon tissues of colitis mice. (A) mRNA expression of IL-1β, IL-6, iNOS and IL-10. (B) The production of IL-1β and IL-6. Data are expressed as the mean± SEM (n = 6). #p < 0.05 or ##p < 0.01 or ###p < 0.001 vs control group; *p < 0.05 or **p < 0.01 or ***P < 0.001 vs DSS group.

Network Pharmacology Predicts Potential Bioactive Components and Anti-Inflammatory Mechanisms of GPS Against Colitis

To explore the potential bioactive components and anti-inflammatory mechanisms of GPS against colitis, a network pharmacology analysis was performed. First, compounds contained in GPS were analyzed using a UPLC-ESI-MS/MS system, and 29 peaks representing individual chemical constituents were identified by comparing their retention time (tR), m/z, and ultraviolet (UV) absorption characteristics (lambda max) with relevant references (Figures 4, S2 and Table S2). Subsequently, compound-, IBD-, and anti-inflammatory-related targets were screened according to the guidelines of previous studies.^30–33 In brief, 918 GPS-related targets were obtained from database Pubchem, SwissTargetPrediction and TargetNet based on the chemical components of GPS identified in Table S2; 8021 IBD-related targets were obtained from database OMIM, GeneCards, DisGeNET, and differential genes extracted from GSE9452 chip; and 1013 anti-inflammatory-related targets were obtained from database OMIM, GeneCards, and CTD, respectively. As shown in Figure 5A, 361 crossover targets were screened and identified as potential targets of GPS for anti-inflammatory activities to treat colitis. Then, a PPI network was constructed to systematically summarize the interactions of the 361 crossover GPS targets related to IBD therapy. As shown in Figure 5B, 77 predicted hub genes were obtained, of which the top 15 were MAPK3, RELA, STAT3, MAPK1, JUN, SRC, TNF, MAPK14, EP300, IL6, FOS, HSP90AA1, Akt1, and TP53, which are considered crucial targets of GPS against colitis. To further screen the pathways corresponding to GPS targets, a KEGG analysis was performed, and the results showed that PI3K/Akt and MAPK signaling pathways were closely related to these GPS-target genes (Figure 5C). In addition, a compound-target network was constructed to predict the anti-inflammatory bioactive components of GPS against colitis. The results showed that luteolin, apocynin, caffeic acid, caffeic acid methyl ester, luteoloside, picroside II, aucubin, cinnamic acid, vanillic acid, and sweroside might be candidate bioactive substances for the treatment of IBD (Figure 5D).

Figure 4 UPLC-Q-TOF-MS total ion current chromatogram of GPS in the positive ion mode. (2) Cinnamic acid; (3) Sweroside; (4) Apocynin; (6) Luteoloside; (15) Aucubin; (17) Picroside II; (19) Luteolin; (24) Caffeic acid; (27) Caffeic acid methyl ester; (29) Vanillic acid.

Figure 5 Network pharmacology analysis of potential mechanisms of GPS against colitis in mice. (A) Venn diagram of compound targets of GPS, inflammation-related targets and IBD-related targets. (B) predicted hub targets network. The darker the red color indicates the higher degree. (C) Top 20 of KEGG pathway enrichment in targets of GPS treatment of IBD. (D) Compounds- predicted hub targets network. The diamond nodes represented the compounds, and the larger the node and the darker the color, the greater the degree.

Verification of the Anti-Inflammatory Mechanism of GPS in vivo and in vitro

To verify the regulatory effect of GPS on PI3k/Akt and MAPK signaling pathways, we examined the protein expression of p-Akt, p-p38, p-ERK, and p-JNK in LPS-induced RAW 264.7 cells and in the colon tissues of DSS-induced colitis mice. As shown in Figure 6, GPS treatment significantly suppressed inflammation-induced upregulation of p-Akt, p-p38, p-ERK, and p-JNK in LPS-induced RAW 264.7 cells and colon tissues of DSS-induced colitis mice, verifying that GPS was effective in regulating PI3k/Akt and MAPK signaling pathways in vivo and in vitro.

