A mechanism by which gut microbiota elevates permeability and inflammation in obese/diabetic mice and human gut

Abstract

Objective: Ample evidence exists for the role of abnormal gut microbiota composition and increased gut permeability ('leaky gut') in chronic inflammation that commonly co-occurs in the gut in both obesity and diabetes, yet the detailed mechanisms involved in this process have remained elusive.

Design: In this study, we substantiate the causal role of the gut microbiota by use of faecal conditioned media along with faecal microbiota transplantation. Using untargeted and comprehensive approaches, we discovered the mechanism by which the obese microbiota instigates gut permeability, inflammation and abnormalities in glucose metabolism.

Results: We demonstrated that the reduced capacity of the microbiota from both obese mice and humans to metabolise ethanolamine results in ethanolamine accumulation in the gut, accounting for induction of intestinal permeability. Elevated ethanolamine increased the expression of microRNA-miR-101a-3p by enhancing ARID3a binding on the miR promoter. Increased miR-101a-3p decreased the stability of zona occludens-1 (Zo1) mRNA, which in turn, weakened intestinal barriers and induced gut permeability, inflammation and abnormalities in glucose metabolism. Importantly, restoring ethanolamine-metabolising activity in gut microbiota using a novel probiotic therapy reduced elevated gut permeability, inflammation and abnormalities in glucose metabolism by correcting the ARID3a/miR-101a/Zo1 axis.

Conclusion: Overall, we discovered that the reduced capacity of obese microbiota to metabolise ethanolamine instigates gut permeability, inflammation and glucose metabolic dysfunctions, and restoring ethanolamine-metabolising capacity by a novel probiotic therapy reverses these abnormalities.

Trial registration number: NCT02869659 and NCT03269032.

Keywords: diabetes mellitus; gut inflammation; inflammation; intestinal microbiology; obesity.

Figure 1

Obese microbiota is causal to instigate elevated gut permeability, inflammation and metabolic dysregulation. (A, B) Mice receiving FMT from both db/db and DIO mice showed significant increase in FITC 4kDa-dextran (A), endotoxin (LPS) (B) leakage from gut to blood compared with their lean control FMT recipient mice. (C–F) In addition, these mice also show significantly increased levels of systemic markers of elevated gut permeability (LBP (C) and sCD14 (D)), as well as microbial 16S rDNA (fold change (FC)) in serum (E) mRNA expression of inflammatory markers (Il1β, Il6 and Tnfα) (F) compared with their controls. (G) The TLR4 activity (absorbance at 650 nm) was significantly increased in HEK-Blue mTLR4 cells treated with serum of db/db and DIO FMT recipient mice compared with the lean FMT controls. (H–L) Obese FMTs also increased fasting hyperglycaemia (H) along with impaired meal tolerance test (MTT) (I), insulin tolerance test (ITT) (J), increased serum insulin (K) and insulin resistance index (HOMA-IR) (L, M) Random forest analysis of gene expression data revealed that obese FMTs and FCMs treatment dramatically reduced Zo1 expression in intestine and enteroids, respectively, compared with their lean FMTs/ FCMs treated controls. (N–Q) Further, the expression of Zo1 mRNA (N–P) and protein (O) in the ileum (N, O), enteroids (P) and Caco-2 cells (Q) of obese FMTs recipient mice and FCMs treated enteroids and Caco-2 cells, respectively, compared with their controls. Values presented are mean (n=5–8 mice per group) and error bars as the SEM. Enteroids and Caco-2 cell culture experiments were performed in triplicates and repeated 2–3 times. *p<0.05, **p<0.01, ***p<0.001 are statistically significant analysed by the t-test and/or ANOVA. ANOVA, analysis of variance; DIO, diet-induced obese; FCM, faecal conditioned media; FMT, faecal microbiota transplantation; HOMA-IR, Homeostatic Model Assessment for Insulin Resistance; LBS, LPS-binding protein; LPS, lipopolysaccharide.

