Viruses are the most abundant and diverse genetic entities on Earth; however, broad surveys of viral diversity are hindered by the lack of a universal assay for viruses and the inability to sample a sufficient number of individual hosts. This study utilized vector-enabled metagenomics (VEM) to provide a snapshot of the diversity of DNA viruses present in three mosquito samples from San Diego, California. The majority of the sequences were novel, suggesting that the viral community in mosquitoes, as well as the animal and plant hosts they feed on, is highly diverse and largely uncharacterized. Each mosquito sample contained a distinct viral community. The mosquito viromes contained sequences related to a broad range of animal, plant, insect and bacterial viruses. Animal viruses identified included anelloviruses, circoviruses, herpesviruses, poxviruses, and papillomaviruses, which mosquitoes may have obtained from vertebrate hosts during blood feeding. Notably, sequences related to human papillomaviruses were identified in one of the mosquito samples. Sequences similar to plant viruses were identified in all mosquito viromes, which were potentially acquired through feeding on plant nectar. Numerous bacteriophages and insect viruses were also detected, including a novel densovirus likely infecting Culex erythrothorax. Through sampling insect vectors, VEM enables broad survey of viral diversity and has significantly increased our knowledge of the DNA viruses present in mosquitoes.
Citation: Ng TFF, Willner DL, Lim YW, Schmieder R, Chau B, Nilsson C, et al. (2011) Broad Surveys of DNA Viral Diversity Obtained through Viral Metagenomics of Mosquitoes. PLoS ONE 6(6): e20579. https://doi.org/10.1371/journal.pone.0020579
Editor: Jianming Qiu, University of Kansas Medical Center, United States of America
Received: April 27, 2011; Accepted: May 4, 2011; Published: June 6, 2011
Copyright: © 2011 Ng et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was funded through a grant to FR from the Canadian Institute for Advanced Research (CIFAR) and National Science Foundation grant DEB-1025915 to MB. TFFN was funded by the William and Elsie Knight Oceanographic Fellowship. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Broad surveys of natural viral diversity are technically challenging due to the inability to sample a sufficient number of individuals from different host species and the difficulty of characterizing previously undescribed viruses. An effective strategy for exploring viral diversity would need to simultaneously identify a wide range of viral types in a large number of individuals. Since female mosquitoes draw blood from a wide range of vertebrate hosts including humans, non-human primates, other mammals and birds , and also feed on plant nectar, they effectively sample numerous important viral reservoirs. Here we describe the use of metagenomics to investigate viruses found in insect vectors and the hosts they feed upon. This method, called vector-enabled metagenomics (VEM), combines the power of metagenomics for discovering novel viruses with the natural ability of insect vectors to integrate viral diversity over space, time, and many hosts .
To date, the majority of mosquito virus studies have focused on the detection of specific, well-described RNA arboviruses , , less is known about the diversity of DNA viruses in mosquitoes. Viruses present in mosquitoes can include viruses that are biologically or mechanically transmitted by these vectors, as well as other viruses that are not transmitted by mosquitoes but are drawn indiscriminately from host reservoirs. Characterizing new viruses is difficult due to limitations of current detection methods . Many viruses cannot be cultured in the laboratory, and methods such as degenerate PCR and pan-viral microarrays rely on the detection of highly conserved regions in known viral genomes for viral discovery . To circumvent these issues, recent studies have demonstrated the effectiveness of viral particle purification and shotgun sequencing (viral metagenomics) for describing novel viruses , . Using viral metagenomics, novel viruses have been characterized from nasopharyngeal aspirates , fecal samples , , blood , , and tissue samples such as lungs ,  and tumors . However, to date, no published studies have applied metagenomic sequencing to explore the diversity of viruses present in mosquitoes. In this study, we performed metagenomic sequencing on viruses purified from three mosquito samples from San Diego, California to provide a snapshot of the diversity of DNA viruses found in mosquitoes.
Results and Discussion
Novel and largely unexplored virus sequences in mosquitoes
By performing viral metagenomics on mosquitoes, this study aimed to broadly survey the viruses present in the many hosts that mosquitoes feed upon. For each of the three mosquito samples, viruses were purified and approximately half a million sequences were generated from purified viral DNA (Table S1). Based on the most significant tBLASTx similarities, the mosquito viromes contained sequences related to a wide range of animal, insect, plant, and bacterial viruses (Fig. 1B). Sequences with nucleotide-level identity to previously described viruses were limited to mosquito densoviruses, human papillomavirus 23 (HPV23), and a few phages (95–100% identities, Tables S3 and S4). The majority of the virome sequences were completely unknown and most recognizable viral sequences had only low levels of similarity to known viruses (32–70% amino acid identity), suggesting a highly novel and diverse viral community sampled by the mosquitoes.
A) Classification based on tBLASTx (E-value <0.001) against the Genbank non-redundant database. B) Breakdown of the viral sequences into four major categories: animal, plant, insect viruses (densoviruses and other insect viruses) and bacteriophages. Samples were obtained from 3 sites in San Diego: Buena Vista Lagoon (SD-BVL), River Bank (SD-RB), and Wild Animal Park (SD-WAP).
