Through analysis of fecal samples, intestinal dysbiosis has been observed in rheumatoid arthritis (RA) patients (1,2). However, comparison studies of intestinal bacterial populations from RA patients and healthy controls are not performed until after arthritis is clinically diagnosed. These analyses provide little insight on microbial changes that occur early in the disease process which may have a significant effect on RA pathogenesis (3). Further investigation is needed to determine whether intestinal dysbiosis has a pathogenic role in RA or if it is a result of RA development. Without the ability to reliably diagnose inflammatory arthritis in pre-clinical stages, animal models are a vital tool for determining the role of intestinal dysbiosis in RA.
Intestinal Bacteria Modulates Inflammatory Arthritis: Methods
A paper published in March 2018 by researchers at the University of Colorado used the Collagen-Induced Arthritis (CIA) and Collagen Antibody-Induced Arthritis (CAIA) models to investigate how gut microbiota modulates inflammatory arthritis. To analyze intestinal bacterial populations during arthritis development, CIA was induced in DBA/1J mice following Chondrex, Inc.’s recommended CIA protocol and fecal samples were collected every 7 days, from day 0 through the end of the study (day 35). Bacterial populations in fecal samples were analyzed using 16S rRNA sequencing. Intestinal tissue homogenates were assayed for Th17 cytokine levels (IL-17A, IL-22, IL-23).
The researchers then went on to use broad-spectrum antibiotic treatments to analyze how microbial depletion affects CIA pathogenesis. CIA mice were given either:
Fecal samples from each group were analyzed to determine intestinal bacterial composition and intestinal tissue homogenates were assayed for Th17 associated cytokines.
Traditional CIA Experiment Results: Changes in Mucosal Bacteria and Th17 Cytokines
Significant changes in mucosal bacteria accompanied the development of CIA, with arthritis being clinically apparent on days 23-24. Bacterial populations in fecal samples from days 14 and 21 differed significantly from fecal samples collected on day 0. Notably, Lactobacillaceae was “significantly expanded” and the Bacteroidales S24-7 family was “significantly reduced” on days 14 and 21. Fecal samples between day 7 and 14 were also significantly different from each other, indicating substantial changes early in CIA development. However, fecal samples from day 35 were not significantly different than previous time points, except for an increase in Lachnospiraceae. This data was compared with a control group of mice which received immunizations of only complete Freund’s adjuvant (CFA). Interestingly, the CFA control group mice also had significant changes in bacterial populations, the pattern was different than that found in CIA mice.
Using 4 kDa FITC-Dextran, the researchers investigated the affect of dysbiosis on intestinal permeability. CIA mice had increased intestinal permeability on days 14 and 35, but CFA immunized control mice did not. This suggests that increased intestinal permeability may be involved in CIA development. Furthermore, IL-17A and IL-22 (associated with Th17 cell activation) were found to be significantly increased in intestinal tissue homogenates of CIA mice on day 14, before the onset of arthritis. CFA immunized mice did not exhibit increased levels of Th17 related cytokines, indicating that Th17 cells in the mucosal environment may play a critical role in the immune response to immunized collagen. Th17 related cytokines on day 35 returned to the level seen on day 0, mirroring the regression of S24-7 and Lactobacillaceae populations.
Microbial Depletion Affects CIA Development
To elucidate the time-dependent role of microbiota in RA development, the researchers used broad spectrum antibiotics to deplete the intestinal microbiota at various points during the CIA protocol and observed how this affected CIA development. The early antibiotic treatment groups exhibited a 40% decrease in arthritis severity compared to CIA controls. The incidence of arthritis was 100%. Surprisingly, the late antibiotic treatment group saw a >95% reduction in arthritis severity when compared to controls with around a 50% incidence.
In the early antibiotic treatment groups, IL-17A and IL-22 at day 14 were significantly lower when compared to controls. However, by day 35, the Th17 related cytokine levels were significantly increased compared to CIA controls. On the other hand, the late antibiotic treatment group had significantly reduced Th17 related cytokine levels on day 35 compared to CIA controls. To ensure that antibiotic treatment did not directly have an anti-inflammatory effect, the Arthrogen-CIA 5-Clone Cocktail was used with antibiotic treatment. Arthritis severity was unaffected in CAIA mice with antibiotic treatments, indicating that changes in microbial bacterial compositions caused by antibiotic treatment, rather than the antibiotics, modulated the intestinal immune environment and affected Th17 related cytokine levels.
In addition to Th17 related cytokines, the researchers measured serum anti-type II collagen IgG antibody levels in each antibiotic treatment group. All antibiotic treatment groups had a mild reduction in anti-type II collagen IgG antibodies (anti-CII antibodies), especially IgG2b antibodies. However, the anti-CII antibodies from the late treatment group were found to have a reduced capacity to activate C3 (an important mediator of the complement cascade) as compared to the early antibiotic treatment group or control group. Given that reduction in IgG antibodies could not explain the loss of complement activation, the researchers investigated the glycosylation profiles of the autoantibodies. Analysis of the glycosylation profiles (important for activating C3) of the anti-CII IgG antibodies in each group revealed that antibodies from the late treatment group had a “distinct glycosylation pattern” that was lower in galactose than other groups. The data indicates that the changes in intestinal bacterial, induced through antibiotic treatment, affects the glycosylation of anti-CII antibodies. Given that antibody glycosylation has been hypothesized to be an important factor in complement activation, this provides a possible mechanism for the reduced arthritis in the late antibiotic treatment group.
This study, published in Arthritis & Rheumatology, provides a look into how intestinal microbiota changes over the course of Collagen-Induced Arthritis. Most notably, this study showed that:
Taken together, this study supports a hypothesis implicating intestinal microbiota in the formation of autoimmune responses. The results clearly show that microbial depletion can modulate CIA by altering Th17 related cytokine expression and affecting the ability of antibodies to activate the complement cascade. Future studies using germ free mice could be used discern the involvement of specific bacterial species in the formation of autoimmune responses.