Microbial imbalances linked to early pregnancy loss: a comparative analysis of vaginal microbiota.

Objective: To explore the role and related functions of vaginal microbiota in early pregnancy loss.

Methods: This study was a case-control study with a comparison group (reference group). We recruited 178 women, including 73 who had experienced at least one early clinical pregnancy loss and 105 patients with one live birth and no history of pregnancy loss. Data on demographics, disease history, menstrual and reproductive history was collected. The case group patients were sampled immediately upon presenting with pregnancy loss at their first visit. The reference group patients underwent samples when they chose to participate voluntarily. All vaginal discharge was performed DNA Preparation and Metagenomics Sequencing. DNA extraction was performed using the phenol/trichloromethane method and the DNA fragments were then size-selected to 300-700 bp using magnetic beads. The selected fragments were repaired and ligated with indexed adaptors. The captured DNA was amplified again by PCR and circularized to create a single-stranded circular (ssCir) library. The ssCir library was subsequently amplified through rolling circle amplification (RCA) to produce DNA nanoballs (DNBs). The DNBs were then loaded onto a flow cell and sequenced using the DNBSEQ Platform. Nonparametric tests, including Kruskal-Wallis and Wilcoxon tests, were employed. Relative abundance between groups was compared, and differential species selection was performed using the LEfSe software with linear discriminant analysis.

Results: 1. PCoA analysis based on Bray-Curtis distances at the species level revealed a difference between the groups (p = 0.011). At the genus level, α-diversity, assessed using the Shannon, Simpson, and Inverse Simpson indices, indicated higher bacterial richness and diversity in the control group (Shannon: mean 0.554 vs. 0.383, p = 0.0044; Simpson: mean 0.254 vs. 0.179, p = 0.0043; Inverse Simpson: mean 1.636 vs. 1.414, p = 0.0043); At the genus level, 107 microbial genera were identified, 18 of which displayed statistically significant differences. At the species level, 23 microbial species showed significant differences between the two groups. 2. We analyzed the differences in the most abundant phyla, genera, and species, with a particular focus on the top 20 most abundant genera and species. Firmicutes and Proteobacteria were significantly more prevalent among patients with pregnancy loss (PL). Among the top 20 most abundant genera, Streptococcus and Porphyromonas were significantly more abundant in patients with PL, whereas Bifidobacterium was significantly more prevalent in the reference group. Among the 20 most abundant species, Lactobacillus crispatus was significantly more prevalent in patients with PL, whereas common in the control group. 3. Principal Coordinates Analysis (PCoA) of Bray-Curtis distances, highlight their distinct clustering patterns, suggesting a notable difference between the metabolic pathways of the two groups. Key pathways with a negative correlation to PL include those related to amino acid biosynthesis, lipid metabolism, and nucleotide biosynthesis.

Conclusion: Our study highlights the association between vaginal microbiota dysbiosis and EPL, identifying specific microbial taxa that may contribute to pregnancy loss. These findings underscore the importance of the vaginal microbiome in reproductive health and open up new avenues for research into microbiome-based diagnostics and therapies. By integrating microbial, immune, and environmental data, future research has the potential to uncover the mechanisms underlying EPL and develop targeted interventions to improve pregnancy outcomes.