Blautia Coccoides is a Newly Identified Bacterium Increased by Leucine Deprivation and has a Novel Function in Improving Metabolic Disorders

Abstract Gut microbiota is linked to human metabolic diseases. The previous work showed that leucine deprivation improved metabolic dysfunction, but whether leucine deprivation alters certain specific species of bacterium that brings these benefits remains unclear. Here, this work finds that leucine deprivation alters gut microbiota composition, which is sufficient and necessary for the metabolic improvements induced by leucine deprivation. Among all the affected bacteria, B. coccoides is markedly increased in the feces of leucine‐deprived mice. Moreover, gavage with B. coccoides improves insulin sensitivity and reduces body fat in high‐fat diet (HFD) mice, and singly colonization of B. coccoides increases insulin sensitivity in gnotobiotic mice. The effects of B. coccoides are mediated by metabolizing tryptophan into indole‐3‐acetic acid (I3AA) that activates the aryl hydrocarbon receptor (AhR) in the liver. Finally, this work reveals that reduced fecal B. coccoides and I3AA levels are associated with the clinical metabolic syndrome. These findings suggest that B. coccoides is a newly identified bacterium increased by leucine deprivation, which improves metabolic disorders via metabolizing tryptophan into I3AA.

12-week-old male C57BL/6J WT mice were injected with Ad-NC or Ad-shAhR via tail vein, respectively.After 3 days, Ad-NC and Ad-shAhR mice were either fed with control diet (Cont) or (-) leucine diet ((-) Leu) for another 7 days    The higher score, the higher confidence.
(A-D) 10-week-old male C57BL/6J WT mice were treated with leucine deprivation diet ((-) Leu) or control diet (Cont) for 7 days.The cecal microbiota were collected for bacterial 16S rDNA (A and B, n=6 biological replicates per group) and metagenomic sequencing (C and D, n=3 biological replicates per group).(A) Shannon index analysis of effective reads.(B) The relative abundance of Bacteroidetes, Verrucomicrobia, Actinobacteria, Cyanobacteria, and Tenericutes at the phylum level.(C) Principal-coordinate analysis (PCoA) of the gut microbiota structure based on Bray-Curtis distance.(D) Bacterial taxonomic profiling in the top 20 bacteria at genus level.(E) Scanning electron microscopy of B. coccoides GA1.(F-H) Whole-genome sequence of B. coccoides GA1 was determined by de novo sequencing.(F) Complete genome graph of B. coccoides GA1.From inner to outer: GC skew, GC content, tRNA/rRNA, CDS (reverse and forward strand).(G) Phylogenetic tree analysis of B. coccoides GA1.(H) The Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of the encoded genes of B. coccoides GA1.All values are expressed as the mean ± SEM.Statistical comparisons were carried out by nonparametric Mann-Whitney U test; *p < 0.05 and **p < 0.01.

Figure S3 .
Figure S3.The metabolic effect of B. coccoides.NCD and HFD-fed mice were orally gavage with PBS, heat-killed B. coccoides GA1 (KBC), or live B. coccoides GA1 (LBC) for 8 weeks, respectively (n = 6-8 biological replicates per group).(A, B) Absolute abundance of B. coccoides or total microbial DNA per gram of feces.(C) Western blot analysis of p-IR, p-AKT, and p-GSK3β levels in liver.The right panel is the densitometry analysis of the relative abundance of phosphorylated proteins normalized to their total protein levels.A.U.: arbitrary units.(D) Food intake.(E) Body weight.All values are expressed as the mean ± SEM.Statistical comparisons were carried out by one-way ANOVA; *p < 0.05, **p < 0.01, and n.s.: no significance.

Figure S4 .
Figure S4.Oral administration of B. coccoides DSM935 improves metabolic disorders in HFD mice.HFD mice were either orally gavage with PBS or live B. coccoides DSM935 strain (DSM935) for 8 weeks (n = 6-8 biological replicates per group).(A) Experimental procedure.(B) Fed and fasting blood glucose levels.(C) Fed and fasting serum insulin levels assayed by ELISA.(D) HOMA-IR index.(E) Glucose tolerance tests.The right panel is AUC.(F) Insulin tolerance tests (0.75 U/kg).The right panel is the AUC.(G) Body weight.(H) Total fat mass.(I) The H&E staining of sWAT.Scale bars, 50 µm.The right panel is the frequency distribution of adipocyte cell size in sWAT and the box plot is average adipocyte diameter.(J) Real-time PCR analysis of browning related genes (Ucp1, Pgc1α, Prdm16, and Cidea) in sWAT.(K) Western blot analysis

