Assessment of bacterial and viral gut communities in healthy and tumoral colorectal tissue using RNA and DNA deep sequencing

Abstract

Background: Colorectal cancer (CRC) is known to present a distinct microbiome profile compared to healthy mucosa. Non-targeted deep-sequencing strategies enable nowadays full microbiome characterization up to species level. Aim: We aimed to analyze both bacterial and viral communities in CRC using these strategies. Materials & Methods: We analyzed bacterial and viral communities using both DNA and RNA deep-sequencing (Novaseq) in colorectal tissue specimens from 10 CRC patients and 10 matched control patients. Following taxonomy classification using Kraken 2, different metrics for alpha and beta diversities as well as relative and differential abundance were calculated to compare tumoral and healthy samples. Results: No viral differences were identified between tissue types, but bacterial species Polynucleobacter necessarius had a highly increased presence for DNA in tumors (p = 0.001). RNA analyses showed that bacterial species Arabia massiliensis had a highly decreased transcription in tumors (p = 0.002) while Fusobacterium nucleatum transcription was highly increased in tumors (p = 0.002). Discussion: Sequencing of both DNA and RNA enables a wider perspective of micriobiome profiles. Lack of RNA transcription (Polynucleobacter necessarius) casts doubt on possible role of a microorganism in CRC. The association of F. nucleatum mainly with transcription, may provide further insights on its role in CRC. Conclusion: Joint assessment of the metagenome (DNA) and the metatranscriptome (RNA) at the species level provided a huge coverage for both bacteria and virus and identifies differential specific bacterial species as tumor associated.