Unclassified sequences likely represent novel viruses
The majority of the sequences in all mosquito viromes were completely unidentifiable based on sequence similarity (>48%, Fig. 1A), which is consistent with other viral metagenomic studies , , , . This suggests that the reservoir of viruses in mosquitoes is novel and largely unexplored. Since sequencing was performed on purified virus particles, these divergent sequences likely originated from uncharacterized viral genomes. However, ongoing advancement of animal virus discovery can help elucidate the identities of these virus sequences found in mosquitoes. For example, when HPV112 was discovered from human skin in the past year , several contigs from this study that were previously classified as “unknown” were able to be recognized as papillomavirus sequences (see below). This example demonstrates that many of the unidentifiable sequences are likely represent novel viruses that are too divergent from known viruses to be recognized by sequence similarity searches. Increasing knowledge of animal and plant virus diversity has the potential to reveal the identities and hosts of these unknown viral sequences in the future.
Distinct viromes of the three mosquito samples
Each mosquito virome contained a different complement of viruses based on several analyses. First, BLAST searches (Fig. 2, Table S2, S3 and S4) showed different viral types in each sample. Second, distinct viromes were supported by specific PCR showing that viruses could generally only be amplified from the sample where they were originally identified (Table S2). Two exceptions to this trend were the densoviruses, which were present in all samples, and Mosquito VEM Anellovirus – SDWAP B, which was detected in both of the samples examined by PCR. Finally, cross-BLASTn analysis was used to determine the percent of sequences shared between the three metagenomes (Table S5). Mosquito SD-RB and SD-WAP had few common sequences (5%), while each shared slightly more sequences with SD-BVL (12% and 11% respectively). Most sequences shared between samples were related to mosquito densoviruses, while unknown sequences and sequences similar to other viral genomes were less likely to be shared. This suggests that each mosquito virome shared a small core component, largely composed of insect viruses infecting mosquitoes. The larger component of the mosquito virome, consisting of animal, plant and bacterial viruses, is more variable, and thus distinct between samples.
The family, host, and name of the most significant tBLASTx similarities in the Genbank non-redundant database are shown, with the colors representing the level of amino acid identity.
The distance between mosquito sampling sites is 30–50 km (Fig. S1), which is far greater than the average flight range for a host-seeking Culex erythrothorax mosquito , . The SD-WAP sample consisted exclusively of C. erythrothorax mosquitoes collected from an inland region of San Diego County in 2009, while both the SD-RB and SD-BVL mosquito samples consisted of mixed mosquito species drawn from coastal regions in 2006. Since mosquitoes draw blood within a radius of a few hundred meters, each sample likely contains viruses from animal hosts within mosquito's flight range. Although the three mosquito samples were of different species composition and were collected in different locations at different times, the distinct viromes demonstrate the diverse and heterogeneous nature of the viral community sampled by the mosquitoes.
Animal viruses identified in mosquitoes
Contiguous sequences (contigs) assembled from the mosquito viromes had similarities to five families of animal viruses, namely Anelloviridae, Circoviridae, Herpesviridae, Poxviridae and Papillomaviridae (Fig. 2), that infect a wide range of hosts including humans, primates, other mammals, and birds. Although other mosquito species can be specific in the hosts they feed on, Culex erythrothorax feed on a variety of mammals and birds opportunistically , allowing them to sample viruses from many different animal hosts. Although this is not an exhaustive investigation of total animal virus diversity, metagenomics performed on insect vectors with broad host ranges provides a way to elucidate a portion of the pan-animal virome.
Sequences related to novel and previously described papillomaviruses were identified in the Mosquito SD-RB virome. Numerous sequences had >95% nucleotide identity to human papillomavirus type 23 (HPV23) (Table S4). Comparison with the HPV23 genome revealed near-complete coverage from the metagenomic sequences (Fig. 3). Additionally, sequences related to human papillomavirus type 112 (HPV112) were identified. Mosquito VEM Papillomavirus - SDRB AE shared 91% nucleotide identity to the E1 gene of HPV112. Mosquito VEM Papillomavirus SDRB AF and AG (MosVemPapAG) shared only 76% and 71% nucleotide identities respectively to the minor capsid protein L2 gene of HPV112, and did not have any significant nucleotide identities to any other HPV types. Phylogenetic analysis based on this partial minor capsid protein region showed that MosVemPapAG is most closely related to HPV112, and belongs to the cutaneous gamma-papillomavirus genus (Fig. 4). Although we cannot confirm the host of MosVemPapAG, it groups phylogenetically with other human papillomaviruses. All papillomavirus sequences identified in the mosquitoes belonged to the cutaneous groups (beta and gamma groups), suggesting that mosquitoes may acquire papillomaviruses from the host's skin during feeding.
A) Human papillomavirus 23 (HPV23), B) Novel anelloviruses, C) Novel viruses with unique genome organization. Open reading frames are highlighted on the genome map and the amount of coverage from the metagenomic reads of the sample the virus was identified in is shown in the center.
Mosquito VEM Papillomavirus – SDRB AF shared similarity to a different region of capsid protein L2, but produced identical tree topography (data not shown). Papillomavirus sequences identified in mosquito virome SD-RB are indicated by arrows, all of which belong to the cutaneous groups.
Although more than 80% of normal human skin harbors papillomaviruses , human papillomaviruses have not previously been described in mosquitoes. This study is the first demonstration of sequences related to a human papillomavirus (HPV23) in mosquitoes, and also provides evidence that mosquitoes can harbor novel papillomaviruses. Papillomaviruses were only identified in one of the mosquito samples, suggesting that these viruses may only be present in mosquitoes sporadically. It was previously noticed that mosquitoes can transmit rabbit papillomavirus ; however, further research is needed to determine the prevalence and transmission potential of different types of human papillomaviruses in mosquitoes.
