Microbiome and metabolic features of tissues and feces reveal diagnostic biomarkers for colorectal cancer

Jiahui Feng; Zhizhong Gong; Zhangran Sun; Juan Li; Na Xu; Rick F Thorne; Xu Dong Zhang; Xiaoying Liu; Gang Liu

doi:10.3389/fmicb.2023.1034325

Microbiome and metabolic features of tissues and feces reveal diagnostic biomarkers for colorectal cancer

Front Microbiol. 2023 Jan 13:14:1034325. doi: 10.3389/fmicb.2023.1034325. eCollection 2023.

Authors

Jiahui Feng¹, Zhizhong Gong¹, Zhangran Sun^{1

2}, Juan Li³, Na Xu¹, Rick F Thorne^{2

4}, Xu Dong Zhang^{2

4}, Xiaoying Liu^{1

2}, Gang Liu¹

Affiliations

¹ School of Life Sciences, Anhui Medical University, Hefei, China.
² Henan International Joint Laboratory of Non-coding RNA and Metabolism in Cancer, Henan Provincial Key Laboratory of Long Non-coding RNA and Cancer Metabolism, Translational Research Institute of Henan Provincial People's Hospital and People's Hospital of Zhengzhou University, Zhengzhou, Henan, China.
³ Department of Oncology, BinHu Hospital of Hefei, Hefei, China.
⁴ School of Biomedical Sciences and Pharmacy, The University of Newcastle, Callaghan, NSW, Australia.

Abstract

Microbiome and their metabolites are increasingly being recognized for their role in colorectal cancer (CRC) carcinogenesis. Towards revealing new CRC biomarkers, we compared 16S rRNA gene sequencing and liquid chromatography-mass spectrometry (LC-MS) metabolite analyses in 10 CRC (T_CRC) and normal paired tissues (T_HC) along with 10 matched fecal samples (F_CRC) and 10 healthy controls (F_HC). The highest microbial phyla abundance from T_HC and T_CRC were Firmicutes, while the dominant phyla from F_HC and F_CRC were Bacteroidetes, with 72 different microbial genera identified among four groups. No changes in Chao1 indices were detected between tissues or between fecal samples whereas non-metric multidimensional scaling (NMDS) analysis showed distinctive clusters among fecal samples but not tissues. LEfSe analyses indicated Caulobacterales and Brevundimonas were higher in T_HC than in T_CRC, while Burkholderialese, Sutterellaceaed, Tannerellaceaea, and Bacteroidaceae were higher in F_HC than in F_CRC. Microbial association networks indicated some genera had substantially different correlations. Tissue and fecal analyses indicated lipids and lipid-like molecules were the most abundant metabolites detected in fecal samples. Moreover, partial least squares discriminant analysis (PLS-DA) based on metabolic profiles showed distinct clusters for CRC and normal samples with a total of 102 differential metabolites between T_HC and T_CRC groups and 700 metabolites different between F_HC and F_CRC groups. However, only Myristic acid was detected amongst all four groups. Highly significant positive correlations were recorded between genus-level microbiome and metabolomics data in tissue and feces. And several metabolites were associated with paired microbes, suggesting a strong microbiota-metabolome coupling, indicating also that part of the CRC metabolomic signature was attributable to microbes. Suggesting utility as potential biomarkers, most such microbiome and metabolites showed directionally consistent changes in CRC patients. Nevertheless, further studies are needed to increase sample sizes towards verifying these findings.

Keywords: biomarkers; colorectal cancer; feces; gut microbiome; metabolomics; tissue.