Figure 6 GPS inhibits activation of ERK, JNK, p38 and Akt in colon tissues of DSS-treated mice and LPS-stimulated RAW264.7 cells. (A) p-ERK, p-JNK, p-p38 and p-Akt expressions in LPS-stimulated RAW264.7 cells. RAW264.7 cells were treated with 1 μg/mL of LPS in the absence or presence of designated concentrations of GPS for 6 h, Data are expressed as mean ± SEM (n = 3). The images shown are representatives of three independent experiments. (B) p-ERK, p-JNK, p-p38 and p-Akt expressions in colon tissues of DSS-treated mice. Colitis was induced in all groups except the control group. GPS and CSA were administered to mice from day 1 to day 7. On day 11, the mice were sacrificed, and protein expression of p-ERK, p-JNK, p-p38 and p-Akt in colon homogenates were determined by Western blotting. Data are expressed as mean ± SEM (n = 6). #p < 0.05 or ##p < 0.01 or ###p < 0.001 vs control group; *p < 0.05 or **p < 0.01 or ***P < 0.001 vs LPS- stimulated or DSS group.

Validation of Anti-Inflammatory Activities of Main Bioactive Components of GPS and Molecular Docking

To further verify the anti-inflammatory activities of the predicted main bioactive components from GPS, their activities were performed on LPS-induced RAW264.7 macrophage cells. As shown in Figure 7A, all of these predicted main bioactive components of GPS exhibited significant inhibitory effects on LPS-induced NO release from LPS-induced RAW264.7 macrophage cells, verifying that these components are anti-inflammatory functional substances of GPS. As the compound-target network revealed that these 10 predicted bioactive components of GPS mainly targeted to the genes MAPK1, MAPK3, MAPK8, MAPK14, and Akt1, we docked these bioactive molecules with MAPK1 and MAPK3 encoding protein ERK, MAPK8 encoding protein JNK, MAPK14 encoding protein p38, and Akt1 encoding protein Akt, respectively. The binding energies were calculated to evaluate the affinity of the components to the protein targets. As shown in Table 1, these predicted bioactive components could bind well to the target proteins Akt, P38, ERK, and JNK, which were better than 5-ASA, a first-line anti-inflammatory drug used to treat IBD in clinics. Among these, luteolin, luteoloside, and picroside II had the lowest binding energies in the molecular docking of all target proteins, respectively. Notably, luteolin, luteoloside, and picroside II showed the lowest binding energies when bound to all target proteins, and their binding energies to JNK, p38, and Akt were comparable to those of JNK, p38, and Akt inhibitors SP600125, SB203580, and AZD5363, respectively. Concurrently, 3D docking diagrams were generated to identify the binding sites of luteolin, luteoloside, and picroside II to all target proteins. As shown in Table S3 and Figure 7B, luteolin, luteoloside, and picroside II interact with target proteins mainly through hydrogen bonding, hydrophobic interaction, and Pi-Pi stacking. Of note, the optimal conformations of luteolin, luteoloside, and picroside II with all target proteins were the “IN” conformations, suggesting that luteolin, luteoloside, and picroside II may occupy the active conformations of those kinases and form a competitive relationship with ATP, which prevents ATP from further phosphorylating those kinases and inhibiting their activities.

Table 1 Docking Results of ERK, JNK, p38 and Akt with References Ligands and 10 Bioactive Components of GPS

Figure 7 Anti-inflammatory effects of main bioactive components of GPS on LPS-induced RAW246.7 cells in vitro. (A) NO release in LPS-stimulated RAW264.7 cells. Data represent the mean ± SD (n = 4). ###p < 0.001 vs control group; **p < 0.01 or ***p < 0.001 vs LPS- stimulated group. (B) The binding modes of ERK, JNK, p38 and Akt with luteolin, luteoloside and picroside II, respectively.

Discussion

Even if multiple drug regimens are already available, incomplete cures and recurrence of IBD remain major problems, making it necessary to develop new safe and effective strategies to treat IBD. THMs are typically preferred as a treatment since they are low in toxicity and their effectiveness has been well tested in long-term use.³⁴ P. scrophulariiflora is a good candidate because this herb was recorded to treat colitis symptoms in TCM, and its sister plant of P. kurroa, an Ayurvedic medicine highly chemically similar to P. scrophulariiflora, has shown an effective therapeutic effect on immunity-related diseases.¹⁰ In this study, we investigated the anti-inflammatory effect of the glycoside fraction of P. scrophulariiflora (GPS) against DSS-induced colitis in mice. Our results showed that oral administration of GPS significantly attenuated the severity of DSS-induced colitis in mice, as indicated by reduced DAI scores, colon shortening, histological damage, and colonic MPO activity. The in vitro assays also showed that GPS significantly suppressed LPS-induced IL-1β, IL-6, and iNOS transcription and the IL-1β, IL-6, and NO production from macrophage RAW264.7 cells.