Anelloviruses and Circoviruses.
All anellovirus sequences identified in the viromes were novel (<70% amino acid identity to known anelloviruses, Fig. 2 and Table S2), suggesting that the animal hosts the mosquitoes feed on contain largely uncharacterized anellovirus diversity. The complete genomes of four putative viral genomes were sequenced (Fig. 3). Phylogenetic analysis based on the complete nucleotide sequence of ORF1 placed SD-BVL and SD-RB anelloviruses into the Torque teno virus (TTV) group, but forming a distinct genetic lineage from other TTV sequences (Fig. 5). Anellovirus genomes from an individual sample were closely related to each other, but unique genomes were found in different mosquito samples. In addition to the complete genomes, partial contigs with similarity to bovine TTV, human TTV, and human SEN virus were identified (Fig. 2 and Table S2).
Genbank accession numbers are shown in parentheses, and the hosts are indicated for any non-human sequences. The newly discovered anelloviruses from the mosquito viromes are indicated by arrows.
A diverse range of circoviruses was identified in the mosquito samples (Fig. 2 and Table S2). No significant pairwise nucleotide identity was shared between the replication genes of any of the circoviruses (data not shown), suggesting that these contigs represent distinct viral genomes. PCR assays for SD-BVL circovirus sequences were positive in the sample they originated from, but negative in the SD-WAP sample (Table S2), suggesting distinct circoviruses were present in each virome.
These results demonstrate that the pan-animal virome contains diverse and largely uncharacterized circoviruses and anelloviruses, which mosquitoes may routinely obtain from viremic hosts during blood feeding. Viruses belonging to the Circoviridae and Anelloviridae families contain small, circular, single-stranded DNA genomes, and are usually identified in blood , . Circoviruses are known to infect birds and pigs , and diverse circoviruses have been identified in aquatic environments , , , as well as in human, chimpanzee and bat feces , , , .The identification of sequences similar to avian circoviruses in mosquitoes is interesting because birds are the reservoir and secondary amplifying hosts of many mosquito-transmitted arboviruses, such as Eastern, Western, Japanese and St Louis equine encephalitis virus and West Nile virus . Anelloviruses are known to infect humans, non-human primates, domestic animals and marine mammals , , , , but the pathology of anelloviruses remains unknown , .
Herpesvirus-like and Poxvirus-like sequences.
In sample SD-RB, four contigs with amino-acid-level sequence similarities to herpesviruses and poxviruses were identified (Fig. 2 and Table S2). However, these contigs are only short portions of the genomes so it is impossible to determine more details about their identities or hosts.
Plant viruses identified in mosquitoes
Geminiviruses and Nanoviruses.
Sequences with similarities to plant viruses were consistently identified in the mosquito viromes (Fig. 2 and Table S2). All three viromes contained sequences related to geminiviruses, and sample SD-WAP had sequences related to nanoviruses. Mosquitoes are known to feed on plant nectar, indicating a potential source of these viruses. However, no plant viruses have been previously described in mosquitoes, so the ability of mosquitoes to transmit plant viruses still needs to be investigated through laboratory and field-based transmission studies. Other insect vectors that feed on plants, such as whiteflies, are known to transmit a diversity of plant viruses . In a related study using VEM to examine the viral community in whiteflies, almost all of the viral sequences shared high levels of nucleotide identity with previously described plant geminiviruses (Ng et al. in review). In contrast, the plant virus sequences in the mosquito viromes showed only weak amino acid level identities to known viruses (46%–53%; Fig. 2). These sequences from mosquitoes may represent extremely novel plant viruses, or could be part of recombinant genomes infecting other hosts.
Insect viruses identified in mosquitoes
Parvoviridae and Poxviridae.
A diverse range of insect viruses was identified in the mosquito viromes (Table S3 and S4). The majority of the sequences were similar to mosquito densoviruses (DNVs), specifically Aedes albopictus densovirus (AalDNV) and Haemogogus equinus densovirus (HeDNV). Since H. equinus and A. albopictus mosquitoes are not indigenous to San Diego, these sequences most likely represent densoviruses that infect the sampled mosquito species, primarily C. erythrothorax. Using PCR targeting the NS1 gene region, we further investigated the presence of densoviruses in the SD-WAP sample, which contained exclusively C. erythrothorax mosquitoes. The 720 base pair sequence of the PCR product (Accession #GU810839) was closely related (96% nucleotide identity) to HeDNV (Fig. 6). This sequence (VEM Culex erythrothorax densovirus; VEMCeDNV) most likely represents a new mosquito densovirus that infects C. erythrothorax mosquitoes.
The VEMCeDNV from the C. erythrothorax mosquitoes in sample SD-WAP is indicated with an arrow.
