Lycium barbarum polysaccharides and capsaicin modulate inflammatory cytokines and colonic microbiota in colitis rats induced by dextran sulfate sodium

Abstract

Active ingredients in the natural products have been considered to be used for alleviating the symptoms of ulcerative colitis, hence the effects of Lycium barbarum polysaccharides (LP) and capsaicin on dextran sulfate sodium (DSS)-induced colitis in rats were investigated. Rats were grouped into normal, DSS induced colitis, and colitis treated with 100 mg LP/kg body weight, 12 mg capsaicin/kg body weight, or combined 50 mg LP/kg body weight and 6 mg capsaicin/kg body weight. Treatment with LP or capsaicin was orally fed by gavage for 4 weeks, and 5% DSS was fed via drinking water for 6 days during week 3. Colon tissue and cecum content were collected for analysis. Treatments with LP and/or capsaicin ameliorated disease activity index scores, severity of colon distortion, and shrinkage of colon length. LP and capsaicin decreased colonic pro-inflammatory cytokine (IFN-γ, IL-17A, and IL-22) levels. Cecal microbiota in colitis rats were enriched with the genus Turicibacter and Lachnospira. The relative abundance of genus Ruminiclostridium_9 and Ruminoclostridium_1 was increased by LP and capsaicin treatment, respectively. Pretreatment with LP or capsaicin inhibits the severity of colonic damage in rats with DSS-induced colitis via anti-inflammation and modulation of colonic microbiota.

Introduction

Ulcerative colitis (UC) is characterized as an autoimmune disease with chronic immune-mediated inflammation, impaired tight junctions in the colon epithelial barrier, and dysbiosis of colonic microbiota.‍⁽¹⁾ The UC patients exerted symptoms including loss of body weight (bw), rectal bleeding, and diarrhea. Although UC is not common in Asian countries, but the prevalent rate of UC in Taiwan was gradually increased from 2.1/10,000 people in 2001 to 12.8/10,000 people in 2015.‍⁽²⁾ There is no casual treatment for UC, and patients were commonly given anti-inflammatory drugs and immune suppressors, or others such as antibiotics, anti-diarrhea compounds, and pain control medication.‍^(3,4) However, these drugs were given for symptom control, but not free from adverse effects.

The precise etiology of UC has not been fully understood, but numerous factors have proposed to be participated in the pathophysiology of UC including interactions between environmental factors, hygiene, diet, and colonic microflora.‍^(5–8) The signaling pathway of Janus kinase/signal transducers and activator of transcription (JAK/STAT) is mainly activated by interferon-γ (IFN-γ), and crucial for activating the immune system.‍⁽⁹⁾ JAK/STAT pathway may be one of the key roles in the pathogenesis of UC.‍^(10,11) T helper 17 (Th17) cells are responsible for protecting the host against extracellular pathogens by producing related cytokines such as interleukin-17A (IL-17A) and IL-22,‍⁽¹²⁾ and regulating inflammation via aggregation of Th17 cells and production of its related cytokines in active inflammatory bowel disease (IBD) stage.‍⁽¹³⁾

Lycium barbarum L., a widely used traditional Chinese herb that was commonly known as goji berry, was reported to have active components such as polysaccharides, zeaxanthin, and β-carotene.‍⁽¹⁴⁾ Lycium barbarum polysaccharides (LP) were shown to own properties of immunomodulation, gastroprotection, neuro­protection, anticancer, and regeneration.‍^(15,16) Capsaicin (CP) was identified as the predominant active component extracted from chili peppers, provided spicy flavor, and exhibited immunomodulatory, analgesic, antioxidative, and anti-hyperlipidemic effects.‍^(17,18) Our study focused on longitudinal effects of LP and/or CP via the phases of prevention/induction/remission for colitis model. We hypothesized that LP and/or CP could contribute to beneficial effects on the healing of rat colitis model induced by dextran sulfate sodium (DSS) via inhibiting inflammatory cytokines and modulating colonic microbiota. As both LP and CP potentially showed beneficial to immunomodulation, suggesting that combined LP and CP could have an additive or synergistic effect. Hence, the effects of LP and/or CP on the development of colitis in rats induced by DSS were studied.