Figure 2

Obese gut accumulates higher ethanolamine, which in turn instigates gut permeability, inflammation and impaired glucose metabolism. (A) Principal component analysis (PCA) of metabolomics data shows that metabolites in the faeces of obese (db/db (red) and DIO (blue)) mice compared with their controls (B6 NC (black) and B6 LFD (gold accent)) mice are significantly distinct. (B) Random forest analysis shows that ethanolamine abundance was significantly higher in the obese gut compared with their controls. (C, D) Ethanolamine most dramatically reduced the expression of Zo1 mRNA in the enteroids (C) and Caco-2 cells (D) among the top 6 selected metabolites (red in b panel) such as isoleucine, leucine, anserine, valine and cholic acid. (E, F) Further, ethanolamine (Et) treatment also dramatically increased the permeability of FITC-dextran (E) and reduced TEER (F) in the monolayers of Caco-2 cells. (G–L) Oral administration of ethanolamine (Et) (1 g/kg body weight for 7 days) in mice significantly reduced expression of Zo1 (mRNA (g) and protein (h)) in ileum along with increased gut permeability (FITC (4 and 40 kDa) and LPS-conjugate- dextran assay (I–K)), along with increased TLR4 activity in mouse serum treated HEK-Blue mTLR4 cells (L). (M–P) It also increased the inflammation (Il1β, Il6 and Tnfα) (M) in the intestine (ileum) and impaired insulin MTT (N) and ITT (O) with increased fasting blood glucose (P) compared with their controls. Values presented are mean of n=5–8 mice and n=2–3 repeated triplicate enteroids and Caco-2 cell culture experiments in each group, and error bars are SEM. *p<0.05, **p<0.01, ***p<0.001 are statistically significant analysed by the t-test and/or ANOVA. ANOVA, analysis of variance; DIO, diet-induced obese; ITT, insulin tolerance test; LFD, low-fat diet; LPS, lipopolysaccharide; MTT, meal tolerance test; TEER, transepithelial electrical resistance.

Figure 3

Ethanolamine abundance in the gut of obese mice and humans increases due to its undermetabolisation by microbiota. (A) The expression of ethanolamine utilising (Eut) operon genes (eutA, eutB, eutC, eutD, eutP, eutQ, eutS, eutT and aggregate of all genes as Eut operon) was significantly decreased in the faeces of db/db and DIO mice compared with lean controls. (B) Interestingly, the reduced expression of these genes was negatively correlated with ethanolamine (Et) abundance, as well as markers of elevated gut permeability (FITC-dextran leakiness, LBP and sCD14) and inflammatory markers (Il1β, Il6 and Tnfα) and positively correlated with the expression of Zo1 in the mouse intestine. (C, D) Interestingly, PCA analyses (C) show that metabolite signature was significantly different in 10 normal weight (control) compared with 10 obese subjects, and differential abundance analyses in volcano graph (D) show that ethanolamine abundance was significantly higher in the gut of obese compared with control subjects. (E) Increased ethanolamine abundance shows a strong positive correlation. (F,G) Notably, the expression of ethanolamine utilising operon genes (F) was significantly reduced in obese stools compared with their controls and showed a negative correlation with BMI (G). Values presented are mean of n=5–8 mice and n=10 lean and n=10 obese subjects in each group, and error bars are SEM. PCA (C), volcano (D) and correlation analyses (E,G) show individual sample values. *p<0.05, **p<0.01, ***p<0.001 are statistically significant analysed by the t-test and/or ANOVA. ANOVA, analysis of variance; BMI, body mass index; LBP, LPS-binding protein; LPS, lipopolysaccharide; PCA, principal component analysis.