Densoviruses have been detected frequently in mosquito cell lines, and more rarely in wild-caught mosquitoes, where they are perpetuated through both horizontal and vertical transmission , , . Densovirus infection is highly lethal in cell lines and early stage larvae, however, infection at the late stages of larval development generally leads to a persistent and transmissible viremia , . Mosquito densoviruses are stable vectors for transformation of mosquitoes , , , which has created interest in using these viruses for mosquito and malaria control, either directly as lethal agents or as possible carriers of transgenes . Viral paratransgenesis takes advantage of the densoviruses to introduce genes that are lethal to mosquitoes or the pathogens that they carry. Viral paratransgenesis efforts can greatly benefit from the discovery of new densoviruses, such as those identified in this study. C. erythrothorax is the most common mosquito in San Diego County, and is suspected to be an emergent vector of West Nile Virus . Further studies of VEMCeDNV will be necessary to determine its efficacy as a biocontrol agent for C. erythrothorax.
Many other insect viruses were also identified, but none were found in all three samples (Table S3). Sample SD-BVL had the highest diversity of densoviruses, followed by SD-RB, then by SD-WAP (Table S3). Many sequences showed less than 85% amino acid identities to known densoviruses and unclassified insect viruses (Table S3), suggesting the presence of many novel insect viruses in mosquitoes.
Phages identified in mosquitoes
The mosquito viromes contained a large diversity of phage sequences (Table S3), including members from Myoviridae, Podoviridae, and Siphoviridae, as well as unclassified phages. Most of the phage sequences found in the mosquitoes only shared amino acid identities to known phages. However, in sample SD-RB, numerous sequences had 100% nucleotide identities to Propionibacterium phage PAD42 and PA6, Acyrthosiphon pisum secondary endosymbiont phage (APSE) 1–6, and Enterobacteria phage lambda (Table S3), suggesting that these known phages (or closely related phages) were present in the mosquito SD-RB virome. The three mosquito viromes differed in terms of the types of phages with BLAST similarities (Table S3), suggesting that each sample had a distinct phage content.
Phages identified in the mosquito viromes may infect the bacterial flora of the mosquito or that of the hosts they have fed upon. Propionibacterium acnes, the host for Propionibacterium phage PAD42 and PA6, is a commensal bacterium of human skin, so it is possible that mosquitoes acquire this bacterium and its phages during blood feeding . Sequences related to Enterobacteria phage might originate from the digestive system of mosquitoes and sequences with identities to phage APSE might infect endosymbiotic bacteria of mosquitoes. Phage APSE-1 through 6 infect Hamiltonella defensa, an endosymbiont of aphids and other sap-feeding insects that protects the aphids from wasp attack by killing the developing wasp larvae . Phage APSE-3 carries a toxin-encoding gene that provides the endosymbiont with defense against wasp larvae , and other APSE phages are also known to encode toxin genes , . Mosquitoes are not known to be hosts for parasitic wasps, but endosymbiotic bacteria such as Wolbachia are known to infect mosquitoes and interfere with the reproductive biology of their host through cytoplasmic incompatibility . Identification of sequences with high nucleotide identity to Phage APSE in this study suggests that mosquitoes potentially harbor other endosymbiotic bacteria and their phages, possibly to increase mosquito survivorship, or that of their eggs, through deterring predation.
This is the first study demonstrating the presence of a broad range of phages in mosquitoes, and distinct phage profiles for each mosquito virome. A diversity of phages with different inferred bacterial hosts was observed in the mosquitoes. Surveys based on the 16S rRNA gene support the notion of high bacterial diversity in mosquitoes , but to date, no metagenomic studies have examined the bacterial communities associated with mosquitoes. Future investigation of the role of bacteria and phages in mosquitoes is important, as they likely affect the host's physiology and fitness.
Viruses with unique genome organizations identified in mosquitoes
Two complete genomes were identified that contained features from different virus families. In sample SD-RB, genome Mosquito VEM CircoNanoGeminivirus - SDRB AJ showed a combination of features from the Circoviridae (animal virus), Nanoviridae (plant virus) and Geminiviridae (plant virus) families (Fig. 3). ORF2 of this virus showed similarity to the pfam viral replication protein 02407 superfamily, and had BLASTp hits to the replication protein from both the Circoviridae (29% amino acid identity to Porcine circovirus 2) and Nanoviridae (44% amino acid identity to Faba bean necrotic yellows virus) families. ORF1 of this virus showed similarity to the pfam geminivirus coat protein 00844 superfamily, and had 27% amino acid identity to the geminivirus Eragrostis curvula streak virus.
In sample SD-BVL, genome Mosquito VEM GeminiFungivirus - SDBVL G (Fig. 3) shared features from both the single-stranded DNA (ssDNA) plant virus Geminiviridae family and the single-stranded DNA (ssDNA) fungal virus, Sclerotinia sclerotiorum hypovirulence associated DNA virus (SSHADV1) . Protein conserved domain searches on ORF2 revealed similarity to the geminivirus replication catalytic domain pfam 00799, and a BLASTp search showed 28% amino acid identity to the geminivirus Tomato mottle leaf curl virus replication protein and 34% amino acid identity to the fungal virus SSHADV1. ORF1 showed 32% amino acid identity to the SSHADV1 coat protein. ORF3 of this virus had 41% amino acid identity to the replication-associated protein from the geminivirus Eragrostis curvula streak virus, and 62% amino acid identity to the replication-associated protein from SSHADV1.
