Alterations of bacterial communities of vocal cord mucous membrane increases the risk for glottic laryngeal squamous cell carcinoma

Abstract

Bacteria are among the important factors that play a role in the balance of human health, and their relationship with some tumors has been well established. However, the association between bacteria colonizing the vocal cords and glottic laryngeal squamous cell carcinoma (GLSCC) remains unclear. Here, we investigated whether bacterial communities of the vocal cord mucous membrane play a role in the development of GLSCC. We collected tumor tissue and normal adjacent tissue (NAT) samples from 19 GLSCC patients, and the bacterial communities were compared with control samples (control) from 21 vocal cord polyps using 16S rRNA high-throughput pyrosequencing. We detected 41 phyla, 93 classes, 188 orders, 373 families, and 829 genera in the vocal cord mucous membrane. A comparison of the bacterial communities in the NAT samples showed higher α-diversity than in the tumor samples. In the tumor samples, seven groups of bacteria, i.e., the phylum Fusobacteria, the class Fusobacteriia, the order Fusobacteriales, the family Fusobacteriaceae, and the genera Fusobacterium, Alloprevotella, and Prevotella, were significantly enriched, as revealed by linear discriminant analysis coupled with effect size measurements (LEfSe). However, bacteria from the phylum Firmicutes were most significantly enriched in the vocal cord polyp tissues. These findings suggest alterations in the bacterial community structure of the vocal cord mucous membrane of GLSCC patients and that seven groups of bacteria are related to GLSCC, indicating that imbalances in bacterial communities increase the risk for the development of GLSCC.

Keywords: 16S rRNA sequences.; bacterial communities; glottic laryngeal squamous cell carcinoma; microbiota; vocal cord.

Figure 1

Relative abundance of major bacterial phyla (A) and genera (B) whose proportions were greater than 1% in the vocal cords. The values are the mean sequence abundances in various groups and levels. Heatmap showing the relative densities of 30 of the most abundant phyla (C) and genera (D) of vocal cord tissue samples. Hierarchical dendrogram depicting the taxonomic assignments of vocal cord samples. The cluster branch groups above represent samples from different groups. The cluster tree on the left represents the cluster of genera. The legend in the upper-right corner of the figure indicates the colors that represent the relative abundances of genera in every sample (presented as a percentage of the total 16S rRNA sequences). Orange indicates a higher relative abundance of the genera, and blue shows a lower relative abundance. The tumor tissue and normal adjacent tissue (NAT) samples were taken from GLSCC patients, and the control tissue was sampled from subjects with vocal cord polyps.

Figure 2

Comparative analyses of the vocal cord bacterial communities of the three groups of tissue samples. Alpha rarefaction plot (A), Chao1 (B), Simpson index (C), and Shannon index (D) of the samples from the tumor tissue, the normal adjacent tissue (NAT), and control tissue were compared. These comparisons were not marked because P> 0.05 between groups. (E) Principal coordinate analysis (PCoA) using the weighted UniFrac distance of the vocal cord microbiota among the study participants. The weighted UniFrac distance significantly varied among the three groups (weighted UniFrac R² = 0.067, P< 0.05).

Figure 3

Alterations in the vocal cord bacterial community compositions at the phylum level. (A) Histogram of the bacterial community structure at the phylum level (top 30). Each column represents a vocal cord sample, and each color represents an individual phylum. (B) Firmicutes. (C) Fusobacteria. (D) Saccharibacteria. Comparisons among groups were performed using Kruskal-Wallis tests. These comparisons were not marked because P> 0.05 between groups.

Figure 4

Alterations in the vocal cord microbiota composition. (A) Histogram of the bacterial community structure distribution at the class level (top 30). Each column represents a vocal cord sample, and each color represents an individual class. (B) Bacteroidia. (C) Bacilli. (D) Fusobacteriia. (E) Actinobacteria. (F) Sphingobacteria. (G) Epsilonproteobacteria. (H) Nitrospira. (I) Chlorobia. Comparisons among groups were performed using Kruskal-Wallis tests. These comparisons were not marked because P> 0.05 between groups.

Figure 5

Alterations in the vocal cord microbiota composition at the order level. (A) Histogram of the bacterial community structure distribution at the order level (top 30). Each column represents a vocal cord sample, and each color represents an individual order. (B) Fusobacteriales. (C) Lactobacillales. (D) Rhodospirillales. (E) Sphingobacteriales. (F) Campylobacterales. (G) Bifidobacteriales. Comparisons among groups were performed using Kruskal-Wallis tests. These comparisons were not marked because P> 0.05 between groups.

Figure 6

Alterations in the vocal cord microbiota composition at the family level. (A) Histogram of the bacterial community structure distribution at the family level (top 30). Each column represents a vocal cord sample, and each color represents an individual family. (B) Fusobacteriaceae. (C) Ruminococcaceae. (D) Rikenellaceae. (E) Campylobacteraceae. (F) Bifidobacteriaceae. (G) Xanthomonadaceae. Comparisons among groups were performed using Kruskal-Wallis tests. These comparisons were not marked because P> 0.05 between groups.

Figure 7

Alterations in the vocal cord microbiota composition at the genus level. (A) Histogram of the bacterial community structure distribution at the genus level (top 30). Each column represents a vocal cord sample, and each color represents an individual genus. (B) Streptococcus. (C) Fusobacterium. (D) Alloprevotella. (E) Escherichia_Shigella. (F) Bifidobacterium. (G) Prevotella. Comparisons among groups were performed using Kruskal-Wallis tests. These comparisons were not marked because P> 0.05 between groups.

Figure 8

Linear discriminant analysis (LDA) coupled with effect size measurements (LEfSe) analysis. (A) Structure of the vocal cord microbiota. The cladogram depicts the relationships among microbiota taxonomic units of the bacteria from the phylum level to the genus level. The colors red, green, and blue represent bacterial taxonomic units that were abundant in vocal cord polyp tissue (control), normal adjacent tissue (NAT), and tumor tissue (tumor). (B) Histogram of LDA scores of taxonomic units demonstrating the contribution of different taxonomic units to the difference.