Gut barrier and microbiota changes with glycine and branched-chain amino acid supplementation in chronic haemodialysis patients

Abstract

Background: We have previously shown that glycine increases fat-free mass in chronic haemodialysis patients with features of malnutrition as compared with branched-chain amino acids (BCAAs). This multicentre randomized double-blind crossover study evaluates the impact of these amino acids on the gut barrier and microbiota.

Methods: Haemodialysis patients were included if they had plasma albumin <38 g/L or weight loss >5% of dry body weight, and daily dietary intakes <30 kcal/kg and <1 g protein/kg. They consumed glycine or BCAA (7 g twice daily) for 4 months and underwent a 1 month washout period, before crossover of supplementations. Faecal microbiota (16S rRNA gene sequencing) and immunoglobulin A (IgA), serum levels of cytokines, surrogate markers of intestinal permeability, appetite mediators, and endocannabinoids were obtained at the start and end of each supplementation. Supplementations were compared by multiple mixed linear regression models, adjusted for age, sex, month of supplementation (0 and 4 in each period), and period (Period 1: first 4 months; Period 2: last 4 months). Microbiota comparisons were performed using principal coordinate analysis and permutational multivariate analysis of variance, Shannon diversity index estimate and analysis of composition of microbiomes analysis, and Wilcoxon tests.

Results: We analysed 27 patients compliant to the supplementations. Multiple mixed linear regression models were significant only for interleukin-6 (P = 0.002), glucagon-like peptide 1 (P = 0.028), cholecystokinin (P = 0.021), and peptide YY (P = 0.002), but not for the other outcomes. The significant models did not show any impact of the type of supplementation (P < 0.05 in all models). Principal coordinate analysis and permutational multivariate analysis of variance (P = 0.0001) showed strong microbiota clustering by subject, but no effect of the amino acids. Bacterial alpha diversity and zero-radius operational taxonomic unit richness remained stable, whatever the supplementation. Lacticaseibacillus paracasei (0.030; Q1-Q3 0.008-0.078 vs. 0.004; Q1-Q3 0.001-0.070) and Bifidobacterium dentium (0.0247; Q1-Q3 0.002-0.191 vs. 0.003; Q1-Q3 0.001-0.086) significantly decreased with the BCAA supplementation.

Conclusions: The BCAA and glycine supplementations had no impact on the serum levels of cytokines, appetite mediators, intestinal permeability, endocannabinoids, or faecal IgA. Overall faecal microbiota composition and microbial diversity did not change with the glycine or BCAA supplementation but decreased the abundance of L. paracasei and B. dentium.

Trial registration: ClinicalTrials.gov NCT02962089.

Keywords: Appetite; Branched-chain amino acid malnutrition; Endocannabinoids; Glycine; Gut microbiota.

Conflict of interest statement

L.G. has received grants from the Swiss National Foundation, Alfred and Alice Lachmann Foundation, and Fresenius Kabi; speaker honoraria from Fresenius Kabi; advisory honoraria from Baxter and Abbott; and travel grants from Nestlé Health Science and Abbott. P.D.C. is a co‐founder of A‐Mansia Biotech S.A. (Belgium) and owner of several patents concerning the use of specific bacteria or components on the treatment of obesity, diabetes, and cardiometabolic disorders. D.T. has received grants from Fresenius Medical Care, Amgen, and Baxter; speaker honoraria from Fresenius Medical Care, B Braun, Abbott International, Baxter, Genzyme, and Sanofi Aventis; travel grants for Fresenius Medical Care, Amgen, and Vifor; and a grant for teaching material from Abbott International. M.P., C.S., N.M., J.M., I.B., A.W.‐G., V.L., L.G., F.R.H., and J.S. have no conflicts of interest.

© 2021 The Authors. Journal of Cachexia, Sarcopenia and Muscle published by John Wiley & Sons Ltd on behalf of Society on Sarcopenia, Cachexia and Wasting Disorders.

Figures

Figure 1

Consort flow diagram. BCAA, branched‐chain amino acid.

Figure 2

Principal coordinate analysis of Bray–Curtis similarity of bacterial communities showing that microbiota strongly cluster by subject. The analysis was based on the square‐root‐transformed relative abundance of zero‐radius operational taxonomic units. Four faecal samples were collected for each patient, at Months 0 and 4 of the glycine supplementation and Months 0 and 4 of the branched‐chain amino acid (BCAA) supplementation. As the overall microbiota similarity did not differ significantly between time points and supplementation, we omitted this legend on the figure for more clarity. The open symbols correspond to the samples of patients in the glycine–BCAA group and the filled symbols to the samples of patients in the BCAA–glycine group.

Figure 3

Alpha diversity of faecal bacterial communities. The dataset was rarefied to 41 500 reads prior to calculating (A) zero‐radius operational taxonomic unit (zOTU) richness and (B) Shannon diversity index; 0 and 4 indicate the sampling month. BCAA, branched‐chain amino acid.

Figure 4

Line plots showing the decrease of the relative abundance of Bifidobacterium dentium and Lacticaseibacillus paracasei, between Months 0 and 4, under branched‐chain amino acid (BCAA). Black lines represent the group treated first with BCAA in the first 4 months of the study and the grey lines the group treated with BCAA in the second 4 months of the study.