Tuftsin-Phosphorylcholine Maintains Normal Gut Microbiota in Collagen Induced Arthritic Mice

Abstract

Rheumatoid arthritis (RA) is characterized by chronic autoinflammation of the joints, with a prevalence of about 1% in Western populations. Evidence in recent years has linked RA to changes in the gut microbiota (dysbiosis). Interestingly, helminths have been shown to have therapeutic activity in RA. Specifically, a glycoprotein containing phosphorylcholine (PC) extracted from helminths was found to have immunomodulatory activity. We have previously developed a novel chimeric compound composed of tuftsin-PC (TPC) that attenuates the joint destruction in mice with collagen-induced arthritis (CIA). Here, we address the interrelationship between TPC immunomodulatory activity and the gut microbiota in CIA mice. Preventive therapy with TPC in mice with arthritis maintained a physiological arthritis score as well as a steady gut microbial environment, similar to that of healthy controls, in contrast to CIA mice with severe disease. The microbial composition differed significantly between healthy and phosphate-buffered saline-treated CIA mice, enabling classifying test samples by machine learning based on levels of a small number of bacterial species. Using these bacterial biomarkers, all TPC-treated CIA mice were classified as healthy. Thus, we describe a clear correlation between TPC treatment, healthy gut microbial communities, and prevention of arthritis. This is the first study to demonstrate the immunomodulatory effect of helminth derivatives in autoimmune diseases and the link to gut microbiota.

Keywords: collagen-induced arthritis; helminths; microbiome; rheumatoid arthritis; tuftsin-phosphorylcholine.

FIGURE 1

TPC lowers the arthritis score and alters gut microbiota in PBS-CIA mice. (A) Arthritis score of healthy (n = 4/group) and CIA mice (n = 7/group) subcutaneously treated with TPC vs. PBS was recorded at 8 time-points throughout the experiment, starting at day 0 (2 days prior to disease induction) and ending at day 34. Values are presented as the mean ± SD. ^∗∗p < 0.01, ^∗∗∗p < 0.001. (B,C) Principal coordinate analysis (PCoA) of fecal microbial communities in CIA and healthy mice treated with TPC or PBS. Community similarity was calculated using the UniFrac distance measure. While no specific clustering by experimental group is evident at day 0 (B), there appear to be 2 clusters at day 34 (C), one representing the microbiota from PBS-treated CIA mice, and the other clustering together the TPC-treated CIA mice, TPC-treated healthy mice, and PBS-treated healthy mice (circled), exhibiting similar bacterial populations.

FIGURE 2

Differentially expressed taxa between healthy and sick (PBS-CIA) mice at day 34. Cladogram indicating the phylogenetic distribution of microbial lineages associated with treatment group. Differences are represented by the color of over-represented bacteria: red indicating the healthy mouse groups (TPC-CIA, PBS-healthy and TPC-healthy mice), green indicating the sick mouse group (PBS-treated CIA mice). Circles represent phylogenetic levels from phylum (outermost circle) to species (innermost circle).

FIGURE 3

Heatmap of OTU abundances in feces of TPC-CIA vs. PBS-CIA and healthy mice. Heatmap constructed by machine learning at day 34 showing the relative proportions of microbial lineages. Darker shades of blue indicate a higher abundance of an OTU on a logarithmic scale. The machine-learning algorithm was trained on the healthy and PBS-CIA groups, and tested on the TPC-CIA group, with all test samples classified as healthy. U, Unclassified. PBS-CIA mice (red, n = 6); TPC-CIA mice (green, n = 3); Healthy - TPC or PBS-treated healthy mice (blue, n = 8). OTUs are presented by annotation and OTU number.