VEM as an effective method for exploring viral diversity
Understanding total viral diversity is important for animal and human health. However, broad surveys of animal viral diversity is difficult due to the large number of individuals to sample, as well as the methodological limitations in characterizing novel viruses. The unique approach described here circumvents these issues by allowing the discovery of multiple novel viruses from many hosts in a single experiment. Through the use of VEM, this study created a baseline of the DNA virus community present in mosquitoes, shedding light on the high diversity of animal, plant, insect, and bacterial viruses that are present in this important vector. Although the discovery of a viral sequence does not always indicate active infection, the initial characterization will enable future studies to investigate viral prevalence and link viruses to hosts and disease symptoms. The application of this technique to mosquitoes from other regions, as well as other types of vectors, will greatly enhance our understanding of viral diversity.
VEM is a versatile technique that can be further refined to answer specific questions. Instead of purifying viruses from whole mosquitoes, VEM can be performed on dissected blood meals, surveying viruses specifically from the animal blood and plant nectar that the mosquitoes feed on and excluding viruses that may be present on the outside of the mosquitoes. Similarly, performing metagenomics on viruses purified from the dissected mosquito salivary glands, or mosquito saliva emitted during sugar feeding  can identify arboviruses with potential for transmission to animals. VEM can also be performed to characterize RNA viruses through the addition of a random-primed reverse transcription step . Finally, the multiple displacement amplification used in this study is known to preferentially amplify small ssDNA circular genomes , , , and the identifiable sequences in this study were dominated by mosquito densoviruses. To identify more double-stranded DNA (dsDNA) viruses, alternate amplification methods without this bias could be used, or ssDNA could be selectively removed from the mosquito samples by Mung Bean nuclease treatment.
In conclusion, this study utilized VEM to demonstrate the presence of a highly novel and diverse reservoir of animal, plant, insect, and bacterial viruses present in mosquitoes. The three different mosquito viromes contained distinct virus profiles, showing heterogeneity in the circulating viral community. By enabling broad surveys of viral diversity from many hosts, the VEM approach described here will be transformative for our understanding of the ecology of animal, plant, insect, and bacterial viruses.
Materials and Methods
Three mosquito samples (Table S1) were collected from San Diego County, CA, USA (Fig. S1) using an EVS CO2 trap baited with dry ice (BioQuip Products, Inc., Rancho Dominguez, CA, USA). SD-BVL and SD-RB mosquito samples were killed by freezing at −80°C, while the SD-WAP mosquitoes were anesthetized with triethylamine and stored at 4°C. Mosquitoes were homogenized in 5 ml of suspension medium (SM) buffer using a Tissumizer (Tekmar Control Systems, Inc., Vernon, Canada) at 5,000–8,000 rpm. Mosquito debris was pelleted by two rounds of centrifugation at 1,500 xg at 4°C for 30 min.
Viral particle purification and metagenomic sequencing
The protocol for viral particle concentration and purification was modified from previous studies , , , and an overview is shown in Table S6. Supernatants from the mosquito homogenates were filtered through a 0.45 µm syringe filter unit (Millipore, Billerica, MA) and viral particles were purified from the filtrate using a cesium chloride (CsCl) step gradient. The purified viral concentrate was examined by epifluorescence microscopy to verify the presence of viral particles, and ensure the absence of contaminating bacterial and eukaryotic cells . The viral fraction was concentrated and washed twice with sterile SM buffer on a Microcon 30 column (Millipore), followed by treatment with 0.2 volumes of chloroform for 10 minutes, then incubation with 2.5 U DNase I per µl for 3 hours at 37°C.
Total DNA was extracted using a CTAB/Formamide protocol . Extracted viral DNA was amplified using Genomiphi for 1.5 hours (GE Healthcare, Piscataway, NJ) based on the manufacturer's instructions. Following amplification, samples SD-BVL and SD-RB were sequenced with 454 GS20 pyrosequencing and sample SD-WAP was sequenced using 454 GS FLX technology. Longer read length in the SD-WAP virome resulted in an increased proportion of sequences with known identities compared to the other two viromes. The NCBI genome project numbers for the three viromes are 28413, 28467, and 49713.
Metagenomic sequences were filtered using PRINSEQ  to remove short reads, and were compared to the GenBank non-redundant database using BLASTn and tBLASTx , . Each sequence was assigned top-level taxonomy (Eukarya, Bacteria, Archaea, or Virus) based on its closest BLAST similarity (E-value <10−3 (Fig. 1A). Sequences with a best BLAST similarity to cellular genomes over ≥50 nt with an E-value <10−3 were removed prior to further classification. Remaining sequences were further classified based on best BLAST similarities to viral genomes (Fig. 1B). A cross-BLAST approach  was used to evaluate the similarity between mosquito metagenomes, where all sequences with BLASTn E-value <10−5 and ≥98% identity were considered shared.
Metagenomic sequences were assembled into contigs using the SeqMan Pro-assembler (DNASTAR, Madison, WI) with match size = 35, minimum match percentage = 95%, match spacing = 15, maximum mismatch end bases = 0. Contigs were compared to the non-redundant database using tBLASTx (E-value <0.001 ,  and contigs representing complete genomes were manually analyzed using SeqBuilder (DNASTAR). For complete genomes, open reading frames (ORFs) were analyzed and annotated using Artemis  and BLASTn and BLASTp were performed to determine identity. The HPV sequences were also assembled to the HPV23 genome (Genbank Accession # U31781) as a reference using Sequencher 4.7 (Gene Codes, Ann Arbor, MI). Densovirus sequences were assembled into contigs using the 454 GS De Novo Assembler (Branford, CT). Viral contigs were deposited to Genbank under the accession number HQ335010-HQ335087.
