Comparisons of Oral, Intestinal, and Pancreatic Bacterial Microbiomes in Patients with Pancreatic Cancer and Other Gastrointestinal Diseases

Authors

Mei Chung, Tufts University
Naisi Zhao, Tufts University
Richard Meier, University of Kansas
Devin C. Koestler, University of Kansas
Guojun Wu, New Jersey Institute for Food, Nutrition, and Health
Erika de Castillo, Forsyth Institute
Bruce J. Paster, Forsyth Institute
Kevin Charpentier, Rhode Island Hospital
Jacques Izard, University of Nebraska-LincolnFollow
Karl T. Kelsey, Brown University
Dominique S. Michaud, Tufts UniversityFollow

ORCID IDs

Chung http://orcid.org/0000-0002-5583-2870

Zhao http://orcid.org/0000-0002-0517-0709

Koestler http://orcid.org/0000-0002-0598-0146

Izard https://orcid.org/0000-0002-5904-5436

Michaud http://orcid.org/0000-0002-8107-3011

Document Type

Article

Date of this Version

2021

Citation

Journal of Oral Microbiology (2021) 13: 1887680

doi: 10.1080/20002297.2021.1887680

Comments

License: CC BY

Abstract

Background: Oral microbiota is believed to play important roles in systemic diseases, including cancer. Methods: We collected oral samples (tongue, buccal, supragingival, and saliva) and pancreatic tissue or intestinal samples from 52 subjects, and characterized 16S rRNA genes using high-throughput DNA sequencing.

Results: Bray–Curtis plot showed clear separations between bacterial communities in the oral cavity and those in intestinal and pancreatic tissue samples. PERMANOVA tests indicated that bacterial communities from buccal samples were similar to supragingival and saliva samples, and pancreatic duct samples were similar to pancreatic tumor samples, but all other samples were significantly different from each other. A total of 73 unique Amplicon Sequence Variants (ASVs) were shared between oral and pancreatic or intestinal samples. Only four ASVs showed significant concordance, and two specific bacterial species (Gemella morbillorum and Fusobacterium nucleatum subsp. vincentii) showed consistent presence or absence patterns between oral and intestinal or pancreatic samples, after adjusting for within-subject correlation and disease status. Lastly, microbial co-abundance analyses showed distinct strain-level cluster patterns among microbiome members in buccal, saliva, duodenum, jejunum, and pancreatic tumor samples.

Conclusions: Our findings indicate that oral, intestinal, and pancreatic bacterial microbiomes overlap but exhibit distinct co-abundance patterns in patients with pancreatic cancer and other gastrointestinal diseases.