Gut microbiota specifically mediates the anti-hypercholesterolemic effect of berberine (BBR) and facilitates to predict BBR's cholesterol-decreasing efficacy in patients

Abstract

Introduction: Gut microbiota has been implicated in the pharmacological activities of many natural products. As an effective hypolipidemic agent, berberine (BBR)'s clinical application is greatly impeded by the obvious inter-individual response variation. To date, little evidence exists on the causality between gut microbes and its therapeutic effects, and the linkage of bacteria alterations to the inter-individual response variation.

Objectives: This study aims to confirm the causal role of the gut microbiota in BBR's anti-hyperlipidemic effect and identify key bacteria that can predict its effectiveness.

Methods: The correlation between gut microbiota and BBR's inter-individual response variation was studied in hyperlipidemic patients. The causal role of gut microbes in BBR's anti-hyperlipidemic effects was subsequently assessed by altered administration routes, co-treatment with antibiotics, fecal microbiota transplantation, and metagenomic analysis.

Results: Three-month clinical study showed that BBR was effectively to decrease serum lipids but displayed an obvious response variation. The cholesterol-lowering but not triglyceride-decreasing effect of BBR was closely related to its modulation on gut microbiota. Interestingly, the baseline levels of Alistipes and Blautia could accurately predict its anti-hypercholesterolemic efficiency in the following treatment. Causality experiments in mice further confirmed that the gut microbiome is both necessary and sufficient to mediate the lipid-lowering effect of BBR. The absence of Blautia substantially abolished BBR's cholesterol-decreasing efficacy.

Conclusion: The gut microbiota is necessary and sufficient for BBR's hyperlipidemia-ameliorating effect. The baseline composition of gut microbes can be an effective predictor for its pharmacotherapeutic efficacy, providing a novel way to achieve personalized therapy.

Keywords: AMPK, AMP-activated protein kinase; Alistipes; BBR, berberine; Berberine (BBR); Blautia; Gut microbiota; H&E, Hematoxylin and Eosin; HFD, high-fat diet; Hypercholesterolemia; Hyperlipidemia; InsR, insulin receptor; Inter-individual response variation; LDL-c, low-density lipoprotein cholesterol; LDLR, low-density lipoprotein receptors; NPS, the non-responsive subjects; PS, the responsive subjects; RF analysis, Random forest analysis; ROC, receiving operating characteristic; SCFAs, short-chain fatty acids; TC, total cholesterol; TG, triglycerides.

Graphical abstract

Fig. 1

BBR effectively decreases blood lipids in hyperlipidemic patients. Eighty-three hyperlipidemic patients were treated with BBR (42 subjects, 1 g/day) or placebo (41 subjects) for 3 months. The lipids changes at each time point compared to the baseline levels were analyzed.

Fig. 2

The cholesterol-lowering efficiency of BBR is closely related with its modulation on gut microbiota. A-B. The principal coordinate analysis (PCoA) (A) and non-metric multi-dimensional scaling (NMDS) analysis (B) of gut microbiota. (C) The Bray-Curtis distance-based clustering analysis. (D) The changes of alpha diversity indices compared to the baseline values. We obtained the initial (week0) and final (week12) fecal samples from 51 patients (28 in BBR group and 23 in placebo group) and analyzed their gut microbiota composition by shotgun sequencing metagenomics. *P < 0.05.

Fig. 3

The baseline abundance of Alistipes and Blautia spp. are effective to predict the cholesterol-lowering efficiency of BBR in hyperlipidemic patients. (A) The genus profile of responsive (PS) and non-responsive (NPS) patients, and Alistipes is the only dominant genus whose baseline abundance is significantly different between PS and NPS patients. (B) The species profile of PS and NPS patients, and three Alistipes spp. are significantly different between PS and NPS patients at the baseline level. (C) Co-occurrence network established by SparCC analysis. The area of each node indicates the accumulated abundance of the species, and the portion of each group was displayed in different colours. The connecting edges indicate positive (orange) or negative (blue) correlations between species. (D) The top 15 species that discriminate the PS and NPS patients based on random forest analysis. (E) Receiver operating characteristic curve (ROC) for the combination of Alistipes and Blautia spp. Area under curve (AUC) and the 95% confidence interval are also shown.

Fig. 4

Parenteral administration or antibiotic treatment largely weakens BBR’s lipid-lowering effect. (A) Serum lipids and glucose levels in each treatment group; error bars denote standard error in measurements. (B) Oral glucose tolerance test (OGTT) and insulin tolerance test (ITT) in each group. (C) Hematoxylin and eosin (H&E) staining of the liver (bar = 10 μm). (D) Steatosis score of different treatment groups; each dot represents a liver sample wherein steatosis was diagnosed. (0–3) was evaluated as follows: 0, no involvement; 1, mild involvement; 2, moderate involvement; and 3, severe involvement. E-F Liver total cholesterol (TC) (E) and triglyceride (F) measurements in different treatment groups. *P < 0.05, ^**P < 0.01, ^***P < 0.001, N.S. = non-significant.

Fig. 5

Fecal material transplantation after BBR treatment prevents HFD-induced hyperlipidemia as effectively as BBR. (A–D) Bodyweight (A), bodyweight gain (B) and the weights of liver (C), subcutaneous fat (D) and epididymal fat (E) of different treatment groups, respectively. (F) Serum levels of TC, TG, LDL-c, and glucose. (G) Oral glucose tolerance test (OGTT) and insulin tolerance test (ITT). (H–J) H&E staining of liver (bar = 10 μm) (H) and hepatic levels of total cholesterol (TC)(I), and triglycerides (TG) (J) from different treatment groups, respectively. (K) HPLC for BBR in BBR soup (bottom panel) and the collected fecal material after BBR administration (top panel), indicating the absence of BBR in fecal materials used for the transplant. *P < 0.05, ^**P < 0.01, ^***P < 0.001, N.S. = non-significant.

Fig. 6

Dysregulation of Blautia downregulates the cholesterol-lowering action of BBR. (A) Bodyweight curve. (B) Bodyweight change. (C) Serum lipids levels. (D) Relative abundance of key genera that were confirmed to be closely related to BBR’s lipid-lowering effects. Data are expressed as mean ± s.e.m. N = 8 for each group. ^#P < 0.05, HFD group vs Chow group; *P < 0.05, **P < 0.01, ***P < 0.001. N.S. = non-significant.