Microbiota-Derived Inosine Suppresses Systemic Autoimmunity via Restriction of B Cell Differentiation and Migration

Abstract

The role of gut microbiota dysbiosis in systemic lupus erythematosus (SLE) pathogenesis remains elusive. Here, it is shown that fecal microbiota transplantation (FMT) from healthy mice to lupus mice ameliorates lupus-like symptoms. Microbiota reconstitution effectively reduces systemic class switch recombination (CSR) and elevates immunoglobulin heavy chain (IGH) naïve isotype. Microbiota profiling reveals an enrichment of Lactobacillus johnsonii post-FMT, with a significant correlation to purine metabolites. Importantly, the L. johnsonii-derived inosine, an intermediate metabolite in purine metabolism, effectively alleviates lupus pathogenesis in mice. Inosine inhibits B cell differentiation and reduces renal B cell infiltration to protect mice from lupus. At the molecular level, inosine reprograms B cells through the extracellular signal-regulated kinase (ERK)-hypoxia-inducible factor-1alpha (HIF-1α) signaling pathway. Therefore, this study highlights the discovery of a novel microbial metabolite modulating autoimmunity and suggests its potential for innovative microbiome-based therapeutic approaches.

Keywords: Lactobacillus johnsonii; B cell differentiation; B cell migration; fecal microbiota transplantation (FMT); inosine; lupus nephritis (LN); purine metabolites; systemic lupus erythematosus (SLE).

Figure 1

Transplantation of fecal microbiota from MRL/mpj mice attenuated lupus‐like phenotypes in pristane‐induced mice. A) Overview of FMT experimental design (n = 10 per group). 8 weeks age C57BL/6j mice were intraperitoneally injected 500 uL pristane, and lupus‐like symptoms started detecting at 18 weeks of age. FMT (0.1 g mL⁻¹, 0.2 mL⁻¹/time, 3 times/week) or saline treatment started at 22 weeks of age and lasted until 45 weeks of age. B,C) Urine protein was detected by B) test strip and C) CBB method. D) The ratio of urine protein to creatinine was detected. E) Plasma anti‐dsDNA level was detected between saline‐treated and FMT‐treated groups. F) Representative flow cytometry diagrams and statistical analyses were conducted to compare the frequencies of CD4 naïve T cells, CD4 TEM cells, CD8 naïve T cells, and CD8 TEM cells in the spleen between the saline‐treated and FMT‐treated groups. G) Representative flow cytometry diagrams and statistical analysis of frequencies of naive B cells and memory B cells in the spleen between saline‐treated and FMT‐treated groups. Renal histopathology was detected. Scale bar, 200 µm. H) Representative flow cytometry diagrams and statistical analysis of frequencies of plasma cells in the spleen between saline‐treated and FMT‐treated groups. I) IgG, IgG2a, and C3 deposition in the kidney were detected. Scale bar, 200µm. J) H&E staining and Masson staining of kidney, and renal histopathology score and interstitial fibrosis score were evaluated. Scale bar, 200 µm. The results are expressed as mean ± SEM. Statistical comparison was based on one‐way ANOVA. *p < 0.05 was considered statistically significant; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 2

FMT reduced CSR and increased IGH naïve isotype in lupus mice. A) Tree maps of the TCR IR of group saline and FMT. Each uCDR3 clone is treated as a pane with different colors, and the size of the panes is based on expression. Panes were gathered from top to bottom as TRA, TRB, TRD, and TRG. B) The Pielou evenness of TRA, TRB, TRD, and TRG chains. C) Tree maps of the BCR IR of group saline and FMT. Panes were gathered from top to bottom and from left to right as IGH, IGK, and IGL. D) The Pielou evenness of IGH, IGK, and IGL chains. E) Networks of IGHs from group saline and FMT. Each node represents a uCDR3 sequence based on expression, and the sequences that underwent CSR or SHM are linked by sticks. F) Expression percentage calculated by uCDR3s in each IGH chain isotype from group saline and FMT. G) IGH class switched isotype percent by uCDR3s. H) The CSR index of IGHA and IGHG2 with IGHM/IGHD (IGHM and IGHD are shown in the graph as a whole) in the two groups. I) IGH naïve mutated isotype percentage and IGH naïve unmutated isotype percentage by uCDR3s. J) The plasma levels of total IgG1, IgG2a, IgG2b, IgG3, IgA, and IgM between saline‐treated and FMT‐treated group. The results are expressed as mean ± SEM. Statistical comparison was based on an unpaired Student t‐test. *p < 0.05 was considered statistically significant; **p < 0.01.

