Background/Purpose: A transmissible agent has long been suspected in SLE. In a discovery cohort we found that,compared with healthy subjects, Lupus patients had a five-fold overall mean greater representation of the Ruminococcus gnavus species of the Lachnospiraceae family of anaerobic gram-positive cocci. Many SLE patients also displayed biomarkers of increased gut permeability that has been associated with bacterial translocation. In a patient, the relative fecal R. gnavus abundance directly correlated with serum levels of IgG anti-R. gnavus antibodies. This anti-bacterial immune response had features indicative of molecular mimicry with anti-native DNA antibodies, which are direct contributors to the pathogenesis of Lupus nephritis (LN). We therefore sought to isolate and characterize the responsible strain-associated antigen in the R. gnavus candidate pathobiont.

Methods: A panel of strains of the Lachnospiraceae family, and other anaerobic commensals, were evaluated by immunoblot, bead-based Luminex assay, as well as direct and competition ELISA. To isolate lipoconjugates from the candidate Gram-positive bacterial commensal strain, we applied a validated extraction protocol using a butanol separation with fractionation after passage over Hydrophobic Interaction Chromatography (HIC). Fractions were then evaluated for immunoreactivity and capacity to stimulate a human TLR2- transfected HEK reporter gene system.

Results: In an assay with a cutoff based on the mean+2SD of 55 healthy controls, we found that one of 8 independent R. gnavus isolates, which we termed RG2, displayed high level serum IgG reactivity with 30 of 50 active LN patients and only 4 of 36 patients without active renal disease (P<0.001 by t test). Fractionation of the RG2 extract by chromatography enabled the isolation of a lipoglycan, which by high-resolution NMR spectroscopy showed the presence of repetitive oligosaccharide building blocks. Sera from LN patients also displayed strong IgG reactivity with the lipoglycan, which was inhibitable by the nuclease-treated RG2 extract. In side-by-side analyses the lipoglycan had the same oligomeric immunoblot banding pattern of the parental bacterial extract, and in a HEK transfected reporter cell system displayed strong dose-dependent (range 2 to 2000 ng/ml) NFkB activation, compared to nonstimulated conditions. This activity was significantly inhibited by an anti-huTLR2 mAb at 10 ug/ml (P=0.006) but not isotype control. Further analyses of the exact carbohydrate content, as well as the overall size and detailed structure of the isolated lipoglycan, are currently under study by NMR, MS, and GC/MS.

Conclusion: We have identified a gut commensal strain-associated cell wall lipoglycan, which appears to represent an immunodominant antigen in a candidate pathobiont implicated in the pathogenesis of LN. Moreover, the TLR2-activation potential of the lipoglycan pool may be due to the presence of lipopeptides, which will require further investigation. Our findings suggest a pathway by which continuous translocation of a bacterial mimic of DNA with innate immunostimulatory properties may contribute to the pathogenesis of LN.

« Back to ACR/ARHP Annual Meeting 2018

ACR Meeting Abstracts - https://acrabstracts.org/abstract/identification-of-a-gut-pathobiont-immunostimulatory-lipoglycan-antigen-linked-to-lupus-nephritis/