Therapeutic Potential of a Novel Bifidobacterium Identified Through Microbiome Profiling of RA Patients With Different RF Levels

Abstract

The potential therapeutic effects of probiotic bacteria in rheumatoid arthritis (RA) remain controversial. Thus, this study aimed to discover potential therapeutic bacteria based on the relationship between the gut microbiome and rheumatoid factor (RF) in RA. Bacterial genomic DNA was extracted from the fecal samples of 93 RA patients and 16 healthy subjects. Microbiota profiling was conducted through 16S rRNA sequencing and bioinformatics analyses. The effects of Bifidobacterium strains on human peripheral blood mononuclear cells and collagen-induced arthritis (CIA) mice were assessed. Significant differences in gut microbiota composition were observed in patients with different RF levels. The relative abundance of Bifidobacterium and Collinsella was lower in RF-high than in RF-low and RF-negative RA patients, while the relative abundance of Clostridium of Ruminococcaceae family was higher in RF-high than in RF-low and RF-negative patients. Among 10 differentially abundant Bifidobacterium, B. longum RAPO exhibited the strongest ability to inhibit IL-17 secretion. Oral administration of B. longum RAPO in CIA mice, obese CIA, and humanized avatar model significantly reduced RA incidence, arthritis score, inflammation, bone damage, cartilage damage, Th17 cells, and inflammatory cytokine secretion. Additionally, B. longum RAPO significantly inhibited Th17 cells and Th17-related genes-IL-17A, IRF4, RORC, IL-21, and IL-23R-in the PBMCs of rheumatoid arthritis patients. Our findings suggest that B. longum RAPO may alleviate RA by inhibiting the production of IL-17 and other proinflammatory mediators. The safety and efficacy of B. longum RAPO in patients with RA and other autoimmune disorders merit further investigation.

Keywords: Bifidobacterium longum; T helper 17 cell; microbiome; rheumatoid arthritis; rheumatoid factor.

Figure 1

Microbial composition at the phylum level in patients with different rheumatoid factor (RF) levels. (A) Comparison of intestinal microflora composition at the phylum level. (B) Relative abundance of Actinobacteria. Comparisons in relative abundance were performed using the Kruskal–Wallis test. (C) Correlation between the relative abundance of Actinobacteria and RF levels. (D) Correlation between the relative abundance of Firmicutes and RF levels. *p < 0.05; **p < 0.01. NM, normal control; NG, rheumatoid arthritis (RA) patients showing 20 or less of RF level; LP, RA patients showing over 20 and not exceeding 60 of RF level; HP, RA patients showing over 60 of RF level.

Figure 2

Microbial composition at the genus level in patients with different RF levels. (A–C) Comparison of the relative abundance of (A) Bifidobacterium, (B) Collinsella, and (C) Clostridium using the Kruskal–Wallis test. (D) Correlations between the relative abundance of each of the 20 most abundant genera and RF levels. Positive correlations are shown in green, and negative correlations are shown in pink. The numbers in bold indicate statistically significant correlations (p < 0.05). *p < 0.05; **p < 0.01; ***p < 0.001. NM, normal control; NG, RA patients showing 20 or less of RF level; LP, RA patients showing over 20 and not exceeding 60 of RF level; HP, RA patients showing over 60 of RF level.

Figure 3

Bifidobacterium longum RAPO suppresses IL-17 expression in human PBMCs. PBMCs were stimulated with anti-CD3 for 72 h. IL-17 levels in the culture supernatant were analyzed using ELISA. Data are presented as the means ± standard deviations (SDs) from three independent experiments. **p < 0.03; ***p < 0.01.

