High-risk genotypes for type 1 diabetes are associated with the imbalance of gut microbiome and serum metabolites

Abstract

Background: The profile of gut microbiota, serum metabolites, and lipids of type 1 diabetes (T1D) patients with different human leukocyte antigen (HLA) genotypes remains unknown. We aimed to explore gut microbiota, serum metabolites, and lipids signatures in individuals with T1D typed by HLA genotypes.

Methods: We did a cross-sectional study that included 73 T1D adult patients. Patients were categorized into two groups according to the HLA haplotypes they carried: those with any two of three susceptibility haplotypes (DR3, DR4, DR9) and without any of the protective haplotypes (DR8, DR11, DR12, DR15, DR16) were defined as high-risk HLA genotypes group (HR, n=30); those with just one or without susceptibility haplotypes as the non-high-risk HLA genotypes group (NHR, n=43). We characterized the gut microbiome profile with 16S rRNA gene amplicon sequencing and analyzed serum metabolites with liquid chromatography-mass spectrometry.

Results: Study individuals were 32.5 (8.18) years old, and 60.3% were female. Compared to NHR, the gut microbiota of HR patients were characterized by elevated abundances of Prevotella copri and lowered abundances of Parabacteroides distasonis. Differential serum metabolites (hypoxanthine, inosine, and guanine) which increased in HR were involved in purine metabolism. Different lipids, phosphatidylcholines and phosphatidylethanolamines, decreased in HR group. Notably, Parabacteroides distasonis was negatively associated (p ≤ 0.01) with hypoxanthine involved in purine metabolic pathways.

Conclusions: The present findings enabled a better understanding of the changes in gut microbiome and serum metabolome in T1D patients with HLA risk genotypes. Alterations of the gut microbiota and serum metabolites may provide some information for distinguishing T1D patients with different HLA risk genotypes.

Keywords: gut microbiota; human leukocyte antigen; serum lipids; serum metabolites; type 1 diabetes mellitus.

Figure 1

Results of diversity and taxonomy. (A) At the family level,the stacked bar plots; (B) Result of β-diversity visualized using principal coordinate analysis based on Bray-Curtis Index (Permutational MANOVA test, F= 1.8078, R2= 0.02483, p<0.066); (C, D) The plots of α-diversity: (C) The chao1 -diversity boxplots (T-test, p= 0.51802). (D) The shannon index boxplots (T-test, p=0.51787).

Figure 2

Results of different taxa (screened by P value <0.05) between two groups by LEfSe analysis. (A) Histogram of LDA value distribution. (B) Cladogram plots depicting the microbial taxa that differed significantly.

Figure 3

Results of metabolites profiles analysis between two groups. (A) OPLS-DA analysis displaying a discriminative trend of metabolite composition and (B) lipid composition between two groups. (C) Differential metabolites (screened by combining P value (<0.05) and fold­ change (>1.2)) mainly involved in purine metabolism (Hypoxanthine;Inosine;Guanine). (D) Differential lipid metabolite were phosphatidylcholines and phosphatidylethanolamines, which involved in glycerophospholipid, linoleic acid, alpha-linolenic acid, arachidonic acid metabolism, and glycosylphosphatidylinositol-anchor biosynthesis.

Figure 4

(A) Results of combined analysis of microbiome and metabolome between two groups; (B) Results of combined analysis of microbiome and lipidome between two groups. **p<=0.01,*p<=0.05.