JVI Accepts, published online ahead of print on 11 …jvi.asm.org/content/early/2009/02/11/JVI.02301-08.full.pdf6 CGA CTC GTA ACA GG, 454-H CGT CCA GGC ACA ATC CAG TC, 454-I CCG 7

Metagenomic analyses of viruses in the stool of children with acute flaccid paralysis. 1

2

Joseph G Victoria1,2

, Amit Kapoor1,2

, Linlin Li1,2

, Olga Blinkova1,2

, Beth Slikas1,2

, 3

Chunlin Wang3, Asif Naeem

4, Sohail Zaidi

4, Eric Delwart

1,2 4

5

1 Blood Systems Research Institute, San Francisco, California 94118 6

2 Dept of Laboratory Medicine, University of California, San Francisco CA 94118 7

3 Stanford Genome Technology Center, Stanford, California 8

4 National Institute of Health, Pakistan, Department of Virology 9

10

Running title: Metagenomic analyses of viruses in stool of children. 11

12

Copyright © 2009, American Society for Microbiology and/or the Listed Authors/Institutions. All Rights Reserved.J. Virol. doi:10.1128/JVI.02301-08 JVI Accepts, published online ahead of print on 11 February 2009

on June 8, 2018 by guesthttp://jvi.asm

.org/D

ownloaded from

http://jvi.asm.org/

Abstract: 1

We analyzed viral nucleic acids in stool samples collected from 35 South Asian children 2

with non-polio acute flaccid paralysis (AFP). Sequence-independent reverse transcription 3

and PCR amplification of viral capsid-protected nuclease-resistant nucleic acids was 4

followed by DNA sequencing and sequence similarity searches to identify known and 5

new viruses. Limited Sanger sequencing (35-240 subclones per sample) identified an 6

average of 1.4 eukaryotic viruses per sample, while pyrosequencing yielded 2.6 viruses 7

per sample. In addition to bacteriophage and plant viruses, we detected known enteric 8

viruses including rotavirus, adenovirus, picobirnavirus, and human enterovirus (HEV) 9

species A-C, as well as numerous other members of the Picornaviridae family including 10

parechovirus, Aichi virus, rhinovirus, and human cardiovirus. The most divergent viral 11

sequences relative to reported viruses included members of a novel Picornaviridae genus 12

and four new viral species (members of the Dicistroviridae, Nodaviridae, Circoviridae 13

families and the Bocavirus genus). Six healthy contacts of AFP cases were similarly 14

analyzed and also contained numerous viruses, particularly HEV-C, including a 15

potentially novel Enterovirus genotype. Determining the prevalence and pathogenicity of 16

the novel genotypes, species, genera, and potential new viral families identified in this 17

study in different demographic groups will require further studies now facilitated by 18

knowledge of their genomes. 19


.org/D

ownloaded from

http://jvi.asm.org/

Introduction: 1

Metagenomic analyses of human gut or stool for bacterial (14, 21) and viral communities 2

(7, 8, 12, 40) have revealed a large degree of microbial diversity. The use of sequence-3

independent amplification of viral nucleic acids (3, 18, 34, 37) for viral metagenomics 4

avoids potential limitations of traditional methods, including the failure of virus to 5

replicate in cell cultures, failed PCR amplification or micro-array hybridization due to 6

high genetic divergence from known viruses, or failure of antibody to known viruses to 7

cross-react. The application of viral metagenomics may also be useful for the study of 8

diseases with unexplained etiologies possibly involving uncharacterized viruses or 9

combinatorial viral infections (12, 20). 10

Acute flaccid paralysis (AFP), characterized by rapid onset of an asymmetric 11

paralysis, can be caused by a variety of viral infections or co-infections (33). AFP in 12

children under 15 years old is currently tracked in countries with endemic poliovirus, 13

including Pakistan, India, Afghanistan and Nigeria as part of the Global Polio Laboratory 14

Network (1). Beside wild-type and revertant vaccine strains of polioviruses, several non-15

polio enteroviruses (NPEV) also have collectively been associated with up to a third of 16

AFP cases in children, including human enterovirus HEV-A serotype EV71 (6, 10, 15, 17

30, 31). In the US, AFP is also observed for 10-41% of the estimated <1% of persons 18

newly infected with West Nile virus exhibiting neurological symptoms (16, 29). Recent 19

studies indicate Chikungunya virus may also cause AFP in rare cases (32). 20

In this study we utilized sequence-independent amplification of partially purified 21

viral nucleic acids from the stool of South Asian children with non-polio AFP. All 22

samples had been tested previously in cell cultures and found to be negative for 23


.org/D

ownloaded from

http://jvi.asm.org/

poliovirus. Sequence data were obtained both by Sanger sequencing of sub-cloned DNA 1

and by 454 pyrosequencing. 2

3

Materials and Methods: 4

Viral particle purification. Stool samples were resuspended in Hank’s balanced salt 5

solution (Gibco BRL) and vortexed thoroughly. 500µl of supernatant from twice 6

repeated 15,000g centrifugation was filtered through a 0.45µm filter (Millipore) to 7

remove eukaryotic and bacterial cell sized particles. The filtrate was then treated with a 8

cocktail of DNase (Turbo DNAse from Ambion, Baseline zero from Epicentre, and 9

Benzonase from Novagen) and RNase (Fermentas) to digest non-particle protected 10

nucleic acid (19). Viral nucleic acids, protected from digestion within viral capsids, were 11

then extracted using the QIAmp Viral RNA extraction kit (Qiagen). 12

Sequence-independent amplification of viral nucleic acids 13

Viral cDNA synthesis of extracted viral RNA/DNA was performed as described 14

previously (37). Briefly, 100 pmol of a primer containing a fixed sequence followed by a 15

randomized octomer at the 3’-end was used in a reverse transcription reaction 16

(Superscript III, Invitrogen). A single round of DNA synthesis was then performed using 17