Figure 4

Bifidobacterium longum RAPO alleviates RA in collagen-induced arthritis (CIA) mice. CIA mice (n = 5 per group) were orally administered B. longum RAPO (1 × 10⁸ CFU/mouse) or methotrexate (MTX; 3 mg/kg) once daily for 7 weeks after the immunization boost. (A) Reductions in arthritis score and arthritis incidence in CIA mice treated with B. longum RAPO. Effects of B. longum RAPO on RA development in CIA mice. (B) Tissue specimens were acquired from the hind paw joints of mice and stained with hematoxylin and eosin and safranin O. Representative histological quality and histological grades are shown. (C) Splenocytes isolated at 7 weeks after immunization were stimulated with phorbol myristate, ionomycin, and GolgiStop for 4 h. The percentages of Th1 (CD4⁺IFN-γ⁺), Th2 (CD4⁺IL-4⁺), and Th17 (CD4⁺IL-17⁺) cells were analyzed using flow cytometry, including also Treg (CD4⁺CD25^high Foxp3⁺). (D) Levels of IgA, IgM, IgG2a, and anti-CII-specific IgG2a antibodies in the serum of CIA mice at 7 weeks after the first immunization. Data are presented as the means ± SDs. *p < 0.05; **p < 0.01; ***p < 0.001 (vs. vehicle-treated group).

Figure 5

Effects of Bifidobacterium longum RAPO in obese CIA mice. Mice were administered orally with B. longum RAPO (1 × 10⁸ CFU/mouse) once daily for 7 weeks after the immunization boost. (A) Arthritis score and incidence of B. longum RAPO-treated mice compared with those of obese CIA mice (n = 5 for each group). (B) B. longum RAPO reduces IL-17 expression in CD4 T cells from the spleen of mice with obese CIA. Flow cytometry of Th1 cells (IFN-r⁺CD4⁺), Th2 cells (IL-4⁺CD4⁺), and Th17 cells (CD4⁺IL17⁺) from the spleen of mice with obese CIA. (C) Effect of B. longum RAPO on RA in mice with obese CIA. Tissue from the hind paw joints was stained with hematoxylin and eosin, as well as safranin O. (D) B. longum RAPO inhibits the proinflammatory cytokines IL-1β, IL-6, IL-17, and TNFα in CIA mice. Representative immunohistochemistry images showing that B. longum RAPO alleviates RA in obese CIA mice. Synovium sections treated with a vehicle, B. longum RAPO, or vehicle were stained for IL-1β, IL-6, IL-17, and TNFα. Scale bar, 100 μm. *p < 0.05; **p < 0.01; ***p < 0.001 (vs. vehicle-treated group).

Figure 6

Effects of Bifidobacterium longum RAPO in human PBMCs and avatar mice of RA patient. (A) The PBMCs of the RA patient were cultured with anti-CD3 antibody for 72 h and the resulting Th1 cells (IFN-r⁺CD4⁺), Th2 cells (IL-4⁺CD4⁺), and Th17 cells (CD4⁺IL17⁺) were analyzed. (B) The PBMCs of the RA patient were cultured with anti-CD3 antibody for 48 h. Hierarchical cluster heatmap of the PBMC-stimulated anti-CD3 antibody of RA patient treated with B. longum RAPO or vehicle. The expression of Th17 pathway was analyzed by RNA sequencing. Significantly differentially expressed gene and significant differences in Th17 pathway activities. The fold change of Th17-related genes decreased in the treatment of the B. longum RAPO. (C) Relative mRNA expression of Th17 pathway genes was analyzed by real-time PCR. (D) NSG mice were administered with B. longum RAPO (1 × 10⁸ CFU/mouse) once daily for 7 weeks after the sensitization injection. (E) Splenocytes from the avatar mice of RA patient treated with B. longum RAPO. The cells were stained with Abs against CD4, IL-17. A graph from a representative experiment showing the frequency of IL-17⁺ cells in CD4 T cells. (F) Joint sections from the avatar mice of RA patient with B. longum RAPO-treated mice were stained with safranin O. *p < 0.05; **p < 0.01; ***p < 0.001 (vs. vehicle-treated group).