As is known, inflammation resulting from an abnormal immune system response is one of the most prominent pathological features and pathogenesis of IBD. Macrophages are the main sentinels used to phagocytose and eliminate foreigners that invade colonic tissue. Activated macrophages by pathogens or inflammatory irritants secrete excessive proinflammatory cytokines to initiate the innate immune response and the consequential adaptive immune response that leads to inflammation.^35,36 Both clinical and animal studies have found that accumulated macrophages are observed in the inflamed gut.^24,37 In addition, the removal of macrophages or the suppression of their immune response preventing the development of colitis has been well verified.^38–40 In the present study, our data showed that GPS significantly reduced the clinical signs and pathological manifestations of experimental colitis and suppressed the expression and production of inflammatory cytokines IL-6 and IL-1β in the colon tissues. It is well known that IL-6 and IL-1β are pro-inflammatory cytokines released by activated macrophages. However, our results from LPS-induced RAW264.7 macrophage cells showed that GPS could also significantly suppress NO release, iNOS, IL-6, and IL-1β transcription, as well as IL-6 and IL-1β production induced by LPS in RAW264.7 macrophage cells. Taken together, these data indicate that GPS has a potent anti-inflammatory effect against colonic inflammation, which is closely associated with inhibition of macrophage activation.

Further, we explored the action mechanism of GPS against colitis and the active components of GPS in its anti-inflammatory effects using the network pharmacology method. It is well known that network pharmacology is a new approach based on the theory of system biology to analyze the network of biological systems and select specific signal nodes to elaborate the principle and law of the interaction between organism and drug.^41,42 Currently, network pharmacology is widely used to predict the main active ingredients and potential targets of medicines with multicomponents and multitargets, and it is becoming a cutting-edge research field in TCMs.^43,44 In this study, a protein–protein interaction network was first constructed to systematically summarize the effect targets of anti-inflammatory properties against colonic inflammation, and 77 hub genes were found. Subsequently, the KEGG analysis showed that these hub genes were mainly enriched in PI3K/Akt and MAPK signaling pathways. Previous studies reported that MAPK and PI3K/Akt were two key signaling pathways for macrophage activation, and protein expression of p-ERK, p-p38, p-JNK, and p-Akt was upregulated in the pathological process of colonic inflammation.^45–47 Inhibition of PI3K/Akt or MAPK signal with inhibitor could suppress macrophage activation and alleviate colitis.^48–51 Based on the predicted results of network pharmacology, we experimentally confirmed that GPS could significantly downregulate the protein expression of p-Akt, p-p38, p-ERK, and p-JNK in LPS-induced RAW264.7 macrophage cells and colon tissues of DSS-treated colitis mice. In parallel, a compound-target network showcasing the correlation between bioactive compounds and GPS-related hub genes was also constructed to identify the anti-inflammatory components from GPS, and the top 10 candidate compounds in this network, namely luteolin, apocynin, caffeic acid, caffeic acid methyl ester, luteoloside, picroside II, aucubin, cinnamic acid, vanillic acid, and sweroside, were identified as the main components responsible for the anti-inflammatory effects of GPS. As expected, our NO release assay confirmed that these compounds all exhibited significant inhibitory activities on NO release from LPS-induced RAW264.7 macrophage cells. In vivo, previous studies have also verified that luteolin,⁵² apocynin,⁵³ caffeic acid,⁵⁴ picroside II,⁵⁵ and vanillic acid⁵⁶ have potent anti-inflammatory effects against experimental colitis in multiple animal models. Molecular docking showed that except for the binding conformations of apocynin, cinnamic acid and vanillic acid to ERK, and vanillic acid to JNK, other compounds had good affinities with the crystal structures of target proteins Akt, P38, ERK, and JNK, which were better than or approximately equal to 5-ASA (a first-line anti-inflammatory drug used to treat IBD) to interact with those target proteins. Collectively, these findings clearly demonstrated that GPS could attenuate DSS-induced colonic inflammation by reducing macrophage activation through inhibition of PI3K/Akt and MAPK signaling pathways, and compounds such as luteolin, apocynin, caffeic acid, caffeic acid methyl ester, luteoloside, picroside II, aucubin, cinnamic acid, vanillic acid, and sweroside were the main components responsible for its anti-inflammatory effects.

Conclusion

In conclusion, our study demonstrated that GPS had potent anti-inflammatory effects against DSS-induced colitis in mice, which is attributed to the main components of GPS compounds, such as luteolin, apocynin, caffeic acid, caffeic acid methyl ester, luteoloside, picroside II, aucubin, cinnamic acid, vanillic acid, and sweroside, to reduce macrophage activation through inhibiting PI3K/Akt and MAPKs signaling pathways, suggesting GPS could serve as the potential agent in the treatment of colitis.