Materials and Methods

Materials

LP (L. barbarum polysaccharides M-5000) was bought from Fengyang Biomedical Co., Ltd. (Taichung, Taiwan), and 50% of polysaccharides from LP were extracted by water extraction. CP was obtained from iHerb Inc. (Irvine, CA). The total capsaicin in CP extract was 85% determined by HPLC method. DSS (TdB Labs, Uppsala, Sweden) with a molecular weight of 40 kDa was purchased from BioCommander International Co., Ltd. (Taipei, Taiwan). Reagents for IFN-γ (439007) and IL-17A (437907) measurements were obtained from BioLegend (San Diego, CA). Analysis for IL-22 (M2200) was bought from R&D Systems, Inc. (Minneapolis, MN).

Animals and study design

A total of 50 male Sprague–Dawley rats aged 7 weeks were bought (BioLASCO Taiwan Co., Ltd., Taipei, Taiwan), and housed at the Laboratory Animal Center of Taipei Medical University for a 12-h light/dark cycle. The environment was kept at 22 ± 2°C and relative humidity of 65 ± 5%. Rats were adapted at the Laboratory Animal Center for one week, and given with water and standard chow diet (Laboratory Rodent 5001 chow diet; PMI Nutrition International LLC, Brentwood, MO) ad libitum. Rats were randomly grouped into: normal (N), DSS induced colitis, colitis treated with 100 mg LP/kg bw (L), 12 mg CP/kg bw (C), or a combination of 50 mg LP/kg bw and 6 mg CP/kg bw (M) (n = 10/group). The combination was given a half dose of each treatment to verify if such combination could have any additive or even a synergistic effect. The Institutional Animal Care and Use Committee of Taipei Medical University had approved the animal protocols (LAC-2018-0344).

Treatment of LP and/or CP was given orally via gavage feeding at the assigned dosage described in the previous section from week 1 to week 4 by dissolving into distilled water. Normal or DSS induced colitis rats were fed with an equivalent amount of distilled water by a gavage. To induce colitis symptoms, DSS (5% w/v) was dissolved in distilled drinking water and provided to rats continuously for 6 days during week 3. One week after the end of DSS induction, all rats were sacrificed after anesthetization.

Specimen preparation

Colon tissue was collected, and colon length was measured. For histological analysis, an excised colon tissue (0.5 cm × 0.5 cm) was immediately fixed in 10% formalin solution. For biochemical evaluation, colon tissue was immediately frozen with liquid nitrogen, and further kept in −80°C. For microbiota assessment, cecum content was stored by snap freezing in liquid nitrogen.

The homogenate of colon tissue (100 mg) was obtained by adding 1 ml of phosphate buffered saline and 5-mm stainless steel beads after rapid homogenization at 30 Hz for 1 min. The homogenate was then centrifuged at 13,000 g at 4°C for 10 min, and collected supernatant was stored at −80°C for analysis of inflammatory cytokines.

Evaluations for disease symptoms and histological observation

Disease activity index (DAI) scores were used to evaluate disease symptoms, and conducted from the day before the induction period at the beginning of week 3 to the day before sacrifice at the end of week 4. The components of DAI scores were assessed as follows: changes of bw loss (0: no changes or increase, 1: bw loss 1–5%, 2: bw loss 5–10%, 3: bw loss 11–15%, and 4: bw loss >15%), fecal consistency (0: normal stiffness, 1: semi loose +, 2: semi loose ++, 3: loose, and 4: diarrhea), and fecal hemorrhage (0: no blood, 1: occult blood +, 2: occult blood ++, 3: bloody feces +, and 4: bloody feces ++).‍⁽¹⁹⁾ Body weight loss was defined as the percentage of the difference between the body weight on the measured day and that on the 1st day of week 3.

After 24-h submersion in 10% formalin, the paraffin-embedded colon tissue was dissected after gradient dehydration, and 4-μm serial sections were further stained with hematoxylin and eosin (H&E). All procedures of histological analysis were performed by the National Laboratory Animal Center. The colonic lesion was observed by a pathologist in a blinded manner using a light microscope, and histological scores were determined as follows: abnormalities of mucosal architecture (0–4), inflammation (0–4), ulceration (0–4), and percentage of area involved (0–4).

Measurement of inflammatory markers

Concentrations of inflammatory markers (IFN-γ, IL-17A, and IL-22) in colon tissue were determined using commercial enzyme linked immunosorbent assay kits. All assay procedures were performed strictly according to the instruction provided by the manufactures. The primary antibody for individual inflammatory markers was coated on a commercial plate, and the secondary antibody conjugated with biotin, horseradish peroxidase conjugated with avidin, and the substrates were subsequently added to a plate for reaction. The final absorbance was read at 450 nm.