Nucleic acids from Mosquito SD-BVL and SD-WAP were amplified by Genomiphi (GE Healthcare), then subjected to PCR with primers designed based on selected contigs from the viromes (Table S7). Sample Mosquito SD-RB was unavailable for PCR testing. PCR primers were designed to amplify large regions of the contigs, encompassing many individual metagenomic sequence reads. PCR products were sequenced to verify the accuracy of the assemblies.
Alignments were performed using ClustalW multiple alignment in Bioedit , and MEGA4 was used for phylogenetic analysis with a neighbor joining method and bootstrap with 1000 replications . Phylogenetic analysis for the papillomaviruses was based on the partial alignment of the minor capsid protein L2 sequence with representative HPVs. The phylogenetic analysis of the anelloviruses was based on alignment of the ORF1 nucleotide sequence with major anellovirus groups , . For the densoviruses, the phylogenetic analysis was based on the nucleotide alignment of the partial NS1 gene PCR sequence with representative densoviruses .
Locations of mosquito samples, produced using Google Earth (http://earth.google.com/). Samples were obtained from 3 sites in San Diego: Buena Vista Lagoon (SD-BVL), River Bank (SD-RB), Wild Animal Park (SD-WAP).
Contigs and genomes with significant tBLASTx similarities to known vertebrate and plant viruses. The virus name, family and host of the most significant tBLASTx sequence in Genbank are shown. Complete genomes are indicated with asterisks (*). For PCR results, “Y” indicates that the contig was detected in that specific sample and “N” indicates that it was not.
Analysis of the contigs with amino acid identities (tBLASTx, evalue<0.001) to bacteriophages and insect viruses.
BLASTn analysis of the contigs with nucleotide identities to members of Papillomaviridae and Parvoviridae.
Cross-BLASTn analysis of the mosquito metagenomes. A sequence was considered to be shared by two metagenomes if each sequence was the best BLASTn similarity for the other when the two metagenomes were compared with BLASTn.
Overview of the VEM methodology for obtaining viral metagenomes from mosquitoes.
We thank the DNA sequencing team of Genome Institute of Singapore for providing DNA sequencing support, as well as Bhakti Dwivedi and Hong Liu for assistance with bioinformatics.
Conceived and designed the experiments: TFFN DLW BC FR MB. Performed the experiments: TFFN DLW YWL BC SA. Analyzed the data: TFFN DLW YWL RS CN YR FR MB. Wrote the paper: TFFN DLW YWL FW MB.
- 1. Molaei G, Andreadis TA, Armstrong PM, Anderson JF, Vossbrinck CR (2006) Host feeding patterns of Culex mosquitoes and West Nile virus transmission, northeastern United States. Emerging Infectious Diseases 12: 468–474.
- 2. Ng TFF, Bixby E, Vallad G, Duffy S, Polston JE, et al. (2011) Exploring the Diversity of Plant DNA Viruses and Their Satellites Using Vector-Enabled Metagenomics on Whiteflies. PLoS ONE. (in Press).
- 3. Huang C, Slater B, Campbell W, Howard J, White D (2001) Detection of arboviral RNA directly from mosquito homogenates by reverse-transcription-polymerase chain reaction. Journal of Virological Methods 94: 121–128.
- 4. Kuno G (1998) Universal diagnostic RT-PCR protocol for arboviruses. Journal of Virological Methods 72: 27–41.
- 5. Delwart EL (2007) Viral metagenomics. Reviews in Medical Virology 17: 115–131.
- 6. Wang D, Coscoy L, Zylberberg M, Avila PC, Boushey HA, et al. (2002) Microarray-based detection and genotyping of viral pathogens. Proceedings of the National Academy of Sciences of the United States of America 99: 15687–15692.
- 7. Edwards RA, Rohwer F (2005) Viral metagenomics. Nature Reviews Microbiology 3: 504–510.
- 8. Allander T, Tammi MT, Eriksson M, Bjerkner A, Tiveljung-Lindell A, et al. (2005) Cloning of a human parvovirus by molecular screening of respiratory tract samples. Proceedings of the National Academy of Sciences of the United States of America 102: 12891–12896.
- 9. Breitbart M, Hewson I, Felts B, Mahaffy JM, Nulton J, et al. (2003) Metagenomic analyses of an uncultured viral community from human feces. Journal of Bacteriology 185: 6220–6223.
- 10. Victoria JG, Kapoor A, Li LL, Blinkova O, Slikas B, et al. (2009) Metagenomic analyses of viruses in stool samples from children with acute flaccid paralysis. Journal of Virology 83: 4642–4651.
- 11. Jones MS, Kapoor A, Lukashov VV, Simmonds P, Hecht F, et al. (2005) New DNA viruses identified in patients with acute viral infection syndrome. Journal of Virology 79: 8230–8236.
- 12. Breitbart M, Rohwer F (2005) Method for discovering novel DNA viruses in blood using viral particle selection and shotgun sequencing. Biotechniques 39: 729–736.
- 13. Ng TFF, Suedmeyer WK, Gulland F, Wheeler E, Breitbart M (2009) Novel anellovirus discovered from a mortality event of captive California sea lions. Journal of General Virology 90: 1256 -1261:
- 14. Willner D, Furlan M, Haynes M, Schmieder R, Angly FE, et al. (2009) Metagenomic analysis of respiratory tract DNA viral communities in cystic fibrosis and non-cystic fibrosis individuals. Plos One 4:
- 15. Ng TFF, Manire C, Borrowman K, Langer T, Ehrhart L, et al. (2009) Discovery of a novel single-stranded DNA virus from a sea turtle fibropapilloma by using viral metagenomics. Journal of Virology 83: 2500–2509.
