Introduction

Inflammatory bowel diseases (IBD) are inflammatory disorders of the gastrointestinal tract that affect millions of individuals worldwide, at increasingly higher rates1. The two major disorders of IBD, Crohn’s disease (CD) and ulcerative colitis (UC), are characterized by both acute and chronic inflammation of the intestine with multifactorial etiology. Although the etiology of IBD remains largely unknown, many studies have shown that its pathogenesis involves a complex interaction between genetic, environmental/microbial factors, and immune responses1. Drugs for IBD, including aminosalicylates, corticosteroids and anti-tumor necrosis factor (TNF) therapies target different aspects of the immune response, but all are hampered by primary and secondary loss of response and have a number of contraindications and adverse effects2,3,4. It is therefore imperative that novel IBD drugs are developed using different strategies. Because innate immune cells such as macrophages and dendritic cells initiate the immune response against microbes and IBD is characterized by an aberrant innate-immune response5,6, this study focused on develo** drug to regulate the most crucial signaling pathways triggered by microbes in innate immune cells.

Traditional Chinese medicine (TCM) has long been used in Asian countries to treat various inflammatory and autoimmune diseases7. In Taiwan, 70% of patients with psoriasis8, 27% of patients with rheumatoid arthritis9, and 37% of patients with IBD10 were TCM users. Of all TCMs, Mu Dan Pi (MDP) is one of the most commonly prescribed single herbs for the treatment of psoriasis and other inflammatory symptoms8. MDP, also known as Moutan Cortex Radicis, is a TCM derived from the root bark of Paeonia suffruticosa Andrews (Genus: Paeonia; Fam: Paeoniaceae)11. Its bioactive components exhibit anticancer, antimicrobial and antioxidant properties12,13,14,15,20, cancer27 and viral infections28. It is known that PGG prevents the phosphorylation of cytosolic IκB proteins and thus alleviates the TLR-NF-κB signaling pathway29,30. By dissecting the TLR-NF-κB and RIG-I/MDA5-IRF3 pathways, we found that MDP plays important roles not only in preventing NF-κB activation, but also in regulating the IRF pathway. The inhibitory effects of MDP and PGG on these two pathways are not equivalent. Compared to the limited activities of PGG in both pathways, MDP more efficiently inhibits IRF reporter activity compared with NF-κB activity. Since IRF is important for type I IFN induction, MDP may be useful in type I IFN-dominant diseases, such as viral infections.

It has been reported that many compounds identified from MDP have anti-inflammatory activities. However, in our screening system, many of them displayed very little or no effect on both NF-κB and IRF reporters, or exhibited inhibitory qualities only at very high concentrations, such as synthetic PGG, which had an IC50 of ~ 10 μM (Fig. 3). In our previous publication, the PGG concentration in WE was measured as 975.90 ± 14.10 ppb (1.04 ± 0.02 μM)13. The amount of PGG in the working concentration of WE was ~ 25 nM and much lower than the IC50 of synthetic PGG compound, suggesting that PGG may not be the major contributor in MDP anti-inflammatory activity. Because MDP powder and WE were obtained from different preparations, it is difficult to compare the efficacy of MDP powder with that of WE in the DSS-induced colitis model (Fig. 7). However, both TCM preparations derived from MDP demonstrated therapeutic effects in colitis, suggesting that both have therapeutic activities. In reference to our measurement of concentrations of PGG in WE13, the amount of PGG in WE that we gave to mice was 1.18 mg/kg (Fig. 7), which is much lower than the amount of synthetic PGG compound (20 mg/kg) in the colitis model (Fig. S1). Evidence from both cellular (Figs. 3 and 6) and animal experiments (Fig. 7) clearly reveal that PGG is not the key compound that independently reduces inflammatory responses. In contrast to PGG, the strong inhibitory activity of WE, which contains relatively low concentrations of individual compounds, has therapeutic effects. This effect may either be the result of a synergistic effect of multiple compounds, or is due to an as-yet unidentified compound. To fully understand the immunomodulating mechanisms underlying MDP, our future studies will seek to identify as-yet unknown components in MDP that can inhibit these two pathways and determine the maximum synergistic effects by reconstituting and combining individual compounds.