Microbiota analysis

Next generation sequencing was performed to evaluate cecal microbiota, and analysis of cecum content was done by Biotools Co., Ltd. (New Taipei, Taiwan). The 16S rDNA clone library was used to identify the library of 16S rRNA using an Illumina MiSeq sequencer,‍⁽²⁰⁾ and the information retrieved from SILVA database ver. 132 was used to classify the taxonomy of microbiota.‍^(21,22) The Quantitative Insights Into Microbial Ecology (QIIME ver. 1.9.1) was applied to analyze the sequences of microbiota.‍^(20–22) The USEARCH ver. 7 pipeline was performed to group the operational taxonomic units (OTUs) for identification of sequences with ≥97% distance-based similarity.‍⁽²³⁾ The OTUs abundance to the minimum sequence depth was rarefied using QIIME script (single_rarefaction.py). Species richness and species diversity for α-diversity were estimated using the output from the QIIME pipeline; Chao1 and abundance-based coverage estimator (ACE) indices were used to assess species richness, and Shannon and Simpson indices were evaluated for species diversity.‍^(24,25) Similarity of cecal microbiota for β-diversity was analyzed by principal coordinate analysis (PCoA) on weighted or unweighted unique fraction (UniFrac) distance and non-metric multidimensional scaling (NMDS) using the vegan package in R software (ver. 3.3.1; Microsoft Corp., Redmond, WA), whereas dissimilarity of cecal microbiota for β-diversity was evaluated by Bray-Curtis dissimilarity using the QIIME (ver. 1.9.1) pipeline.‍^(26–28) The differential enrichment of taxa between the groups was determined by linear discriminant analysis (LDA) effect size (LEfSe),‍⁽²⁹⁾ and the threshold of the absolute logarithmic LDA score was set at 3 for discriminative features.

Statistical analysis

Statistical comparisons were analyzed using SPSS 19.0 (IBM Corp., Armonk, NY), and data were expressed as mean ± SEM. One-way analysis of variance (ANOVA) and Tukey’s test were used to compare the differences among continuous variables. R software (ver. 3.3.1) was used for analysis of cecal microbiota. Non-parametric Kruskal–Wallis test and Wilcoxon rank-sum test were applied to compare the relative abundance of cecal microbiota. The differences in all α-diversity indices and PCoA on weighted or unweighted UniFrac distance of β-diversity were measured by one-way ANOVA and Tukey’s test. The differences in NMDS of β-diversity were assessed by analysis of similarities (ANOSIM) and multi-response permutation procedure (MRPP). The differences in Bray–Curtis dissimilarity of β-diversity were determined by permutational multivariate analysis of variance (ADONIS in R-vegan function). The p value <0.05 was considered statistically different.

Results

Gross assessment of colitis

The colon length was significantly shrunk in colitis rats (Table 1, Fig. 1A), and colon weight to length (W/L) ratio was increased compared to that in the normal group (Table 1). Both colon length and W/L ratio were ameliorated in all 3 treatment groups. Rats in the U group exhibited diarrhea and hemorrhage, leading to a sharp increase in DAI scores until the last day of the induction period (day 6) (Fig. 1B). All 3 treatment groups showed a reduction of DAI scores between day 4 and day 12 compared to the U group.

Table 1.

The changes of colon weight, colon length, and weight to length (W/L) ratio by Lycium barbarum polysaccharides and capsaicin in DSS-induced colitis rats

	N	U	L	C	M
Colon weight (g)	2.21 ± 0.15	2.39 ± 0.10	2.19 ± 0.10	2.14 ± 0.06	2.25 ± 0.12
Colon length (cm)	25.0 ± 1.1	19.9 ± 0.6*	23.4 ± 0.6‍#	22.6 ± 0.6‍#	23.6 ± 0.8‍#
Colon W/L (mg/mm)	8.7 ± 0.4	11.8 ± 0.5*	9.4 ± 0.4‍#	9.5 ± 0.3‍#	9.6 ± 0.5‍#

Values are mean ± SEM (n = 10). The differences of colon length, weight, and W/L ratio were compared using one-way analysis of variance and Tukey’s test. N, normal group; U, DSS induced colitis group; L, colitis with 100 mg LP/kg bw group; C, colitis with 12 mg CP/kg bw group; M, colitis with 50 mg LP/kg bw and 6 mg CP/kg bw group. *p<0.05 compared to the N group. ‍^#p<0.05 compared to the U group.