- 16. Rosario K, Nilsson C, Lim YW, Ruan YJ, Breitbart M (2009) Metagenomic analysis of viruses in reclaimed water. Environmental Microbiology 11: 2806–2820.
- 17. Ekström J, Forslund O, Dillner J (2010) Three novel papillomaviruses (HPV109, HPV112 and HPV114) and their presence in cutaneous and mucosal samples. Virology 397: 331–336.
- 18. Tietze NS, Stephenson MF, Sidhom NT, Binding PL (2003) Mark-recapture of Culex erythrothorax in Santa Cruz County, California. Journal of the American Mosquito Control Association 19: 134–138.
- 19. Walton WE, Workman PD, Tempelis CH (1999) Dispersal, survivorship, and host selection of Culex erythrothorax (Diptera : Culicidae) associated with a constructed wetland in southern California. Journal of Medical Entomology 36: 30–40.
- 20. Antonsson A, Forslund O, Ekberg H, Sterner G, Hansson BG (2000) The ubiquity and impressive genomic diversity of human skin papillomaviruses suggest a commensalic nature of these viruses. Journal of Virology 74: 11636–11641.
- 21. Dalmat HT (1958) Arthropod transmission of rabbit papillomatosis. Journal of Experimental Medicine. 108.
- 22. Biagini P, Pierre G, Cantaloube JF, Attoui H, de Micco P, et al. (2006) Distribution and genetic analysis of TTV and TTMV major phylogenetic groups in French blood donors. Journal of Medical Virology 78: 298–304.
- 23. Shibata I, Okuda Y, Yazawa S, Ono M, Sasaki T, et al. (2003) PCR detection of Porcine circovirus type 2 DNA in whole blood, serum, oropharyngeal swab, nasal swab, and feces from experimentally infected pigs and field cases. Journal of Veterinary Medical Science 65: 405–408.
- 24. Todd D (2000) Circoviruses: immunosuppressive threats to avian species: a review. Avian Pathology 29: 373–394.
- 25. Lopez-Bueno A, Tamames J, Velazquez D, Moya A, Quesada A, et al. (2009) High diversity of the viral community from an Antarctic lake. Science 326: 858–861.
- 26. Rosario K, Duffy S, Breitbart M (2009) Diverse circovirus-like genome architectures revealed by environmental metagenomics. Journal of General Virology 90: 2418–2424.
- 27. Li L, Victoria JG, Wang C, Jones M, Fellers GM, et al. (2010) Bat guano virome: predominance of dietary viruses from insects and plants plus novel mammalian viruses. Journal of Virology 84: 6955–6965.
- 28. Li L, Kapoor A, Slikas B, Oderinde BS, Wang C, et al. (2010) Multiple diverse circoviruses infect farm animals and are commonly found in human and chimpanzee feces. Journal of virology 91: 74–86.
- 29. Blinkova O, Rosario K, Li L, Kapoor A, Slikas B, et al. (2009) Frequent detection of highly diverse variants of cardiovirus, cosavirus, bocavirus, and circovirus in sewage samples collected in the United States. Journal of Clinical Microbiology 47: 3507–3513.
- 30. Blinkova O, Victoria J, Li Y, Keele BF, Sanz C, et al. (2010) Novel circular DNA viruses in stool samples of wild-living chimpanzees. J Gen Virol 91: 74–86.
- 31. Weaver SC, Barrett ADT (2004) Transmission cycles, host range, evolution and emergence of arboviral disease. Nature Reviews Microbiology 2: 789–801.
- 32. Biagini P, Uch R, Belhouchet M, Attoui H, Cantaloube JF, et al. (2007) Circular genomes related to anelloviruses identified in human and animal samples by using a combined rolling-circle amplification/sequence-independent single primer amplification approach. Journal of General Virology 88: 2696–2701.
- 33. Hino S, Miyata H (2007) Torque teno virus (TTV): current status. Reviews in Medical Virology 17: 45–57.
- 34. Leary TP, Erker JC, Chalmers ML, Desai SM, Mushahwar IK (1999) Improved detection systems for TT virus reveal high prevalence in humans, non-human primates and farm animals. Journal of General Virology 80: 2115–2120.
- 35. Davidson I, Shulman LM (2008) Unraveling the puzzle of human anellovirus infections by comparison with avian infections with the chicken anemia virus. Virus Research 137: 1–15.
- 36. Jones DR (2003) Plant viruses transmitted by whiteflies. European Journal of Plant Pathology 109: 195–219.
- 37. Oneill SL, Kittayapong P, Braig HR, Andreadis TG, Gonzalez JP, et al. (1995) Insect densoviruses may be widespread in mosquito cell-lines. Journal of General Virology 76: 2067–2074.
- 38. Ren XX, Hoiczyk E, Rasgon JL (2008) Viral paratransgenesis in the malaria vector Anopheles gambiae. Plos Pathogens 4:
- 39. Barreau C, Jousset FX, Bergoin M (1997) Venereal and vertical transmission of the Aedes albopictus parvovirus in Aedes aegypti mosquitoes. American Journal of Tropical Medicine and Hygiene 57: 126–131.
