Hydroxyapatite-mediated separation of double-stranded DNA, single-stranded DNA, and RNA genomes from natural viral assemblages

Abstract

Metagenomics can be used to determine the diversity of complex, often unculturable, viral communities with various nucleic acid compositions. Here, we report the use of hydroxyapatite chromatography to efficiently fractionate double-stranded DNA (dsDNA), single-stranded DNA (ssDNA), dsRNA, and ssRNA genomes from known bacteriophages. Linker-amplified shotgun libraries were constructed to generate sequencing reads from each hydroxyapatite fraction. Greater than 90% of the reads displayed significant similarity to the expected genomes at the nucleotide level. These methods were applied to marine viruses collected from the Chesapeake Bay and the Dry Tortugas National Park. Isolated nucleic acids were fractionated using hydroxyapatite chromatography followed by linker-amplified shotgun library construction and sequencing. Taxonomic analysis demonstrated that the majority of environmental sequences, regardless of their source nucleic acid, were most similar to dsDNA viruses, reflecting the bias of viral metagenomic sequence databases.

FIG. 1.

Flow diagram depicting HAP-mediated separation of known and environmental viral nucleic acids.

FIG. 2.

Conversion of ssDNA genomes to linear dsDNA. Single-stranded DNA templates are converted to linear dsDNA molecules by sequential treatment with E. coli DNA polymerase I in the presence of random hexamers.

FIG. 3.

Separation of known viral nucleic acids using HAP chromatography. Ten nanograms of M13mp18 ssDNA (lane 2), phi6 dsRNA (lane 3), MS2 ssRNA (lane 4), and lambda dsDNA (lane 5), respectively, were combined (lane 6) and visualized, along with 1-kb Plus DNA Ladder (lanes 1 and 7) on a 0.8% E-gel. One hundred nanograms of each nucleic acid were combined, loaded onto a HAP column, and eluted using 0.12 M (lane 8), 0.18 M (lane 9), and 0.40 M/1.0 M (lane 10) phosphate buffer. Lanes 8 to 10 show 1/10th (∼10 ng) of the HAP-purified nucleic acids.

FIG. 4.

Separation of viral-community nucleic acids from a subsurface hypoxic region of the Chesapeake Bay. Viral nucleic acids purified from the Chesapeake Bay (Collection CBAY1 is shown) were fractionated on a HAP column using 0.12 M, 0.20 M, 0.40 M, and 1.0 M phosphate buffer and visualized with 1 μl of High Mass DNA Ladder (L1) and 1-kb Plus DNA Ladder (L2).

FIG. 5.

Neighbor-joining phylogenetic tree produced by APIS. An example of a Chesapeake Bay ssDNA viral query sequence (depicted in red) clading with sequences from other dsDNA environmental viral metagenomes. The viral metagenome called UDEL_CHESAPEAKE_VIRIOPLANKTON_SMPL was generated from a separate study (3). The viral metagenome called SDSU_SCRIPPS_PIER_MARINE_VIRUS_SMPL was generated by a study conducted by Breitbart et al. (7). Bootstrap values are indicated at branches. The scale bar indicates 0.1 amino acid substitution per position.