Lipidome Modulation by Dietary Omega-3 Polyunsaturated Fatty Acid Supplementation or Selective Soluble Epoxide Hydrolase Inhibition Suppresses Rough LPS-accelerated Glomerulonephritis in Lupus-prone Mice

Introduction: Lipopolysaccharide (LPS)-accelerated autoimmune glomerulonephritis (GN) in NZBWF1 mice is a preclinical model potentially applicable for investigating lipidome-modulating interventions against lupus. LPS can be expressed as one of two chemotypes: smooth LPS (S-LPS) or rough LPS (R-LPS) which is devoid of O-antigen polysaccharide sidechain. Since these chemotypes differentially affect toll-like receptor 4 (TLR4)-mediated immune cell responses, these differences may influence GN induction.

Methods: We initially compared the effects of subchronic intraperitoneal (i.p.) injection for 5 wk with 1) S-LPS, 2) R-LPS, or 3) saline vehicle (VEH) (Study 1) in female NZBWF1 mice. Based on the efficacy of R-LPS in inducing GN, we next used it to compare the impact of two lipidome-modulating interventions, ω-3 polyunsaturated fatty acid (PUFA) supplementation and soluble epoxide hydrolase (sEH) inhibition, on GN (Study 2). Specifically, effects of consuming ω-3 docosahexaenoic acid (DHA) (10 g/kg diet) and/or the sEH inhibitor 1-(4-trifluoro-methoxy-phenyl)-3-(1-propionylpiperidin-4-yl) urea (TPPU) (22.5 mg/kg diet ≈ 3 mg/kg/day) on R-LPS triggering were compared.

Results: In Study 1, R-LPS induced robust elevations in blood urea nitrogen, proteinuria, and hematuria that were not evident in VEH- or S-LPS-treated mice. R-LPS-treated mice further exhibited kidney histopathology including robust hypertrophy, hyperplasia, thickened membranes, lymphocytic accumulation containing B and T cells, and glomerular IgG deposition consistent with GN that was not evident in VEH- or SLPS-treated groups. R-LPS but not S-LPS induced spleen enlargement with lymphoid hyperplasia and inflammatory cell recruitment in the liver. In Study 2, resultant blood fatty acid profiles and epoxy fatty acid concentrations reflected the anticipated DHA- and TPPU-mediated lipidome changes, respectively. The relative rank order of R-LPS-induced GN severity among groups fed experimental diets based on proteinuria, hematuria, histopathologic scoring, and glomerular IgG deposition was: VEH/CON< R-LPS/DHA ≈ R-LPS/TPPU<<< R-LPS/TPPU+DHA ≈ R-LPS/CON. In contrast, these interventions had modest-to- negligible effects on R-LPS-induced splenomegaly, plasma antibody responses, liver inflammation, and inflammation-associated kidney gene expression.

Discussion: We show for the first time that absence of O-antigenic polysaccharide in R-LPS is critical to accelerated GN in lupus-prone mice. Furthermore, intervention by lipidome modulation through DHA feeding or sEH inhibition suppressed R-LPS-induced GN; however, these ameliorative effects were greatly diminished upon combining the treatments.