Karlyshev, A.V. (2018) Application of the next generation genome sequencing for metagenomic analysis of dairy products. In: VIII International Conference on Environmental, Industrial and Applied Microbiology - BioMicroWorld2018; 24 - 25 Jun 2018, Torremolinos-Malaga, Spain. (Unpublished)

Abstract

Introduction. Yogurts and kefir products are known for their beneficial properties due to the presence of probiotic microorganisms. The beneficial effects depend both on qualitative and quantitative composition of the microflora. Composition of kefir grains and changes in microbial content during manufacture of kefir drink were previously studied using metagenomics and metabolomics [1]. In this study we compared both total and live bacterial content of six commercial dairy products (three kefir and three yogurt samples) using next generation sequencing. The data indicated remarkable differences between total and live bacterial content among the products, likely to be the result of a manufacturing process and/or storage. Methodology. Sample 1-3 and 4-6 represented different brands of kefir and yogurts respectively. Total DNA extracted from original product samples and after growth on solid medium was used for amplification of V1-V2 variable regions of the16S rRNA genes. The amplicons were sequenced using IonTorrent PGM with Ion 400 HiQ View sequencing kit and 316v2 chips. The generated sequences were run via MG RAST [2] and IonReporter metagenomics tools for qualitative and quantitative assessment of bacterial composition estimating both total and live bacteria. Results. In contrast to IonReporter, usage of MG RAST server resulted in over-representation of ‘unclassified’ bacteria. In sample 2 (before growth) MG RAST was unable to identify 91% of bacteria, whilst IonReporter assigned 99% of them as Streptococcus spp. Limited discriminative power of MG RAST was also detected with sample 1 (after growth) with 72% of bacteria reported as ‘unclassified’, which were identified to be Lactobacillus spp by IonReporter. Both programs identified Lactobacillus spp in sample 6 after growth. The MG RAST also mistakenly identified Lactococci in sample 4 containing Streptococci (confirmed by whole genome sequencing). The data suggest the predominant live bacteria in most samples being either enriched with or exclusively Lactobacillus spp., which is likely to be due to difference in bacterial growth rates either during product manufacturing. Discussion and Conclusion. The data suggest higher reproducibility, selectivity and discriminative power of IonReporter compared to a widely used metagenomics server MG RAST, producing much higher proportion of ‘unclassified’ bacteria and higher error rate, especially when distinguishing between closely related species (e.g. Streptococcus and Lactoccoccus spp). A detailed analysis of the samples with critical assessment of the identification methods, and implications of the results for food industry will be presented.

Actions (Repository Editors)

Item Control Page