Gut microbial dysbiosis has been reported in patients with lupus. Results obtained with mouse models suggest that dysbiosis contributes to lupus pathogenesis through the pathobionts that induce inflammation by translocating out of the gut and/or producing proinflammatory metabolites. We and others have shown that fecal microbiota transfers (FMT) from lupus-prone mice induced autoantibodies and immune activation in non-autoimmune mice. On the other hand, we have shown that the production of autoantibodies and associated expansion of follicular T (Tfh) cells and germinal center (GC) B cells can be eliminated by treating lupus-prone mice with 2-deoxy-D-glucose (2DG), a glycolysis inhibitor. Here, we investigated the effect of 2DG on the fecal microbiome in two models of lupus with different etiologies, the (NZB × NZW)F1 and (NZW x BXSB)F1 mice.

Anti-dsDNA IgG-positive (NZB × NZW)F1 and (NZW x BXSB)F1 mice were treated with 2DG. The composition of their fecal microbiome was determined by 16S rDNA sequencing and their metabolome by LC-MS analysis, and compared to age-matched controls. Fecal samples from these mice were used for FMT 3 times per week in pre-autoimmune lupus-prone mice of the same strain that were pre-treated with antibiotics for 2 weeks. FMT lasted for 26 weeks in (NZB × NZW)F1 mice and 9 weeks in (NZW x BXSB)F1 mice. Age-matched controls were gavaged with PBS. Autoantibodies were measured by ELISA and indirect immunofluorescence. Immunophenotypes were assessed by flow cytometry in the spleen and mesenteric lymph nodes. Renal pathology was evaluated by light microscopy on PAS-stained sections and immunofluorescence on frozen sections with antibodies to complement C3, IgG2a, F4/80 and CD3.

We showed that a 2DG treatment started in reduced the changes in bacterial populations that occurred as disease developed in control mice in both models. 2DG also altered the distribution of fecal metabolites in these treated mice. Next, we investigated the effect of serial FMT from 2DG-treated or control mice into pre-autoimmune lupus-prone mice of the same strain that were pre-treated with antibiotics. In both models, FMT from 2DG-treated mice was protective, with a reduction of anti-dsDNA IgG production, immune cell activation, and renal pathology as compared to FMT from control mice.

Overall, our results demonstrated for the first time that the therapeutic effect of glucose inhibition in lupus is transferable through the gut microbiota. These results suggest that the enhanced glucose metabolism in lupus-prone mice promotes the expansion of pathogenic gut bacteria either directly or indirectly through the immune system that normalized by glucose inhibition.

Lay summary High glucose metabolism sustains the activation of the immune system in lupus. Inhibition of glucose metabolism with a drug called 2DG reverses the production of pathogenic autoantibodies in mice. Here we showed that 2DG also changes the gut microbiome in lupus- prone mice. Further, transfers of fecal bacteria from 2DG-treated lupus mice protected younger mice to develop lupus. The results showed that the gut microbiome contributes significantly to the pathogenic effects of glucose metabolism is lupus, and suggest that the beneficial effect of reducing glucose metabolism includes the restoration of a healthy gut microbiome.