- 40. Carlson J, Suchman E, Buchatsky L (2006) Densoviruses for control and genetic manipulation of mosquitoes. Insect Viruses: Biotechnological Applications 68: 361–392.
- 41. Rwegoshora RT, Kittayapong P (2004) Pathogenicity and infectivity of the Thai-strain densovirus (AThDNV) in Anopheles minimus s.l. Southeast Asian Journal of Tropical Medicine and Public Health 35: 630–634.
- 42. Cheng LP, Chen SX, Zhou ZH, Zhang JQ (2007) Structure comparisons of Aedes albopictus densovirus with other parvoviruses. Science in China Series C-Life Sciences 50: 70–74.
- 43. Lood R, Morgelin M, Holmberg A, Rasmussen M, Collin M (2008) Inducible Siphoviruses in superficial and deep tissue isolates of Propionibacterium acnes. Bmc Microbiology 8:
- 44. Van der Wilk F, Dullemans AM, Verbeek M, van den Heuvel J (1999) Isolation and characterization of APSE-1, a bacteriophage infecting the secondary endosymbiont of Acyrthosiphon pisum. Virology 262: 104–113.
- 45. Oliver KM, Degnan PH, Hunter MS, Moran NA (2009) Bacteriophages encode factors required for protection in a symbiotic mutualism. Science 325: 992–994.
- 46. Moran NA, Degnan PH, Santos SR, Dunbar HE, Ochman H (2005) The players in a mutualistic symbiosis: Insects, bacteria, viruses, and virulence genes. Proceedings of the National Academy of Sciences of the United States of America 102: 16919–16926.
- 47. Degnan PH, Moran NA (2008) Diverse phage-encoded toxins in a protective insect endosymbiont. Applied and Environmental Microbiology 74: 6782–6791.
- 48. Yen JH, Barr AR (1971) New hypothesis of the cause of cytoplasmic incompatibility in Culex pipiens L. Nature 232: 657–658.
- 49. Pidiyar VJ, Jangid K, Patole MS, Shouche YS (2004) Studies on cultured and uncultured microbiota of wild Culex quinquefasciatus mosquito midgut based on 16s ribosomal RNA gene analysis. American Journal of Tropical Medicine and Hygiene 70: 597–603.
- 50. Yu X, Li B, Fu Y, Jiang D, Ghabrial SA, et al. (2010) A geminivirus-related DNA mycovirus that confers hypovirulence to a plant pathogenic fungus. Proceedings of the National Academy of Sciences of the United States of America 107: 8387–8392.
- 51. Hall-Mendelin S, Ritchie SA, Johansen CA, Zborowski P, Cortis G, et al. (2010) Exploiting mosquito sugar feeding to detect mosquito-borne pathogens. Proceedings of the National Academy of Sciences of the United States of America 107: 11255–11259.
- 52. Zhang T, Breitbart M, Lee WH, Run JQ, Wei CL, et al. (2006) RNA viral community in human feces: prevalence of plant pathogenic viruses. Plos Biology 4: 108–118.
- 53. Pinard R, de Winter A, Sarkis GJ, Gerstein MB, Tartaro KR, et al. (2006) Assessment of whole genome amplification-induced bias through high-throughput, massively parallel whole genome sequencing. Bmc Genomics 7: 21.
- 54. Kim K-H, Chang H-W, Nam Y-D, Roh SW, Kim M-S, et al. (2008) Amplification of uncultured single-stranded DNA viruses from rice paddy soil. Appl Environ Microbiol 74: 5975–5985.
- 55. Haible D, Kober S, Jeske H (2006) Rolling circle amplification revolutionizes diagnosis and genomics of geminiviruses. Journal of Virological Methods 135: 9–16.
- 56. Thurber RV, Haynes M, Breitbart M, Wegley L, Rohwer F (2009) Laboratory procedures to generate viral metagenomes. Nature Protocols 4: 470–483.
- 57. Patel A, Noble RT, Steele JA, Schwalbach MS, Hewson I, et al. (2007) Virus and prokaryote enumeration from planktonic aquatic environments by epifluorescence microscopy with SYBR Green I. Nature Protocols 2: 269–276.
- 58. Sambrook J, Russell DW (2001) Molecular cloning: A laboratory manual. New York: Cold Spring Harbor Laboratory Press.
- 59. Schmieder R, Edwards R (2011) Quality control and preprocessing of metagenomic datasets. Bioinformatics.
- 60. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990) Basic local alignment search tool. Journal of Molecular Biology 215: 403–410.
- 61. Altschul SF, Madden TL, Schaffer AA, Zhang JH, Zhang Z, et al. (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Research 25: 3389–3402.
- 62. Rutherford K, Parkhill J, Crook J, Horsnell T, Rice P, et al. (2000) Artemis: sequence visualization and annotation. Bioinformatics 16: 944–945.
- 63. Hall TA (1999) BioEdit:a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucl Acids Symp Ser 41: 95–98.
- 64. Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4: Molecular evolutionary genetics analysis (MEGA) software version 4.0. Molecular Biology and Evolution 24: 1596–1599.
- 65. Zhai YG, Lv XJ, Sun XH, Fu SH, Gong ZD, et al. (2008) Isolation and characterization of the full coding sequence of a novel densovirus from the mosquito Culex pipiens pallens. Journal of General Virology 89: 195–199.