Figure 3

Gut microbiota alteration reduced immune cell infiltration in mice kidneys. A) scRNA‐seq t‐distributed stochastic neighbor embedding (t‐SNE) plot exhibiting renal cells from FMT‐treated pristane‐induced mice (FMT, n = 3) and saline‐treated controls (Saline, n = 2). Cells colored by cell types. Macrophages and dendritic cells: Mac/DC, distal convoluted tubule: DCT, memory T cell: Tm, intercalated cell: IC, vascular endothelial cell: vEC. B) Dot plot shows signature genes for each cell type. The dot color represents the scaled amount of expression for each gene. The size of the dot represents the percentage of cells expressing the gene. C) Bar plot showing the proportion of cells from different groups in each cell type. D) Umap plot exhibiting individual renal B cells colored by sub‐clusters. E) A bar plot shows the proportion of cells from different groups in each sub‐cluster of B cell. F) Violin plot showing the expression level of signature genes for each sub‐cluster of B cells. Colors represent different genes. G) Dot plot visualizes the GO biological process terms significantly enriched in each sub‐cluster of B cells from saline and FMT. H) Dot plot showing genes of interest for each cell type. The dot color represents the scaled amount of expression for each gene. The size of the dot represents the percentage of cells expressing the gene. I) The mRNA level of cd38, ccr7, s1pr1, and ebi2 in kidney. The results are expressed as mean ± SEM. Statistical comparison was based on one‐way ANOVA. *p < 0.05 was considered statistically significant; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 4

FMT modulated gut microbiome in pristane‐induced mice. Metagenome sequencing analysis of DNA from fecal pellets in saline‐treated and FMT‐treated pristane‐induced mice (n = 10 per group) in 44 weeks. A) Beta‐diversity using principal coordinate analysis of Bray‐Curtis was performed between groups. B) Relative abundance of the top 10 phyla in saline‐treated and FMT‐treated groups. C) Relative abundance of the top 10 genera in saline‐treated and FMT‐treated groups. D) The cladogram generated using LEfSe shows statistically significant differences between saline‐ and FMT‐treated groups. E) A Histogram of the LDA scores generated with LEfSe is shown on a logarithmic scale. The most differential taxa are displayed in the FMT group in green and the saline group in red. F) Plasma FITC‐dextran level in 45 weeks of age among healthy control (HC), saline‐treated and FMT‐treated groups. G) Cultures of translocated bacteria in liver, kidney, and MLN among HC, saline‐treated and FMT‐treated groups. H) Immunofluorescence of tight junction Occludin and ZO‐1 were detected in the small intestines. The results are expressed as mean ± SEM. Statistical comparison was based on one‐way ANOVA in (F). *p < 0.05 was considered statistically significant; **p < 0.01; ***p < 0.001.

