Amplification of uncultured single-stranded DNA viruses from rice paddy soil

Abstract

Viruses are known to be the most numerous biological entities in soil; however, little is known about their diversity in this environment. In order to explore the genetic diversity of soil viruses, we isolated viruses by centrifugation and sequential filtration before performing a metagenomic investigation. We adopted multiple-displacement amplification (MDA), an isothermal whole-genome amplification method with phi29 polymerase and random hexamers, to amplify viral DNA and construct clone libraries for metagenome sequencing. By the MDA method, the diversity of both single-stranded DNA (ssDNA) viruses and double-stranded DNA viruses could be investigated at the same time. On the contrary, by eliminating the denaturing step in the MDA reaction, only ssDNA viral diversity could be explored selectively. Irrespective of the denaturing step, more than 60% of the soil metagenome sequences did not show significant hits (E-value criterion, 0.001) with previously reported viral sequences. Those hits that were considered to be significant were also distantly related to known ssDNA viruses (average amino acid similarity, approximately 34%). Phylogenetic analysis showed that replication-related proteins (which were the most frequently detected proteins) related to those of ssDNA viruses obtained from the metagenomic sequences were diverse and novel. Putative circular genome components of ssDNA viruses that are unrelated to known viruses were assembled from the metagenomic sequences. In conclusion, ssDNA viral diversity in soil is more complex than previously thought. Soil is therefore a rich pool of previously unknown ssDNA viruses.

FIG. 1.

(A) Results of real-time PCR with specific primers. (B) Real-time PCR of C005 (left) and C112 (right). (A) Specific primers for C005 (lanes 1 and 2) and C112 (lanes 3 and 4) produced 438- and 380-bp amplicons, respectively. Viral DNA extracted from concentrated virus preparations (lanes 1 and 3) and DNA amplified by MDA without a denaturing step (lanes 2 and 4) were used as templates. Lane L, 100-bp ladder. (B) Circle a, DNA after MDA without the denaturing step (used for the RX library); circle b, DNA after MDA with the denaturing step (RH library); circle c, DNA before MDA. Similar concentrations of DNA (1.4, 1.3, and 1.0 ng/μl for circles a, b, and c, respectively) were used as templates. Nonlinear points indicated by red filled circles in the left-hand standard curve were not included in the standard curve. R² values are 0.982 (left) and 0.988 (right).

FIG. 2.

(A) Preferential amplification of circular DNA during MDA without the denaturing step. (a) Plasmids left uncut (circular DNA) and cut with the Eam1105I restriction enzyme (linear DNA). Lane 1, S11 and uncut DNA; lane 2, S11 and cut DNA; lane 3, S36 and uncut DNA; lane 4, S36 and cut DNA; lane L1, 1,000-bp ladder. (b) PCR products amplified from plasmids cut with Eam1105I. Specific primers for S11 (lane 5) and S36 (lane 6) produced 391- and 378-bp amplicons, respectively. Lane L1, 100-bp ladder. (B) Quantification of circular and linear DNAs during MDA without the denaturing step. A 0.037-ng/μl concentration of each circular or linear DNA was mixed with 10 ng/μl lambda DNA, and MDA was performed without the heating step. Concentrations of circular and linear DNAs were estimated by real-time PCR. Filled circles, circular S11; filled squares, circular S36; empty circles, linear S11; empty squares, S16. Circular DNA was amplified much more efficiently than linear DNA. x axis, time in minutes; y axis, DNA concentration in picograms per microliter.

FIG. 3.

Virus-like particles (A) and bacteria (B) from rice paddy soil stained with Sybr gold and observed by EFM. Scale bars, 5 μm. (A) No bacterial contamination is observed.

FIG. 4.

Phylogenetic trees of the amino acid sequences from the Rep_Viral (A) and Gemini_AL3 (B) domains obtained in this study and related members with those domains. The alignment lengths for tree construction were 75 and 108 amino acids for trees A and B, respectively. The scale bars indicate the distance score calculated with the scoring matrix BLOSUM62. The prefix C in the sequence name means that the sequence is from a contig, and the prefixes RH or RX indicates the library where the sequence was retrieved. (A) Reference sequences: CFDV, Coconut foliar decay virus (accession no. Q66005); SCSVa, -b, and -c, Subterranean clover stunt virus (accession no. Q87009, Q87013, and Q9ICP7); BBTVa, -b, and -c, Banana bunchy top virus (accession no. Q8QTK9, Q83026, and Q65378); PCV2, Porcine circovirus 2 (accession no. Q8BB16); GCV, Goose circovirus (accession no. Q8AYY2); CCV, Canary circovirus (accession no. Q912W1); CuCV, Columbid circovirus (accession no. Q91GA3); EHa, -b, and -c, Entamoeba histolytica HM-1 (accession no. XP_648754, XP_648748, XP_650115); CV, Canarypox virus (accession no. NP_955176); AYVV-aDNA1, Ageratum yellow vein virus-associated DNA 1 (accession no. Q8QME8); MVDVa and -b, Milk vetch dwarf virus; FBNYVa, -b, -c, Fava bean necrotic yellows virus (accession no. O91250, O91250, and Q66862); TCSV-aDNA1, Tobacco curly shoot virus-associated DNA 1 (accession no. Q7T7M5); GI, Giardia intestinalis (accession no. Q9NJY0); p4M, plasmid of Bifidobacterium pseudocatenulatum (accession no. NP_613078). (B) Reference sequences: CSMV, Chloris striate mosaic virus (accession no. P18921); PSV, Panicum streak virus (accession no. Q00338); MSV, Maize streak virus (accession no. P03568); WDV, Wheat dwarf virus (accession no. P06847); TYDV, Tobacco yellow dwarf virus (accession no. P31617); TYLCV, Tomato yellow leaf curl virus (accession no. P36279); PHYVV, Pepper huasteco yellow vein virus (accession no. Q06923); BGMV, Bean golden mosaic virus (accession no. P05175), ACMV, African cassava mosaic virus (accession no. P14982); SLCV, Squash leaf curl virus (accession no. P2904); BCTV, Beet curly top virus (accession no. P14991); pPASa and -b and pPAUa and -b, plasmids of Candidatus P. asterisasteris (accession no. Q2NIE5, Q2NIE4, Q0QLC1, and Q0QLC5). For the accession numbers of sequences from this study in the figures and alignments of the translated sequences, see the supplemental material.

FIG. 5.

Genome organizations of the three putative circular genomic components reconstructed from soil viral metagenomic sequences. The components show a genomic organization similar to that of Porcine circovirus2 (PCV2), a representative circovirus.