Fig. 1.

Gross assessment of colon tissue and disease symptoms in rats. (A) Representative colon tissue. (B) Disease activity index (DAI) scores. Data are mean ± SEM (n = 10). The differences of the DAI scores were compared using one-way ANOVA and Tukey’s test. N, normal group; U, DSS induced colitis group; L, colitis with 100 mg LP/kg bw group; C, colitis with 12 mg CP/kg bw group; M, colitis with 50 mg LP/kg bw and 6 mg CP/kg bw group. *p<0.05 compared to the N group. ‍^#p<0.05 compared to the U group.

Histological assessment of colitis

Histological sections of colon tissue in the N group were found normal epithelial structure without significant changes (Fig. 2A). However, rats in the U group showed deterioration and crypt damage in the epithelial layer of the colon. In addition, submucosal layer of the colon in the DSS induction group was profound with a massive infiltration of neutrophils (Fig. 2A), leading to a significant increase of histological scores (Fig. 2B). All treated rats showed an amelioration in the epithelial damage of the colon caused by DSS. Epithelial regeneration and lesser infiltration of neutrophils were observed in the treated groups (Fig. 2A), and a significant decrease of histological scores was found in 3 treated groups compared to those in the DSS induction group (Fig. 2B).

Fig. 2.

The histological evaluation of colon tissue in rats. (A) Representative colon tissue stained with H&E. (B) Histological scores in colon tissue. Data are mean ± SEM (n = 5). Histological scores were analyzed using one-way ANOVA and Tukey’s test. N, normal group; U, DSS induced colitis group; L, colitis with 100 mg LP/kg bw group; C, colitis with 12 mg CP/kg bw group; M, colitis with 50 mg LP/kg bw and 6 mg CP/kg bw group. The crypt damage of colonic epithelial layer in Fig. 2A was marked with arrows. Asterisk (*) in Fig. 2A indicates the infiltration of the neutrophils. Dotted-line areas in Fig. 2A represent the regeneration of the epithelial layer in the colon. Scale bars indicate 200 μm. In Fig. 2B, *p<0.05 compared to the N group. ‍^#p<0.05 compared to the U group.

Inflammatory cytokines

The concentrations of IFN-γ, IL-17A, and IL-22 (Fig. 3) in colon tissue were significantly elevated in the DSS-induced colitis rats compared to those in the normal rats (p<0.05). Rats pretreated with LP or CP significantly suppressed the elevation of colonic IFN-γ, IL-17A, and IL-22 levels (p<0.05) while compared to those in the U group. Although combined LP and CP significantly decreased colonic IL-22 levels (p<0.05), treatment with mixture of both did not exhibit similar results in colonic IFN-γ and IL-17A levels.

Fig. 3.

The concentrations of colonic inflammatory cytokines in rats. (A) IFN-γ. (B) IL-17A. (C) IL-22. Values are mean ± SEM (n = 10). The differences of colonic inflammatory cytokines were compared using one-way ANOVA and Tukey’s test. N, normal group; U, DSS induced colitis group; L, colitis with 100 mg LP/kg bw group; C, colitis with 12 mg CP/kg bw group; M, colitis with 50 mg LP/kg bw and 6 mg CP/kg bw group. *p<0.05 compared to the N group. ‍^#p<0.05 compared to the U group.

Cecal microbiota

The relative abundance of cecal microbiota from phylum to genus (Supplemental Fig. 1*) and Firmicutes/Bacteroidetes ratio (Supplemental Fig. 2*) did not show significant differences among 5 groups. The α-diversity indices of species richness Chao 1 (Supplemental Fig. 3A*) and Simpson (Supplemental Fig. 3D*) did not show significant differences among all the groups. However, there were significant differences for ACE (Supplemental Fig. 3B*) and Shannon (Supplemental Fig. 3C*) indices of α-diversity between the M and N groups. Principal coordinate analysis of β-diversity on unweighted UniFrac distance (Supplemental Fig. 4A*) was similar among different groups, but on unweighted UniFrac distance (Supplemental Fig. 4B*) was significantly different between the M and N groups (p<0.05). The β-diversity in the N group revealed a different cluster in NMDS plot (Supplemental Fig. 4C*) using ANOSIM or MRPP and in Bray-Curtis dissimilarity (Supplemental Fig. 4D*) using ADONIS compared to the DSS induction group or 3 treatment groups (p<0.05).