Figure S5 .
Figure S5.Colonization of B. coccoides in gnotobiotic mice improves insulin sensitivity.(A-I) 10-week-old male C57BL/6J WT mice were given antibiotic cocktail (Abx) in drinking water or given autoclaved water (Con).After 4 weeks, the Con or Abx mice were gavage with PBS or B. coccoides, respectively (n=7-9 biological replicates per group).o.g.: oral gavage.(A) Experimental procedures.(B) The DNA contents of cecal microbiota at day 0, 3, and 14 after Abx treatment, respectively.(C) The relative abundance of B. coccoides at day

Figure S6 .
Figure S6.B. coccoides metabolites enrichment, I3AA levels, and AhR targeted gene expression.(A) The growth media of B. coccoides after 48 h culture was collected to incubate primary hepatocytes for 48 h at indicate concentrations, and then stimulated with 100 nM insulin for 20 min (n=5-6 replicates per group).Western blot analysis of p-IR, p-AKT, and p-GSK3β levels in primary hepatocytes.The right panel is the densitometric analysis of the relative abundance of phosphorylated proteins normalized to their total protein levels.A.U.: arbitrary units.(B) Enrichment analysis of differential metabolites.(C-F) HFD mice were either orally gavage with PBS or live B. coccoides (LBC) daily for 8 weeks (n=5-7 mice per group).(C, D) I3AA levels in liver (C) or sWAT (D) detected by LC-MS/MS.(E, F) Real-time PCR analysis of AhR, Cyp1a1, or Cyp1b1 mRNA levels in liver (E) or sWAT (F).(G, H) The I3AA levels in cecal feces (G) and sera (H) of mice fed with HFD or NCD for 16 weeks (n=5-7 biological replicates per group).(I) Serum I3AA levels in mice fed with control

Figure S7 .
Figure S7.Metabolic effects of I3AA in vitro and in vivo.(A, B) Primary hepatocytes (A) or differentiated adipocytes (B) were incubated with indicate dose of I3AA for 48 h.Real-time PCR analysis of AhR and its target gene Cyp1a1 mRNA levels (n= 5-6 replicates per group).(C) HFD mice were oral gavage with indicated dose of I3AA for 2 weeks.Real-time PCR analysis of AhR and its target gene Cyp1a1 mRNA levels in the liver (n= 3 biological replicates per group).(D-I) HFD mice were orally gavage with PBS or 10 mg/kg I3AA for 4 weeks, respectively (n=6-7 biological replicates per group).(D) Experimental procedure.(E) Serum I3AA levels detected by LC-MS/MS.(F) Body weight.(G) Food intake.(H) Energy expenditure (EE).(I) Fecal energy loss.All values are expressed as the mean ± SEM.Statistical comparisons were carried out by unpaired two-tailed Student's t test; *p < 0.05 and **p < 0.01.

Figure S8 .
Figure S8.I3AA increased primary hepatocytes insulin signaling is depend on AhR.Primary hepatocytes were transfected with negative control small interfering RNA (-siAhR) or siRNA targeting at mouse AhR (+ siAhR) for 24h, and then incubated with or without 500 μM I3AA for the next 48 hours.At last, cells were additionally incubated with 100 nM insulin for 20 min in (B) to detect insulin signaling (n= 5-6 replicates per group).(A) Real-time PCR analysis of AhR and Cyp1a1 mRNA levels.(B) Western blot analysis of p-IR, p-AKT, p-GSK3β, and AhR levels.The right panel is the densitometry analysis of the relative abundance of phosphorylated proteins normalized to their total protein or Actin levels.A.U.: arbitrary units.All values are expressed as the mean ± SEM.Statistical comparisons were carried out by two-way ANOVA; *p < 0.05, **p < 0.01.

