Oral Microbiome in Nonsmoker Patients with Oral Cavity Squamous Cell Carcinoma, Defined by Metagenomic Shotgun Sequencing

SIMPLE SUMMARY: Smoking is the commonest cause of oral cavity squamous cell carcinoma (OC-SCC), but the cause of OC-SCC in nonsmokers is unknown. Our primary goal was to use metagenomic shotgun sequencing (MSS) to directly define the taxonomic composition and functional potential of oral metagenome in nonsmokers with OC-SCC. We found three bacterial phyla, one genus, and one species were enriched in OC-SCC while two phyla, five genera, and 18 species were enriched in controls. Pathways related to metabolism of flavin, biotin, thiamine, heme, sugars, fatty acids, peptidoglycans, and tRNA were more abundant in OC-SCC while those related to metabolism of nucleotides and essential amino acids were more abundant in controls. The MSS method achieved similar results to the commonly used 16S rRNA gene survey in compositional differentiation but differed greatly from prediction using 16S rRNA genes in functional differentiation of microbiomes in OC-SCC and controls. ABSTRACT: Objectives: Smoking is the commonest cause of oral cavity squamous cell carcinoma (OC-SCC), but the etiology of OC-SCC in nonsmokers is unknown. Our primary goal was to use metagenomic shotgun sequencing (MSS) to define the taxonomic composition and functional potential of oral metagenome in nonsmokers with OC-SCC. Methods: We conducted a case–control study with 42 OC-SCC case and 45 control nonsmokers. MSS was performed on DNA extracted from mouthwash samples. Taxonomic analysis and pathway analysis were done using MetaPhlAn2 and HUMAnN2, respectively. Statistical difference was determined using the Mann–Whitney test controlling false discovery rate. Results: There was no significant difference in age, sex, race, or alcohol consumption between OC-SCC and control patients. There was a significant difference in beta diversity between OC-SCC and controls. At the phylum level, Bacteroidetes and Synergistetes were overly represented in OC-SCC while Actinobacteria and Firmicutes were overly represented in controls. At the genus level, Fusobacterium was overly represented in OC-SCC compared with controls, while Corynebacterium, Streptococcus, Actinomyces, Cryptobacterium, and Selenomonas were overly represented in controls. Bacterial pathway analysis identified overrepresentation in OC-SCC of pathways related to metabolism of flavin, biotin, thiamin, heme, sugars, fatty acids, peptidoglycans, and tRNA and overrepresentation of nucleotides and essential amino acids in controls. Conclusions: The oral microbiome in nonsmoker patients with OC-SCC is significantly different from that of nonsmoker control patients in taxonomic compositions and functional potentials. Our study’s MSS findings matched with previous 16S-based methods in taxonomic differentiation but varied greatly in functional differentiation of microbiomes in OC-SCC and controls.