The enrichment of rat cecal microbiota was compared between the normal and colitis (Fig. 4A) or between colitis with LP treatment (Fig. 4B), CP treatment (Fig. 4C), or combination treatment (Fig. 4D) groups using the LEfSe algorithm. Comparing the N and U groups, the genus Alloprevotella and Coprococcus_2 were found to be more enriched with LDA score >4 in the N group, and the genus Turicibacter and Lachnospira were more abundant (absolute logarithmic LDA score >3) in the U group (Fig. 4A). While comparing the U and L groups, the genus Prevotellaceae_NK3B31 was more enriched in the U group (absolute logarithmic LDA score >4), and the genus Ruminoclostridium _9 was more enriched in the L group (Fig. 4B). As comparing the U and C groups, the order Bacteroidales were enriched (absolute logarithmic LDA score >4) in the U group, and the genus Ruminoclostridium_1, Oscillibacter, Marvinbryantia, and Faecalibaculum were enriched (absolute logarithmic LDA score >3) in the C group (Fig. 4C). None of the genus or species in the U or M group reached the discriminative threshold (absolute logarithmic LDA score >3), but the value for the genus Ruminococcus_torques in the M group was close to the discriminative threshold (Fig. 4D).

Fig. 4.

The enrichment of cecal microbiota analyzed by linear discriminant analysis (LDA) effect size (LEfSe) algorithm. (A) Comparisons between normal and dextran sulfate sodium induced colitis groups. (B) Comparisons between the colitis and Lycium barbarum polysaccharides (LP) treated groups. (C) Comparisons between the colitis and capsaicin (CP) treated groups. (D) Comparisons between the colitis and LP + CP treated groups (n = 5). The cladogram (right) reveals phylum to species of the taxonomic ranks from inside to outside. The letters indicate the position on the cladogram, and the abbreviations before the name of microbiota represent the taxonomic rank. The letters from a to h (Fig. 4A) indicate 8, a to d (Fig. 4B) represent 4, a to l (Fig. 4C) indicate 13, and a to f (Fig. 4D) represent 6 different taxa, respectively. The p value <0.05 is considered statistically different when the absolute logarithmic LDA score >3 (left). N, normal group; U, DSS induced colitis group; L, colitis with 100 mg LP/kg bw group; C, colitis with 12 mg CP/kg bw group; M, colitis with 50 mg LP/kg bw and 6 mg CP/kg bw group; c, class; f, family; g, genus; o, order; p, phylum; s, species.

Discussion

We used DSS as a negatively charged polysaccharide compound to induce intestinal erosion and inflammation and increase epithelial permeability in the murine model, which closely resembled UC symptoms in humans.‍⁽³⁰⁾ The DSS-induced colitis model has been applied to screen potential drugs or natural derived substances for the treatment of colitis. In our study, rats successfully developed diarrhea, body weight loss, and bloody stool, which responded to an elevation in DAI scores by providing 5% DSS in drinking water continuously for 6 days. The length of the colon also showed distinctive shrinkage after DSS administration. Pretreatment with LP and/or CP not only decreased the severity of colitis clinical symptoms, but also protected against the damage of colonic crypts and mucosa caused by DSS.

The pathogenesis of IBD has been proposed by numerous studies, and the uncontrolled immune responses mediated by pro-/anti-inflammatory cytokines played a crucial role in the activation of UC symptoms.‍^(31,32) Active IBD patients had increased concentrations of pro-inflammatory cytokines such as IFN-γ, IL-17A, IL-22, IL-1β, and IL-6 in the colon.‍⁽³³⁾ Pro-inflammatory IFN-γ generated by T or NK cells‍⁽³⁴⁾ was an indicator for autoimmune diseases.‍⁽³⁵⁾ A previous study found that IFN-γ deficient mice treated with DSS did not exert significant symptoms of colitis,‍⁽³⁶⁾ suggesting that IFN-γ could be one of the key cytokines to exacerbate mucosal inflammation.