Figure 4

Ethanolamine and obese microbiota enhance the expression of miR-101a-3p, which in turn induce gut permeability by reducing ZO-1 expression. (A) PCA graph of global miRNA profiles from enteroids treated with FCMs of db/db and DIO mice show significantly distinct miRNA expression profiles compared with lean controls. (B–D) Random forest analyses of miRNA data revealed that miR-101a-3p expression was significantly increased in enteroids treated with FCMs of db/db and DIO mice compared with lean controls (B), which was verified by real-time qPCR analyses in enteroids (C) and Caco-2 cells (D). (E, F) The expression of miR-101a-3p expression was also significantly higher in both ileum (E) and colon (F) of donor db/db and DIO mice. (G–J) The expression of miR101a-3p was also increased in the ileum (G, I) and colon (H, J) of FMT recipients (G, H) and ethanolamine (Et) treated mice (I, J) compared with their controls. (K–N) Intriguingly, enema of lentivirus expressing miR-101a-3p mimetic significantly increased gut permeability (FITC-dextran (K), LBP (L) and sCD14 (M) in serum), and that serum increases TLR4 activity in HEK-Blue mTLR4 cells (N) which corresponds to increased inflammation (Ilβ, Il6 and Tnfα (o)) and reduced Zo1 mRNA (P) and protein (Q) in the mice gut compared with their scrambled miR lentivirus infected mice. Values presented are the mean of n=5–8 mice in each group, and error bars are the SEM. PCA (A) and random forest analyses (B) show individual sample values. *p<0.05, **p<0.01, ***p<0.001 are statistically significant analysed by the t-test and/or ANOVA. ANOVA, analysis of variance; DIO, diet-induced obese; FCM, faecal conditioned media; FMT, faecal microbiota transplantation; PCA, principal component analysis.

Figure 5

Ethanolamine increases miR-101a-3p expression by increasing its promoter activity by enhancing transcription factor ARID3a binding on it. (A) Ethanolamine increased miR-101a-3p promoter depicted by luciferase assay in Caco-2 cells transfected with a vector carrying miR-101a-3p promoter (−1 to −2000 bp of transcript start site (TSS)) compared with empty vector-transfected cells. (B) Further, ethanolamine treatment significantly increased miR-101a-3p promoter activity in the Caco-2 cells transfected with vectors carrying −1 to −1000 bp and −1 to 1500 bp fragments, while no change was observed in cells transfected with a vector carrying −1 to 500 bp and empty vector. (C) Unbiased and untargeted ChiP- pull-down analyses revealed that a transcription factor-ARID3a was the highest protein pulled out with −500–1000 bp fragment compared with a scrambled nucleotide DNA sequence. (D, E) Interestingly, expression of Arid3a mRNA (D, E) and protein (F) was found significantly higher in the gut of donor db/db and DIO mice (D) as well as in FMT recipients (E, F) (G–I) In addition, obese/T2D FCMs (G–I) and ethanolamine (J–N) treatments significantly increased the expression of Arid3a in the enteroids (G, J), Caco-2 cells (H, I, K, L) and mouse intestine (M, N) compared with controls. (O–Q) Notably, ethanolamine-mediated activation of miR-101a-3p promoter activity (O), miR-101a-3p expression (P), and suppression in Zo1 expression (Q) were abolished in ARID3a siRNA transfected Caco-2 cells compared with scrambled siRNA transfected cells. Values presented are the mean of n=5–8 mice, n=2–3 repeated triplicate enteroids and Caco-2 culture experiments in each group; error bars are the SEM. *p<0.05, **p<0.01, ***p<0.001 are statistically significant analysed by independent t-test and/or ANOVA. ANOVA, analysis of variance; FCM, faecal conditioned media; FMT, faecal microbiota transplantation; T2D, type 2 diabetes.

Figure 6

Ethanolamine induces miR-101a-3p, which in turn reduces Zo1 expression by decreasing its mRNA stability. (A, B) Correlation networking analyses of miRNA and intestinal cell-specific gene expression profiles in intestines from obese (db/db (A) and DIO (B)) FMT recipient mice and enteroids treated with obese FCMs show that miR-101a-3p and Zo1 show the highest negative correlation. (C) Representation of three miR-101a-3p binding sites on human Zo1 mRNA 3’UTR sequence (seed positions). (D) miR-101a-3p mimetic (Agomir) significantly reduced expression of Zo1 mRNA while miR-101a-3p inhibitor oligonucleotide reversed it. (E) The miR-101a-3p mimetic significantly reduced the stability of Zo1 mRNA, while the miR-101a-3p inhibitor reversed it. (F) The miR-101a-3p inhibitor abolished the ethanolamine effects of reducing Zo1 mRNA expression. Values presented are the mean of n=2–3 repeated triplicate of Caco-2 culture experiments in each group, and error bars are the SEM. *p<0.05, **p<0.01, ***p<0.001 are statistically significant analysed by t-test and/or ANOVA. 3'-UTR, 3'-untranslated region; ANOVA, analysis of variance; FCM, faecal conditioned media; FMT, faecal microbiota transplantation; DIO, diet-induced obese.