Figure S9 .
Figure S9.B. coccoides cannot induce metabolic improvements in global AhR knockout mice.(A-E) The basic phenotypes in wild-type (WT) and AhR knockout (Ahr -/-) mice (n=6-7 biological replicates per group).(A) Real-time PCR analysis of AhR mRNA levels in liver or sWAT.(B) Western blot analysis of AhR protein levels in liver or sWAT.Red arrow indicates specific protein bands of AhR.(C) Food intake.(D) Body weight.(E) Fed blood glucose levels.(F-N) WT and Ahr -/- mice were fed with HFD for 12 weeks, and then were either orally gavage with PBS or live B. coccoides (LBC) for 8 weeks (n=4-8 biological replicates per

Figure S10 .
Figure S10.B. coccoides induced improvement in insulin sensitivity is depend on liver AhR.HFD mice were injected with AAV-shGFP (-AAV-shAhr) or AAV-shAhr (+ AAV-shAhr) by tail vein, respectively.After 3 weeks, all mice were gavage with live B. coccoides (LBC) for 4 weeks (n=5-6 biological replicates per group).(A) Western blot analysis of liver AhR expression.The bottom panel is the densitometry analysis of AHR protein levels.A.U.: arbitrary units.(B) Fed and fasting blood glucose levels.(C) Fed and fasting serum insulin levels assayed by ELISA.(D) HOMA-IR index.(E) Glucose tolerance tests.The right panel is the AUC.(F) Insulin tolerance tests (0.75 U/kg).The right panel is the AUC.(G) Western blot analysis of p-IR, p-AKT, and p-GSK3β levels in liver with or without insulin stimulation.The right panel is the densitometric analysis of the relative abundance of phosphorylated proteins normalized to their total protein levels.(H) Body weight.(I) Total fat mass.All values are expressed as the

Figure S11 .
Figure S11.Leucine deprivation induced metabolic improvements are partially dependent on liver AhR.

(
n=5-6 biological replicates per group).(A) Western blot analysis of AhR protein levels in the liver.The right panel is the densitometry analysis of AhR protein levels.A.U.: arbitrary units.(B) Fed and fasting blood glucose levels.(C) Fed and fasting serum insulin levels assayed by ELISA.(D) HOMA-IR index.(E) Glucose tolerance tests.The right panel is AUC.(F) Insulin tolerance tests (0.5 U/kg).The right panel is the AUC.(G) Body weight.(H) Total fat mass.(I) The H&E staining of sWAT.Scale bars, 50 µm.The right panel is the frequency distribution of adipocyte cell size in sWAT and the box plot is average adipocyte diameter.(J) Real-time PCR analysis of the gene expression related to sWAT browning, including Ucp1, Pgc1α, Prdm16, and Cidea.(K) Western blot analysis of UCP1 protein levels.The right panel is the densitometry analysis of UCP1 protein levels.All values are expressed as the mean ± SEM.Statistical comparisons were carried out by two-way ANOVA; *p < 0.05, **p < 0.01, ***p < 0.001, and n.s.: no significance.

Figure S12 .
Figure S12.Leucine deprivation-B.coccoides-I3AA-AhR axis affects FGF21 levels both in the liver and serum.(A) 12-week-old male C57BL/6J WT mice were fed with control (Cont) diet or leucine deprivation ((-) Leu) diet for indicated time (n=5-6 biological replicates per group).(B) HFD mice were either orally gavage with PBS or B. coccoides DSM935 strain for 8 weeks (n=5-6 biological replicates per group).(C) HFD mice were either orally gavage with PBS or 10 mg/kg I3AA for 4 weeks (n=5-6 biological replicates per group).(D) HFD mice injected with Ad-NC or Ad-shAhR via tail vein, respectively.And then, the Ad-NC or Ad-shAhR mice were either orally gavage with PBS or 10 mg/kg I3AA daily for 10 days (n=4-6 biological replicates per group).(A-D) Western blot analysis of FGF21 protein levels in the liver or serum.The right panel is the densitometry analysis of FGF21 protein levels.A.U.: arbitrary units.All values are expressed as the mean ± SEM.Statistical comparisons were carried out by unpaired two-tailed Student's t test; *p < 0.05 and n.s.: no significance.

Figure S13 .
Figure S13.B. coccoides reshapes the structures of the gut microbiota in HFD mice.NCD and HFD mice were either orally gavage with PBS, heat-killed B. coccoides (KBC) or live B. coccoides (LBC) for 8 weeks.Cecal microbiota were used for bacterial 16S rDNA sequencing (n = 5 biological replicates per group).(A) Principal-coordinate analysis (PCoA) of the microbial composition.(B) Relative abundance of obtained species.(C) Bacteroidetes:Firmicutes ratio.(D) Analysis of the top 30 bacterial genus.(E) The abundance of detailed bacteria genus.All values are expressed as the mean ± SEM.Statistical comparisons were carried out by the Kruskal-Wallis test followed by a Dunn's post hoc test; *p < 0.05 and **p < 0.01.