Exploring animal food microbiomes and resistomes via 16S rRNA gene amplicon sequencing and shotgun metagenomics

Abstract

As a diverse and complex food matrix, the animal food microbiota and repertoire of antimicrobial resistance (AMR) genes remain to be better understood. In this study, 16S rRNA gene amplicon sequencing and shotgun metagenomics were applied to three types of animal food samples (cattle feed, dry dog food, and poultry feed). ZymoBIOMICS mock microbial community was used for workflow optimization including DNA extraction kits and bead-beating conditions. The four DNA extraction kits (AllPrep PowerViral DNA/RNA Kit, DNeasy Blood & Tissue Kit, DNeasy PowerSoil Kit, and ZymoBIOMICS DNA Miniprep Kit) were compared in animal food as well as the use of peptide nucleic acid blockers for 16S rRNA gene amplicon sequencing. Distinct microbial community profiles were generated, which varied by animal food type and DNA extraction kit. Employing peptide nucleic acid blockers prior to 16S rRNA gene amplicon sequencing was comparable with post-sequencing in silico filtering at removing chloroplast and mitochondrial sequences. There was a good agreement between 16S rRNA gene amplicon sequencing and shotgun metagenomics on community profiles in animal feed data sets; however, they differed in taxonomic resolution, with the latter superior at resolving at the species level. Although the overall prevalence of AMR genes was low, resistome analysis of animal feed data sets by shotgun metagenomics revealed 10 AMR gene/protein families, including beta-lactamases, erythromycin/lincomycin/pristinamycin/tylosin, fosfomycin, phenicol, and quinolone. Future expansion of microbiome and resistome profiling in animal food will help better understand the bacterial and AMR gene diversity in these commodities and help guide pathogen control and AMR prevention efforts.IMPORTANCEWith the growing interest and application of metagenomics in understanding the structure/composition and function of diverse microbial communities along the One Health continuum, this study represents one of the first attempts to employ these advanced sequencing technologies to characterize the microbiota and AMR genes in animal food. We unraveled the effects of DNA extraction kits on sample analysis by 16S rRNA gene amplicon sequencing and showed similar efficacies of two strategies at removing chloroplast and mitochondrial reads. The in-depth analysis using shotgun metagenomics shed light on the community compositions and the presence of an array of AMR genes in animal food. This exploration of microbiomes and resistomes in representative animal food samples by both sequencing approaches laid important groundwork for future metagenomic investigations to gain a better understanding of the baseline/core microbiomes and associated AMR functions in these diverse commodities and help guide pathogen control and AMR prevention efforts.

Keywords: 16S rRNA gene amplicon sequencing; DNA extraction; animal food; microbiome; mock community; peptide nucleic acid; resistome; shotgun metagenomics.

Fig 1

(A) Relative abundance of bacterial orders identified using 16S rRNA gene amplicon sequencing in animal food samples processed with different DNA extraction kits in trail 1. Samples were classified as “unblocked” (U) or “blocked” (B) based on the addition of PNA to the PCR reaction mixture and “unfiltered” or “filtered” based on whether microbiomes were depleted of chloroplast and mitochondrial sequence reads bioinformatically. (B) PCoA based on Bray-Curtis distances for the animal food samples processed using four DNA extraction kits and analyzed by 16S rRNA gene amplicon sequencing. Samples were grouped based on their blocker and filtering status. Ellipses depict 95% confidence intervals for the unblocked and unfiltered (i), blocked and unfiltered (ii), blocked and filtered (iii), and unblocked and filtered (iv) samples. Gray lines connect samples originating from the same DNA extracts.

Fig 2

Box-and-whisker plots of (A) chloroplast and mitochondrial relative abundances, (B) Observed genera counts, (C) Simpson’s diversity indices, and (D) Pielou’s evenness measures for animal food microbial communities with and without PNA blockers during library preparation or with and without filtering of 16S rRNA gene amplicon sequencing data sets. Pairwise comparisons between group means were performed using the Wilcox rank-sum test (ANCOM analysis).

Fig 3

(A) Relative abundances of bacterial orders identified using 16S rRNA gene amplicon sequencing and shotgun metagenomics in animal food samples in trial 2. (B) PCoA based on Bray-Curtis distances for the animal food samples using the two sequencing approaches. Ellipses depict 95% confidence intervals for color-coded sample types.

Fig 4

Relative abundance of microbial reads successfully classified at different taxonomic ranks by the two sequencing approaches, 16S rRNA gene amplicon sequencing versus shotgun metagenomics. D – domain, P – phylum, C – class, O – order, F – family, G – genus, and S – species.