Dysbiosis of the Saliva Microbiome in Patients With Polycystic Ovary Syndrome

Abstract

Significant differences in salivary microbiota communities between polycystic ovary syndrome (PCOS) patients and healthy controls have been reported, and interestingly, some salivary microbiota exhibit diurnal oscillation in healthy people. However, whether the diurnal oscillation of salivary microbiota is present in PCOS patients is unknown. In this study, we describe the differences in the saliva microbiome between the PCOS group and the control group at different time points over 24 h. 16S rRNA gene amplicon sequencing was performed on salivary and fecal samples from 10 PCOS patients and 10 healthy controls, and salivary samples were collected at 6-h intervals over 24 h (Zeitgeber (ZT)0, ZT6, ZT12, and ZT18). Among the salivary samples, those from the PCOS group showed significant differences from those of the control group at each time point. Differences were evident in taxa level and metabolic pathways. Interestingly, we found that PCOS disrupted the diurnal rhythm of the salivary microbiota abundance, as determined in the group of healthy women. In addition, no similar changes were found in PCOS patients and controls between the oral and fecal microbiota, including differential microbiota at the phylum level. In this study, significant differences in the composition of the salivary microbiota between PCOS and healthy women were detected at different time points. We also showed that the diurnal rhythm of relative abundance of the salivary microbiota was disrupted in patients with PCOS, which might be related to development of oral-related diseases and systematic metabolic disorders.

Keywords: 16S rRNA; diurnal rhythm; fecal microbiota; polycystic ovary syndrome; salivary microbiome.

Figure 1

Relative abundance of salivary microbiota at different taxomal levels and comparison between the control and PCOS groups at different time points. The mean relative abundances at the phylum (A) and genus (B) levels in different groups are represented. Different colors represent different microbiota, and area sizes represent the relative content of the microbiota composition.

Figure 2

Differences in the diurnal oral microbe community structure between the control group and the PCOS group. PCoA of the salivary pellets colored by different groups (A) collected over 24 h from healthy controls and PCOS patients, analyzed using the unweighted UniFrac method. (B) PCoA analysis of the salivary microbiota, colors represent Zeitgeber (ZT) times.

Figure 3

Identification of specific bacterial taxa in PCOS patients and controls at different time points. Linear discriminant analysis (LDA) effect size (LEfSe) was used to identify biomarkers with significant differences between the two groups: Control.ZT0 vs. PCOS.ZT0 (A), Control.ZT6 vs. PCOS.ZT6 (B), Control.ZT12 vs. PCOS.ZT12 (C), and Control.ZT18 vs. PCOS.ZT18 (D). LDA score assessed the size of the differentiation between the two groups with a score threshold of 4.0. The enriched taxa in the PCOS group (right side) are shown with a positive LDA score, whereas the taxa increased in the control group (left side) have a negative score.

Figure 4

KEGG pathway analysis showing significant differences between the PCOS and control groups at different time points. Functional prediction of the altered salivary microbiota between the two groups by PICRUSt analysis was based on KEGG pathways: Control.ZT0 vs. PCOS.ZT0 (A), Control.ZT6 vs. PCOS.ZT6 (B), Control.ZT12 vs. PCOS.ZT12 (C), and Control.ZT18 vs. PCOS.ZT18 (D). Differences were considered significant at p < 0.05 using the Student’s t-test.

Figure 5

Relative abundance of significant taxa in the PCOS and control groups across the time points. The diurnal oscillation of salivary microbiota was present in the control and PCOS groups, including p_Proteobacteria, p_Bacteroidetes, p_Actinobacteria, c_Bacteroidales, o_Bacteroidales, o_Lactobacillales, f_Prevotellaceae and g_unidentified-Prevotellaceae. The RM Friedman test indicated that various individual taxa were significantly different across time points in the control group (A–H) and the PCOS group (E, G). * or ^# Dunn’s multiple comparison test indicated significant differences between time points in either the control or PCOS group (* represents the control group and ^# represents the PCOS group). Error bars=SEM.