Journal of General Virology (2011), 92, 1870–1879 DOI 10.1099/vir.0.030585-0 Discovery and initial analysis of novel viral genomes in the soybean cyst nematode Sadia Bekal,1 Leslie L. Domier,2 Terry L. Niblack1 and Kris N. Lambert1 Correspondence 1Department of Crop Sciences, University of Illinois, Urbana, IL 61810, USA Kris N. Lambert 2United States Department of Agriculture, Agricultural Research Service, Department of Crop [email protected] Sciences, University of Illinois, Urbana, IL 61810, USA Nematodes are the most abundant multicellular animals on earth, yet little is known about their natural viral pathogens. To date, only two nematode virus genomes have been reported. Consequently, nematode viruses have been overlooked as important biotic factors in the study of nematode ecology. Here, we show that one plant parasitic nematode species, Heterodera glycines, the soybean cyst nematode (SCN), harbours four different RNA viruses. The nematode virus genomes were discovered in the SCN transcriptome after high-throughput sequencing and assembly. All four viruses have negative-sense RNA genomes, and are distantly related to nyaviruses and bornaviruses, rhabdoviruses, bunyaviruses and tenuiviruses. Some members of these families replicate in and are vectored by insects, and can cause significant diseases in animals and plants. The novel viral sequences were detected in both eggs and the second juvenile stage of SCN, suggesting that these viruses are transmitted vertically. While there was no evidence of integration of viral sequences into the nematode genome, we indeed detected transcripts from these viruses by using quantitative PCR. These data are the first finding of virus genomes in parasitic nematodes. This discovery highlights the need for further exploration for nematode viruses in all tropic groups of these diverse and abundant animals, to determine how the Received 15 January 2011 presence of these viruses affects the fitness of the nematode, strategies of viral transmission and Accepted 11 April 2011 mechanisms of viral pathogenesis. INTRODUCTION Caenorhabditis briggsae and their genome sequenced. The discovery of natural nematode viruses supports the The use of genomic approaches for viral discovery has assertion that nematode viruses might be common but facilitated the rapid identification of unknown viruses overlooked (Fe´lix et al., 2011). from previously recalcitrant organisms and environments (Culley et al., 2006; Edwards & Rohwer, 2005). One group Other than C. elegans and C. briggsae, very little is known of organisms that has not been adequately explored for about nematode viruses. An iridovirus was found in the infectious viruses are the roundworms, forming the insect parasitic nematode (Thaumamermis cosgrovei) and phylum Nematoda. This is surprising since nematodes its isopod insect hosts Porcellio scaber and Armadillidium are among the most highly studied and numerous animals vulgare (Poinar et al., 1980). Another study described the on earth (Riddle et al., 1997; Telford et al., 2008; Wilson, infection of root-knot nematodes with a virus-like 2003). Plant parasitic nematodes are known to vector plant pathogen that caused visible and debilitating symptoms viruses, but these viruses do not replicate within their in nematodes; however no viral particles were visualized nematode vectors (Brown et al., 1995; Gray & Banerjee, (Loewenberg et al., 1959). In addition, a number of 1999). Several lines of evidence suggest that roundworms electron microscopic studies reported virus-like particles in are naturally infected by viruses. The model organism nematodes, but no biochemical or genomic analysis was Caenorhabditis elegans has been experimentally infected conducted (Foor, 1972; Poinar & Hess, 1977; Zuckerman with broad-host-range animal viruses, and such infections et al., 1973). This dearth of information on nematode- have been shown to be useful tools for dissecting the infecting viruses may be due to the difficulty in culturing or genetics of host–virus interactions (Liu et al., 2006; Lu collecting large numbers of infected parasitic nematodes et al., 2005; Shaham, 2006). Recently, two nodaviruses have required for traditional virus characterization. been identified in natural populations of C. elegans and The soybean cyst nematode (SCN), Heterodera glycines,isa microscopic sedentary obligate plant pathogen that feeds The GenBank/EMBL/DDBJ accession numbers for the novel SCN viral upon the roots of soybean and other closely related plants. sequences reported in this study are HM849038–HM849041. SCN is the most destructive pathogen of soybean in the 1870 030585 G 2011 SGM Printed in Great Britain RNA viruses in nematodes USA and has a complicated interaction with its host SCN tenuivirus (ScTV) (Table 1, Fig. 1). The Illumina (Wrather & Koenning, 2006). SCN hatches from its egg as cDNA sequences were derived from a paired-end library, a pre-parasitic second-stage juvenile (J2), which then therefore when the paired reads were mapped back on to migrates through the soil, finds a root, burrows into it and the SCN virus genomes the quality of the virus genome initiates feeding on vascular cells. The nematode completes assemblies could be assessed by noting the frequency of three more moults within the root to form both adult male correctly aligned paired cDNA sequences. The large and females. Determining whether a nematode is infected by number of SCN virus sequences (883- to 3310-fold a virus is a difficult task due to several factors: SCNs intra- coverage) produced continuous alignments of paired ends plant parasitic life cycle and microscopic size makes it showing no discontinuities in the alignments (data not difficult to observe or collect the parasitic stages free of plant shown), indicating the sequences were correctly assembled. material and in quantities sufficient for observation of The presence of long ORFs also supports the correct possible symptoms. Fortunately, new methods of high- assembly of the viral genomes (Table 2). throughput DNA sequencing are facilitating viral genome We designated the first virus genome detected in the SCN discovery. These methods produce immense amounts of transcriptome as ScNV after the top BLASTX result showed data, require relatively low amounts of starting material this genome was similar to the Midway virus (MIDWV) (Ansorge, 2009), and allow many potentially infected (Takahashi et al., 1982). MIDWV and Nyamanini virus individuals to be tested in a single experiment. Using these (NYMV) are closely related members of the proposed new methods, we describe the discovery of four nematode genus Nyavirus, and while related to viruses in the family viral genomes from SCN that were identified during a high- Bornaviridae, appear to be a distinct taxonic unit throughput sequencing of the SCN transcriptome. To our (Mihindukulasuriya et al., 2009). The assembled ScNV knowledge, this is the first description of nematode virus RNA genome sequence was 11 359 nt in length and had a genomes in a parasitic nematode. 1216-total fold coverage when the egg and J2 matching Illumina reads were combined, suggesting abundant RESULTS replication (Table 1). ScNV is predicted to have at least five ORFs (Fig. 1), which is the minimal number of ORFs needed for a virus in the order Mononegavirales. The Identification of virus genomes in the H. glycines arrangement of ORFs in the genomes of members of the transcriptome order is nucleoprotein (N), phosphoprotein (P), matrix Because of the importance of SCN as an agricultural protein (M), glycoprotein (G) and the RdRP large non- pathogen, a project was initiated to sequence the structural protein (L), with occasional additional ORFs + nematode’s transcriptome. Poly(A) RNA was extracted added upstream of the L-encoding ORF depending on the from eggs and surface-sterilized J2, and the resulting cDNA virus (Neumann et al., 2002). In ScNV the predicted amino was sequenced using 454 and Illumina sequencers. acid sequences of ORFs I, IV and V showed significant Subsequently, the cDNA was assembled into contigs, which similarity to N of Borna disease virus (BDV), and the G were compared to known proteins in the National Center and L proteins of MIDWV, respectively (Table 2). The for Biotechnology Information (NCBI) databases using predicted ORF IV amino acid sequence contains a puta- BLASTX (Altschul et al., 1990). Several large (.6 kb) contigs tive N-glycosylation site, amino-terminal signal peptide containing regions similar to viral RNA-dependent RNA sequence and a carboxyl-terminal membrane anchor polymerase (RdRP) enzymes were identified. Further sequence, which are consistent with ORF IVs encoding a bioinformatic investigation and assembly revealed nearly membrane-anchored G protein. ORF II and III showed no complete genomes of four different, likely ssRNA viruses significant similarity to proteins in the NCBI database. present in both egg and J2 stages: SCN nyavirus (ScNV), However, the predicted isoelectric point of the protein SCN rhabodovirus (ScRV), SCN phlebovirus (ScPV) and encoded by ORF II is 5.1 making it the most acidic protein Table 1. Properties of SCN-associated virus-like genomes and genome fragments SCN virus Most similar virus E value* CoverageD Length (nt) ScNV (HM849038d) Midway virus (MIDWV) 5e2160 548E, 713J, 1261T 11 359 ScRV (HM849039d) Northern cereal mosaic virus (NCMV) 3e2102 568E, 315J, 883T 12 698 ScPV (HM849040d) Uukuniemi virus (UUKV) 0 460E, 2850J, 3310T 6705§ ScTV (HM849041d) Rice stripe virus (RSV) 3e298 393E,