We evaluated colonic IL-17A and IL-22 levels as the representatives of Th17 cells derived cytokines rather than other cytokines because several studies had indicated that IL-17A was correlated with autoimmune diseases and inflammation‍^(13,37–39) Both IL-17A and IL-17F played important roles in host defense, but IL-17F was a weaker inducer of proinflammatory cytokine expression, whereas IL-17A had a stronger potential in inducing inflammation compared to IL-17F,‍⁽³⁷⁾ and IL-17A was positively correlation with disease severity.‍⁽³⁸⁾ Colonic IL-22 protein expression was significantly increased in chronic and initial UC patients compared to the normal controls, and between chronic and initial UC patients, while there were no significant differences for colonic IL-23 expression between chronic and initial UC patients, suggesting that IL-22 can be served as a better biomarker with more sensitivity for determining the severity of UC and initial or chronic period of UC.‍⁽³⁹⁾ However, IL-21 showed no significantly positive correlation with the percentage of Th17 cells in UC patients, but was positively associated with the percentage of follicular T helper cells.‍⁽⁴⁰⁾ Furthermore, the sensitivity of cytokines could be varied at different stages of UC. Rats were killed a week after 6-day colitis induction in our study, and this stage was considered as late acute phase. In most previous studies, mice were sacrificed immediately after 5–7 days of DSS induction,‍^{(41,42,51–53)} or 3 days after a single dose of TNBS (2,4,6-trinitrobenzenesulfonic acid) for inducing colitis symptoms during acute phase of colitis,‍⁽⁷⁾ which could be initial or active phase of colitis. Therefore, colonic IL-17A and IL-22 levels could be more suitable biomarkers for our animal model.

Th17 cells forming IL-17A, IL-22, and its other related cytokines are crucial for host defense against infection with extracellular bacteria and fungi.‍⁽¹³⁾ It was previously reported that the infiltration of Th17 cells and increases of its related cytokines in active IBD patients could be important for the activation of IBD inflammation.‍⁽⁹⁾ Th17 related cytokines such as IL-17A and IL-22 in the colon and serum were positively associated with C-reactive protein levels, DAI scores, and histopathological scores in IBD patients.‍⁽¹³⁾ Our results also revealed that DSS-induced colitis rats significantly elevated colonic IFN-γ, IL-17A, and IL-22 levels. Several studies revealed that the DSS induced colitis murine model increased mRNA or protein expression of the pro-inflammatory cytokines.‍^(41–43) A previous study found that the spleen and mesenteric lymph nodes significantly increased IL-17A producing cells in mice induced colitis by 3% DSS during the late acute phase.‍⁽⁴⁴⁾ However, exacerbation of colitis severity and elevation of pro-inflammatory cytokine mRNA expression were profound in DSS-induced colitis murine models with neutralization of IL-17A using an antibody or knockout of IL-17A, indicating that inhibitory or protective role of IL-17A exhibited in the development of colitis.‍^(45,46) A clinical human study showed that moderate to severe IBD patients using secukinumab (an anti-IL-17 antibody) had more adverse results compared to placebo-controlled IBD patients.‍⁽⁴⁷⁾ In addition, several cases reported that the patients receiving secukinumab for the treatment of dermatological or rheumatoid diseases developed the onset of IBD.‍^(48–50) Therefore, it is reasonable to hypothesize that IL-17A could contribute to a protective factor in the early course of IBD, and an increase in IL-17A levels after DSS administration may act as a damper corresponding to an increase in IFN-γ levels. More studies for IL-17A may be required to determine its role in the pathogenesis of IBD.

