Abstract
BACKGROUND: The composition of the gut microbiome has been recorted to be strongly associated with gestational diabetes mellitus (GDM), but mutational characterization of the microbiome in patients with GDM has been overlooked. Here, we revealed the genetic variation landscape of the gut microbiome and assessed its clinical significance in a cohort of patients with GDM. METHODS: We employed a macrogenomic dataset made up of a discovery cohort of 54 cases and a validation cohort of 220 cases to screen for high-abundance microbial flora and identified single nucleotide variants (SNVs) and insertions/deletions (indels). Subsequently, we analyzed the mutation spectra of genomes of the intestinal flora by using the previously identified SNVs and identified mutation signatures. Additionally, we utilized the Random Forest algorithm to identify key differential SNVs and elucidated their biological functions and associations with the clinicopathological parameters of GDM. RESULTS: We screened 15 key microbial flora and found that the GDM group had more SNVs and indels in the intestinal flora than the control group, with a significant increase in C > T and T > C base mutations and were more susceptible to sequence mutations. Compared to the control group, the GDM group underwent a more significant evolution, as evidenced by the presence of a unique mutational spectrum and mutational characteristics. Random Forest algorithm analysis showed that the combined characterization of five gut microbial species and 21 SNV-related markers was effective in distinguishing between GDM and control subjects in both discovery (area under the curve (AUC) = 0.86) and validation (AUC = 0.73) sets. These markers also revealed that GDM is strongly associated with sphingolipids, galactose, and proteins containing the DUF structural domain. CONCLUSIONS: The GDM intestinal flora has unique mutational features that correlate significantly with clinicopathological involvement and may be involved in the development of the disease.