In this study, WE significantly inhibited NF-κB and IRF reporters and alleviated downstream IL-6 and TNF-α production in a dose-dependent manner, without inducing cell death. Among the five known active components identified from MDP, PGG was the most efficient in this study in inhibiting NF-κB and IRF reporters (with half-maximal inhibitory concentrations [IC50] of 5–10 µM in each condition) and in downregulating production of the proinflammatory cytokines. Methyl gallate inhibited only the IRF promoter at a very high concentration (100 µM). At 100 μM, all other compounds, including gallic acid, paeonol, and paeoniflorin failed to inhibit either the NF-κB or the IRF reporter. MDP powder and WE, but not PGG, reduced colon inflammation in the DSS-induced colitis model. All of our study findings suggest that MDP powder and its water extract are highly effective at resolving immune cell activation and infiltration in the gut, and thereby reducing colitis severity. Thus, MDP and its water extract offer promise as a novel therapeutic strategy for treating inflammatory diseases and IBD.

Materials and methods

Antibodies and reagents

Antibodies used in this study were F4/80 (Cat. MCA497G, Lot 1806, Bio-Rad, Hercules, CA, USA; 1:200) and glyceraldehyde 3-phosphate dehydrogenase (Cat. 100118, Lot 42928, GeneTex; 1:5,000). All other antibodies came from Cell Signaling (1:2000), including p-TBK1 (Cat. 5483, Lot 11), TBK1 (Cat. 3504, Lot 4), p-IRF3 (Cat. 29047, Lot 4), IRF3 (Cat. 4302, Lot 4), p-IKKε (Cat. 8766, Lot 3), p-IKKα/β (Cat. 2697, Lot 16), IKKα (Cat. 27609), p-IκBα (Cat. 2859, Lot 18), IκBα (Cat. 4814, Lot 17). LPS, poly(I:C) and Pam3CSK4 were obtained from Invivogen. PGG, methyl gallate, gallic acid, paeoniflorin, and paeonol were purchased from Sigma (Sigma-Aldrich).

MDP powder and preparation of WE

MDP powder was purchased as a commercialized powder form for clinical use from Sun Ten Pharmaceutical Company, Ltd. (Taipei, Taiwan). WE was prepared by the same company following our requested procedures as follows: 100 g P. suffruticosa was boiled with 1.5 L water at 100 °C for 30 min, then concentrated to 100 ml under reduced pressure. A clear supernatant (WE) was further processed by centrifugation at 14,000×g for 20 min, which contained the weight of a total of 250 mg residue per ml after dehydration in vacuo. WE components were examined by liquid chromatography tandem-mass spectrometry (LC–MS/MS), using previously described methods13.

Cells and culture conditions

THP1-Dual cells were purchased from InvivoGen and cultured according to the manufacturer’s manual. BMDMs were isolated from mice and differentiated for 5–6 days, following standard procedures31. Adherent BMDMs were cultured overnight at a density of 1 × 106 cells per ml before being treated with 5, 2.5, 0.5, or 0.25% WE (equivalent to 12.5, 6.25, 1.25, or 0.625 mg MDP residue per ml, respectively) or with the indicated chemicals for 6 h. The cells were then stimulated with TLR ligands at the indicated concentrations32.

Reporter assays

THP1-Dual cells were pretreated with WE or indicated chemicals for 6 h, then stimulated with TLR2 ligand (0.1 μg/ml Pam3CSK4) or RIG-I/MDA5 ligand (0.05 μg/ml poly(I:C)/LyoVec), following the manual’s procedures. After undergoing stimulation for 24 h, NF-κB and IRF induction was quantified by measuring the levels of secreted alkaline phosphatase (SEAP) and luciferase activity in the culture supernatant using QUANTI-Blue (InvivoGen) and QUANTI-Luc (InvivoGen), respectively. Changes from baseline were calculated as -fold changes, normalized against unstimulated control cells.

