Abstract
The role of gut microbiota in colorectal cancer is subject to extensive research. Before usage of biorepositories for microbiome studies, it is crucial to evaluate technical feasibility of microbiome profiling from various biospecimens. The aim of this study was to assess the feasibility of DNA-extraction and microbiome profiling of samples from different sample sites, tissue sites and storage duration of a colorectal cancer biobank. Mucosa samples, mucosal scrapings and feces as well as different tissue sites (tumor, normal mucosa) were analyzed. 16S rRNA gene-based microbiome profiling with taxonomic assignment was performed on the Illumina MiSeq (Illumina, San Diego, USA) platform from stored snap frozen samples. For statistical analysis, α- and β-diversity measures, PCoA, permutational multivariate analysis of variance and graphical representation were performed. Microbiome analysis could be successfully performed in most of the samples (overall 93.3%) with sufficient numbers of high-quality reads. There were no differences between sample sites, while in some measures significant differences were found between tumor and normal mucosa (α-diversity, Shannon/Simpson Indices p = 0.028/0.027, respectively). Samples stored for up to eight years were used and storage conditions had no significant influence on the results. Tumor and tissue samples of a biobank stored long term can be successfully used for microbiome analysis. As large sample sizes are needed for association studies to evaluate microbial impact on tumorigenesis or progression of colorectal cancer, an already established biorepository may be a useful alternative to prospective clinical studies.