Human gut microbiota is a neglected endocrine organ in humans that participates in the interactions between the gut and xenobiotics, and impacts the pathogenesis of chronic host diseases such as type 2 diabetes mellitus (T2DM). The advancement of next-generation sequencing (NGS) technologies over the last decade has dramatically increased the understanding of gut microbiota, their roles, and functions in T2DM. However, the antidiabetic drug-microbiome interactions in T2DM patients are poorly understood.

The first aim of this Ph.D. research was to identify bacterial and functional signatures in the early T2DM stage. Faecal metagenomic and metaproteomic analyses were conducted using Suzhou T2D cohort with 77 newly diagnosed T2DM patients (ND-T2D), 80 prediabetes individuals (Pre-DM), and 97 individuals with normal glucose tolerance (NGT). The three groups had substantial distinct signatures. For instance, Pre-DM and ND-T2D individuals had similar alterations, including lower abundances of Clostridium species and higher abundances of Megasphaera elsdenii compared with NGT individuals. Escherichia coli was enriched in Pre-DM individuals, while several human antimicrobial peptides were enriched in NGT.

The main aim of the thesis was to dissect the crosstalk between gut microbiota and oral antidiabetic drugs (OADs). A randomized, open-labelled, two-arm trial was conducted using Chinese ND-T2D patients exposed to a 6-month acarbose (n=50) or vildagliptin (n=50) monotherapy. Acarbose significantly affected human gut microbiota compared with vildagliptin. Meanwhile, the OADs targeting the gastrointestinal tract with different mechanisms caused similar human gut microbiota changes. The 3- or 6-month acarbose treatment induced similar changes in microbial alpha diversity, species, and pathway abundances based on a published ND-T2D cohort. Otherwise, baseline gut microbiota, rather than baseline Glucagon-like peptide-1 (GLP-1) levels, was strongly associated with GLP-1 responses to vildagliptin. A randomized, double-blinded, placebo-controlled trial was also conducted with ND-T2D patients (n=409) from 20 centres in China. The enrolled participants were randomly assigned into berberine (BBR)+placebo, probiotics+BBR, probiotics+placebo, and placebo intervention groups. The randomized controlled trial (RCT) showed that BBR treatment significantly influenced human gut microbiota. Furthermore, probiotics+BBR treatment significantly improved postprandial hyperlipidaemia (PL), suggesting the potential value of BBR combined with probiotic therapy. Last, in vitro experiments revealed the potential mechanism of combined therapy for improving PL.

Overall, these findings demonstrate the gut microbial changes in Chinese Pre-DM and ND-T2D patients, as well as bi-directional crosstalk between gut microbiota, host, and OADs. Future large-scale, well-designed longitudinal trials on T2DM and multi-omics technologies are needed to assess the mechanism and interactions between antidiabetic drugs and human gut microbiota and enhance the translation from microbiome-based clinical research into precise T2DM medicine.