Viability assays

The viability of THP1-Dual cells was analyzed using the MTS assay. THP1-Dual cells were pretreated with water extract or chemicals before RIG-I/MDA5 ligand stimulation. After stimulation for the indicated times, the cells were washed twice with fresh medium. Cell viability was analyzed using an MTS assay kit (Abcam). After 2 h of incubation, cell viability was determined by the amount of 490 nm absorbance. The percentage of cell death was calculated by normalization with cells that were either untreated or treated with the ligand only.

DSS-induced colitis model

The same aged-, sex- and weight-matched C57BL/6N mice were kept in a pathogen-free environment and were aged 8–12 weeks at the time of the experiments. DSS-induced colitis was induced by the administration of 3% DSS (36,000–50,000 molecular weight; MP Biomedicals, Santa Ana, CA, USA) in the drinking water. DSS drinking water was made available to the mice for 5 days, followed by 7 days of normal water, as previously reported33. In the drug treatment experiment, mice were randomly assigned to one of two groups; the control (vehicle) group and the treatment group. In the treatment group, mice were fed daily with MDP (20 mg/kg), MDP water extract (312.5 mg MDP residue/kg), or PGG (20 mg/kg), starting at the same time as DSS treatment. MDP was purchased from Sun Ten Pharmaceutical Company and its clinical dose suggested by the company was 20 mg/kg. Body weights and stool samples were monitored daily starting from day 0 of DSS treatment, as previously described34. The baseline clinical score was determined at the moment DSS treatment was started (day 0). Scoring for stool consistency and occult blood was done as described previously34. Briefly, stool scores were determined as follows: 0 = well-formed pellets, 1 = semi-formed stools that did not adhere to the anus, 2 = semi-formed stools that adhered to the anus, 3 = liquid stools that adhered to the anus. Bleeding scores were determined as follows: 0 = no blood by using Hemoccult SENSA (Beckman Coulter), 1 = positive Hemoccult SENSA, 2 = visible blood traces in stool, 3 = rectal bleeding. The stool and bleeding scores were averaged to calculate the clinical score. On day 12, the colon was removed, measured, photographed, and fixed in 10% formalin. Intestinal inflammation was assessed histologically using previously described methods33. The histological scores were determined as follows: 0 = presence of occasional inflammatory cells in the lamina propria, 1 = increased numbers of inflammatory cells in the lamina propria, 2 = confluence of inflammatory cells extending into the submucosa, 3 = transmural extension of the infiltrated cells. Tissue damage was scored as follows: 0 = no mucosal damage, 1 = lymphoepithelial lesions, 2 = surface mucosal erosion or focal ulceration, 3 = extensive mucosal damage and extensive deeper structural damage. The combined histological score ranged from 0 (no changes) to 6 (extensive infiltrated cells and tissue damage). All animal care and experimental methods were approved by the Institutional Animal Care and Use Committee of China Medical University, Taichung, Taiwan and were performed in line with the Guideline for the Care and Use of Laboratory Animals and approved by Taiwan’s government organization, Council of Agriculture, Executive Yuan.

Enzyme-linked immunosorbent assay (ELISA)

Cytokine kits for detecting mouse IL-6 and TNF-α were obtained from eBioscience. Cytokine measurements were performed according to their respective manuals provided by eBioscience.

Statistics

Results are presented as the mean ± S.E.M. in mouse model experiments and as the mean ± S.D. in experiments using primary macrophages or THP-1 cells. Data were analyzed using the Student’s t‐test (two-tailed) or one-way ANOVA with the Tukey's multiple comparison test. A p-value of < 0.05 was considered significant.