Figure 5

L.johnsonii‐associated purine metabolite enrichment alleviated systemic autoimmunity. An untargeted metabolomics of the feces in saline‐treated and FMT treated pristane‐induced mice in 44 weeks was conducted (n = 10 per group). A) Enriched KEGG pathways in fecal metabolites between saline‐treated and FMT‐treated mice were shown with the P values and the number of metabolites represented in each pathway. The size of each bubble represents the number of metabolites differentially expressed for each pathway, with a scale on the lower right (number). B) The matchstick map according to the differential metabolites obtained from each group of difference comparison combinations in positive mode was drawn, and the up‐down of metabolites and the substances with large difference multiples were indicated. C) Pearson's correlation analysis between metabolites in purine metabolism and significantly differentiated strains was conducted. D) Plasma inosine level was tested between Saline‐treated and FMT‐treated groups. E) Schematic diagram of GF mice orally treated saline and L. johnsonii (Lj, 2 OD, 0.3 mL⁻¹/time, 6 times in total). Plasma inosine level was tested between saline‐treated and Lj‐treated groups. F) Overview of Lj and inosine treatment experimental design (n = 6 per group) in pristane‐induced C57 mice. Orally gavage Lj (2 OD, 0.3 mL⁻¹/time, 3 times/week), inosine (50 mg k⁻¹g, 0.3 mL⁻¹/time, 3 times/week), or saline started at 21 weeks of age, and lasted until 36 weeks of age. The amount of inosine in plasma from saline‐treated, Lj‐treated and inosine‐treated pristane‐induced mice was detected. G) Plasma anti‐dsDNA level was detected among HC, saline‐treated, Lj‐treated, and inosine‐treated groups. H) Statistical analyses were conducted to compare the systemic frequencies of CD4 naïve T cells and CD4 TEM cells, CD8 naïve T cells and CD8 TEM cells among HC, saline‐treated, Lj‐treated, and inosine‐treated groups. I,J) Statistical analyses were conducted to compare the systemic frequencies of I) naïve B and J) memory B cells among HC, saline‐treated, Lj‐treated, and inosine‐treated groups. K) Statistical analyses were conducted to compare the systemic frequencies and cell number of GCB among saline‐treated, Lj‐treated, and inosine‐treated groups. The results are expressed as mean ± SEM. Statistical comparison was based on Student t‐test and a one‐way ANOVA. *p < 0.05 was considered statistically significant. **p < 0.01; ***p < 0.001.

Figure 6

Inosine restricted B cell differentiation in vitro. A) Isolated mouse splenic CD19⁺ B cells were triggered by IL‐4 and R848 incubated with different concentrations of inosine (0, 10 µm, 100 µm, 1 mm, 10 mm) for 48 h. Statistical analysis of frequencies of naive B cells, memory B cells, ASC, and plasmablast with different concentrations of inosine. B) Ig isotyping in mouse B cell supernatant treated with 10 mm inosine. C) The mRNA level of genes related to B cell differentiation, including spib, pax5, il21r, bach2, irf4, cd38 and bcl2, prdm1, and bcl6. D) The mRNA level of genes related to B cell migration, including s1pr1, ebi2, ccr7, and ccr6. E) Western blot detected the ERK pathway after increased inosine treatment in mouse splenic B cells after 48 h incubation. F) GO‐KEGG enrichment analysis of the differentially expressed genes in RNA‐seq. G) Differential expressed genes in HIF signaling pathway. H) The mRNA level of hif1α and protein level of HIF‐1α in mouse spleen B cells with different concentrations of inosine (0, 10 µm, 100 µm, 1 mm, 10 mm). I) Representative figures and statistical analysis of frequencies of ASC with or without LW6 (10 µm) in human peripheral blood B cells culture for 48 h. J) Frequencies of ASC with or without LW6 (10 µm) or inosine in mouse splenic B cells cultured for 48h. K) Western blot for HIF‐1α protein with or without PD98059 treatment in mouse splenic B cells after 48 h incubation (20 µm). Data were obtained from three biologically replicated experiments. The results are expressed as mean ± SEM. Statistical comparison was based on one‐way ANOVA. *p < 0.05 was considered statistically significant; **p < 0.01; ***p < 0.001; ****p < 0.0001; ns = not significant.