A Protocol for Extraction of Infective Viromes Suitable for Metagenomics Sequencing from Low Volume Fecal Samples

Abstract

The human gut microbiome (GM) plays an important role in human health and diseases. However, while substantial progress has been made in understanding the role of bacterial inhabitants of the gut, much less is known regarding the viral component of the GM. Bacteriophages (phages) are viruses attacking specific host bacteria and likely play important roles in shaping the GM. Although metagenomic approaches have led to the discoveries of many new viruses, they remain largely uncultured as their hosts have not been identified, which hampers our understanding of their biological roles. Existing protocols for isolation of viromes generally require relatively high input volumes and are generally more focused on extracting nucleic acids of good quality and purity for down-stream analysis, and less on purifying viruses with infective capacity. In this study, we report the development of an efficient protocol requiring low sample input yielding purified viromes containing phages that are still infective, which also are of sufficient purity for genome sequencing. We validated the method through spiking known phages followed by plaque assays, qPCR, and metagenomic sequencing. The protocol should facilitate the process of culturing novel viruses from the gut as well as large scale studies on gut viromes.

Keywords: T4; c2; human gut phageome; human gut virome; isolation; microbiome; phage; phi29; phiX174; purification.

Figure 1

Overview of the virome extraction, amplification, and sequencing procedures. A workflow for gut virome extraction and sequencing was established, the virome isolation part and sequencing part is in blue and green, respectively.

Figure 2

Infective phage recovery determined by plaque assays. The percentages of phages recovered (y axis) were determined by plaque assay at each different sampling point (x axis). The error bars here indicate the standard deviation of 3 replicates.

Figure 3

Taxonomic distribution (relative abundance) of the sequenced viromes. The relative distribution is described at the taxonomical level of orders. Taxonomy of contigs was determined by querying the viral contigs against a database containing taxon signature genes for virus orthologous group hosted at www.vogdb.org. The unassigned category is the contigs that have no relation to any known classified sequences.

Figure 4

(A) Distribution of sequencing reads into the different taxonomic categories viral, human, bacterial, and unknown origin. To check the presence of non-viral DNA sequences, 50,000 random forward reads were evaluated according to their match to a range of viral, bacterial, and human reference genome and protein databases as described in [17]. No reads (in 50,000 reads) matched to the 16S rRNA gene sequences in all the samples. (B) Relative abundance of sequencing reads matching the assembled virus-like contigs compared to the average of the three true samples. At least 10 times coverage/contig was applied here as the threshold for counting. Numbers 1–3refer to viromes extracted from feces from infants 1–3. ck1 and ck2 refer to co-extracted blank (SM buffer) samples.