The colonic microbiota and the pathogenesis of IBD had a complicated relationship because the changes of microflora were correlated with the alterations of immune profiles after DSS administration in the murine model.‍⁽⁵¹⁾ Our results showed that the taxonomic distribution and Firmicutes/Bacteroidetes ratio in rat cecal microbiota were similar among all the groups, but the M group exhibited significant differences in ACE and Shannon indices of the α-diversity and PCoA on weighted UniFrac distance of the β-diversity compared to the N group. However, the results of NMDS plot and Bray-Curtis dissimilarity revealed that DSS treated rats had different clusters of cecal microbiota compared to the N group though the treatment of LP, CP, or mixture of both did not show any changes compared to the U group. Using the LEfSe method, there were 8 different taxa in the cecal microbiota between the normal and DSS induction groups, and the genus Turicibacter and Lachnospira were more abundant in the DSS induction group. Previous studies‍^(52,53) also showed increased levels of colonic pro-inflammatory cytokines and relative abundance of Turicibacter in mice after DSS administration, indicating that the development of colitis may be highly correlated with the alterations of cecal microbiota and an imbalance level of inflammatory cytokines. The genus Lachnospira, the anaerobic bacteria from the family Lachnospiraceae, characterized as butyrate-producing bacteria‍⁽⁵⁴⁾ also showed to be increased after DSS administration,‍⁽⁵⁵⁾ and such family was found to trigger monocytes or macrophages for shifting into the inflamed colon.‍⁽⁵⁶⁾ Four different taxa were identified between the U and L groups, and the genus Ruminiclostridium_9 was greatly enriched in the L group. The genus Ruminiclostridium_9 was strictly anaerobic and could produce endoxylanase and acetylxylan esterase to degrade polysaccharides and further form short-chain fatty acids which showed anti-inflammatory potential.‍⁽⁵⁷⁾ Thirteen different taxa were identified between the U and C groups, and the CP treated rats had more abundant in the genus Faecalibaculum, Marvinbryantia, Oscillibacter, and Ruminiclostridium_1. The genus Faecalibaculum had anti-inflammatory properties and was decreased in IBD patients.‍⁽⁵⁸⁾ Interestingly, a previous study found that Oscillibacter had a positive correlation with DSS administration in mice,‍⁽⁵⁹⁾ and was enriched after the treatment of CP in our study. Further studies for identifying the relation between colonic microbiota and the activation or remission of IBD symptoms could be required due to a relatively complicated and dynamic ecosystem of colonic microbiota.

We did not assess the microbiota profile of LP and CP before DSS induction because the aims of this study were to investigate the protective effects of LP and/or CP against colitis and the mechanism of LP and/or CP on ameliorating colitis via modulation of colonic microbiota. The previous studies have demonstrated the effects of Lycium barbarum polysaccharides or capsaicin on modulating colonic microbiota. The mice study showed LP increased the abundance of Proteobacteria and Firmicutes, and potentially elevated probiotic genera such as Akkermansia, Lactobacillus, and Prevotellaceae.‍⁽⁶⁰⁾ Additionally, CP enhanced Faecalibacterium prausnitzii content and improved Firmicutes/Bacteroidetes ratio in the mice study.‍⁽⁶¹⁾ Although we did not found these alterations in the abundance of colonic microbiota in rats by LP or CP after DSS-induced colitis, colonic microbiota is a complex ecosystem and correlated with many factors. Therefore, further investigation is needed.

In conclusion, pretreatment with LP and/or CP can attenuate disease symptoms in DSS-induced colitis rats. Treatment with 100 mg LP/kg bw continuously for 4 weeks decreases colonic levels of IFN-γ, IL-17A, and IL-22, but increases the relative abundance of genus Ruminiclostridium_9. Oral administration with 12 mg CP/kg bw continuously for 4 weeks attenuates colonic IFN-γ, IL-17A, IL-22, and enriches the relative abundance of genus Ruminoclostridium_1, Oscillibacter, Marvinbryantia, and Faecalibaculum. No synergistic effects on the improvement of colitis are found in the combined treatment of 50 mg LP/kg bw and 6 mg CP/kg bw in terms of disease symptoms, immune responses, and colonic microbiota.

Author Contributions

The design of this study was conceptualized by YZL and JC-JC. YZL conducted this study and biochemical analyses. All authors provided critical inputs to data analyses and the inter­pretation of the data. YZL and JC-JC wrote the manuscript draft. JC-JC edited and revised the manuscript. All authors read and approved the final version for submission.

Abbreviations

ACE

abundance-based coverage estimator

CP

capsaicin

DAI

disease activity index

DSS

dextran sulfate sodium

IBD

inflammatory bowel disease

IFN-γ

interferon-γ

IL

interleukin

LDA

linear discriminant analysis

LEfSe

linear discriminant analysis effect size

LP

Lycium barbarum polysaccharides

NMDS

non-metric multidimensional scaling

PCoA

principal coordinate analysis

Th17

T helper 17

UC

ulcerative colitis

UniFrac

unique fraction

Conflict of Interest

No potential conflicts of interest were disclosed.