Figure 7

The ileum is primarily deficient in ethanolamine-metabolising bacteria with higher leakiness, and a high-fat diet (HFD) and meat-supplemented diet (MSD) reduce them. (A–D) abundance of Eut operon (A) significantly and predominantly reduced in ileum of obese FMT recipient mice compared with other sections like duodenum, jejunum and colon, which were linked with increased expression of ARID3a (B), miR101a-3p (C) and reduced Zo1 (D) mRNA in the ileal section. (E–G) similarly, ethanolamine (et) treatment primarily increased the expression of ARID3a (E) and miR101a-3p (F) and reduced Zo1 (G) mRNA primarily in the ileum compared with other sections like duodenum, jejunum and colon. (H–W) The feeding of HFD (H–O) and MSD (P–W) similarly and significantly reduced the abundance of Eut operon-containing microbes (H, P) along with increased expression of ARID3a (I,Q) and miR101a-3p (J, R), reduced expression of Zo1 (K, S) along with increased gut permeability (FITC-dextran (L,T)), inflammatory markers (Il1β, Il6 and Tnfα (M, U)) and impaired MTT (N, V) and ITT (O, W) compared with their normal chow-fed controls. Values presented are the mean of n=5–8 mice in each group, and error bars are the SEM. *p<0.05, **p<0.01, ***p<0.001 are statistically significant analysed by the t-test and/or ANOVA. ANOVA, analysis of variance; CMT, cecal microbiota transplantation; FMT, faecal microbiota transplantation; HFD, high-fat diet; ITT, insulin tolerance test; MTT, meal tolerance test.

Figure 8

A probiotic therapy restores ethanolamine-metabolising capacity in microbiota which in turn mitigates elevated gut permeability, inflammation and metabolic impairment by restoring homoeostasis in ARID3a/miR-101a-3p/Zo1 axis. (A) Screening of ethanolamine-metabolising capacity of human origin probiotics using a colorimetry assay and measuring expression of ethanolamine using operon genes to find Lactobacillus rhamnosus HL-200 (HL-200) as a potential ethanolamine-metaboliser. (B–G) Feeding of HL-200 to mice significantly reduced the ethanolamine mediated elevation in gut permeability (FITC-dextran (b), LBP (c) and sCD14 (d)), inflammation (Il1β, Il6 and Tnfα (e)) and impaired area under curve (AUC) MTT (F), AUC ITT (G).(H) The serum of HL-200 fed mice also show less TLR4 activity in HEK-Blue mTLR4 cells treated with serum. (I–M) It also increased the abundance of Eut operon expressing microbes (I) and reduced expression of ARID3a (J), and miR-101a-3p (K) and increased Zo1 (both mRNA and protein) (L,M) compared with controls. (N–U) In addition, HL-200 treatment also significantly reduced the meat supplemented diet (MSD) feeding induced elevation in gut permeability (N), serum-mediated TLR4 activity (O), inflammation (Il1β, Il6 and Tnfα (p)), and impair AUC MTT (Q) and AUC ITT (R) along with reduced ARID3a (S) and miR101a-3p (T) along with increased expression of Zo-1 mRNA (U) compared with MSD treated controls and restored them close to control diet-fed mice. Values presented are mean (n=5–8 mice per group) and error bars as the SE of means. Microbial culture experiments were performed in triplicates and repeated 2–3 times. *p<0.05, **p<0.01, ***p<0.001 are statistically significant analysed by the t-test and/or ANOVA. ANOVA, analysis of variance; ITT, insulin tolerance test; MSD, meat-supplemented diet; MTT, meal tolerance test.