Evaluation of Host Depletion and Extraction Methods for Shotgun Metagenomic Analysis of Bovine Vaginal Samples

Abstract

The reproductive tract metagenome plays a significant role in the various reproductive system functions, including reproductive cycles, health, and fertility. One of the major challenges in bovine vaginal metagenome studies is host DNA contamination, which limits the sequencing capacity for metagenomic content and reduces the accuracy of untargeted shotgun metagenomic profiling. This is the first study comparing the effectiveness of different host depletion and DNA extraction methods for bovine vaginal metagenomic samples. The host depletion methods evaluated were slow centrifugation (Soft-spin), NEBNext Microbiome DNA Enrichment kit (NEBNext), and propidium monoazide (PMA) treatment, while the extraction methods were DNeasy Blood and Tissue extraction (DNeasy) and QIAamp DNA Microbiome extraction (QIAamp). Soft-spin and QIAamp were the most effective host depletion method and extraction methods, respectively, in reducing the number of cattle genomic content in bovine vaginal samples. The reduced host-to-microbe ratio in the extracted DNA increased the sequencing depth for microbial reads in untargeted shotgun sequencing. Bovine vaginal samples extracted with QIAamp presented taxonomical profiles which closely resembled the mock microbial composition, especially for the recovery of Gram-positive bacteria. Additionally, samples extracted with QIAamp presented extensive functional profiles with deep coverage. Overall, a combination of Soft-spin and QIAamp provided the most robust representation of the vaginal microbial community in cattle while minimizing host DNA contamination. IMPORTANCE In addition to the host tissue collected during the sampling process, bovine vaginal samples are saturated with large amounts of extracellular DNA and secreted proteins that are essential for physiological purposes, including the reproductive cycle and immune defense. Due to the high host-to-microbe genome ratio, which hampers the sequencing efficacy for metagenome samples and the recovery of the actual metagenomic profiles, bovine vaginal samples cannot benefit from the full potential of shotgun sequencing. This is the first investigation on the most effective host depletion and extraction methods for bovine vaginal metagenomic samples. This study demonstrated an effective combination of host depletion and extraction methods, which harvested higher percentages of 16S rRNA genes and microbial reads, which subsequently led to a taxonomical profile that resembled the actual community and a functional profile with deeper coverage. A representative metagenomic profile is essential for investigating the role of the bovine vaginal metagenome for both reproductive function and susceptibility to infections.

Keywords: cattle; metagenomics; microbiome; shotgun; vagina; veterinary microbiology.

FIG 1

Experimental design. Three sample types were used to examine the efficacies of depletion and extraction methods. Vaginal swabs were vaginal swabs without spiked bacteria. The mock sample contained vaginal samples and five bacterial suspensions. Positive-control samples contained only the five bacterial suspensions. The samples were treated with different depletion and extraction methods to obtain the nucleic acids for shotgun sequencing and subsequent metagenomic analysis.

FIG 2

The relationship between the percentage of 16S rRNA genes and the percentage of microbial reads identified in the bovine vaginal samples. The markers represent all the samples involved in this study (n = 36), including vaginal swabs, mock samples, and positive-controls. Samples were treated with different depletion methods, including None (circle), NEBNext (triangle), PMA (square), and Soft-spin (cross). Red represents samples extracted using DNeasy and blue represents samples extracted using QIAamp.

FIG 3

Alpha (A) and Beta (B) diversity of the samples involved in this study (n = 36), including vaginal swab, mock sample, and positive-control. Samples were treated with different depletion methods, including None (circle), NEBNext (triangle), PMA (square), and Soft-spin (cross). Red represents samples extracted using DNeasy and blue represents samples extracted using QIAamp. ANOVA test was applied to examine the significance of differences between the alpha diversity of samples treated with different depletion methods while the t test was applied to examine the significance of differences between the alpha diversity of samples processed with different extraction methods. Beta diversity was represented by principal coordinate analysis ordination of the Bray-Curtis dissimilarity matrix.

FIG 4

Abundances of spiked bacteria, including Aliarcobacter cryaerophilus, Campylobacter fetus, Enterococcus faecalis, Pseudomonas aeruginosa, and Staphylococcus aureus, in the samples involved in this study (n = 36), including vaginal swabs, mock samples, and positive-controls.

FIG 5

Number of functional annotations identified by the different functional and pathway databases. A t test was applied to examine the significance of differences between the number of functional annotations in samples processed with different extraction methods.