Klenow fragment polymerase (New England Biolabs) with an additional 50pmol of the 18

same primer. PCR amplification of nucleic acids was then performed using primers 19

consisting of the fixed portion of the random primers (37). Primers used for downstream 20

Sanger sequencing were based on primK (GAC CAT CTA GCG ACC TCC ACN NNN 21

NNN N) (34) or RAO-1 (GCC GGA GCT CTG CAG ATA TCN NNN NNN NNN) (18). 22

For the 10 samples submitted for 454 pyrosequencing the following primers were used in 23


.org/D

ownloaded from

http://jvi.asm.org/

the PCR reaction, and corresponding primers containing an additional 8-N at the 3’-end 1

were used for priming during reverse transcription: 454-A ATC GTC GTC GTA GGC 2

TGC TC, 454-B GTA TCG CTG GAC ACT GGA CC, 454-C CGC ATT GGT CGG 3

CAC TTG GT, 454-D CGT AGA TAA GCG GTC GGC TC, 454-E CAT CAC ATA 4

GGC GTC CGC TG, 454-F CGC AGG ACC TCT GAT ACA GG, 454-G CGC ACT 5

CGA CTC GTA ACA GG, 454-H CGT CCA GGC ACA ATC CAG TC, 454-I CCG 6

AGG TTC AAG CGA GGT TG, 454-J ACG GTG TGT TAC CGA CGT CC. 7

Amplification products were either subcloned into bacterial plasmids and (Sanger) 8

sequenced, or PCR products were sequenced directly using GS-FLX 454 pyrosequencing, 9

(Roche). Briefly, PCR fragments were treated as sonicated bacterial DNA for 454 10

pyrosequencing with polishing of the extremities of fragments and ligation of adaptors 11

prior to emulsion PCR as recommended by manufacturer (Roche). 12

Data assembly and processing. 13

For 454 sequencing, primer tag signatures on the random PCR primers were used as 14

identifier tags to assign sequences to the corresponding sample. Sequences were then 15

automatically trimmed of the fixed primer sequences plus 8 additional nucleotides at the 16

3’-end corresponding to random-N portion used in cDNA synthesis and Klenow 17

extension. Trimmed sequences were assembled into contigs using Sequencher 18

(Genecodes) software a criterion of 95% identity over 35bp or greater for 454 19

pyrosequenced products. Similarly, for cloned and sequenced products, primer sequences 20

were removed from plasmid inserts and assembled with a criterion of 85% identity over 21

50bp (Genecodes). The assembled contigs or singlet sequences which failed to assemble 22

were analyzed using BLASTn, BLASTx, and tBLASTx. Sequences with tBLASTx E-23


.org/D

ownloaded from

http://jvi.asm.org/

values lower than 10-3

were classified as either eukaryotic virus, bacteriophage, 1

eukaryotic, bacterial, or other based on the match with the best E-value. Those sequences 2

with tBLASTx, BLASTn and BLASTx E-values greater than 10-3

for the best hit were 3

deemed “unclassifiable”. High quality sequences and contigs have been deposited in 4

Genbank (Accession numbers XXXXXX through XXXXXX). 5

Sequencing of HEV VP1 region 6

Nested consensus primers were designed within the VP3 and 2A genes of HEV-C to 7

amplify the complete VP1 gene. Primers from 5’ to 3’-end are as follows: VP1_C_F1-8

GGBACNCAYRTNATHTGGGA, VP1_C_F2-GCNTGYMMNGAYTTYWSHGT, 9

VP1_C_R1-GGRCACCANVMNCKNACATG, VP1_C_R2- 10

GYTTDGGYTTCATGTACACTC. RT-PCR conditions were as follows: 10µl of 11

extracted viral RNA was incubated with 100pmol of random hexamer oligonucleotide 12

and 0.5mM of each deoxynucleoside triphosphate (dNTP) at 75ºC for 5min. 13

Subsequently, 40U of RNase inhibitor, 10mM dithiothreitol, 1X first-strand extension 14

buffer, and 200U of SuperScript III (Invitrogen) were added to the mixture and incubated 15

at 25ºC for 5 min, followed by 37ºC incubation for 1 hr. For PCR, 5µl of the reaction 16

described above was used in a total reaction volume of 50µl containing 2.5mM MgCl2, 17

0.2mM dNTPs, 1X manufacturer’s buffer, 0.2µM of each primer, and 5U of Taq 18

polymerase (New England Biolabs). For the first round, the PCR cycle included a 4 min 19

denaturation at 95°C; 30 cycles of 95°C for 45 s, 52°C for 1 min, 72°C for 1 min; a final 20

extension at 72°C for 7 min. 1µl of first round PCR product was used as template in a 21

second round amplification under identical conditions, with appropriate second round 22

primers. 23


.org/D

ownloaded from

http://jvi.asm.org/

Phylogenetic analysis 1

Reference HEV sequences were obtained from NCBI and edited for alignment using 2

GENEDOC software. Three representative sequences were used for each serotype, when 3

available. Sequence alignments were generated using the CLUSTAL_W package with 4

the default settings. Aligned sequences were trimmed to match the genomic region of the 5

sequences obtained in this study and used to generate phylogenetic trees in MEGA 4 6

using either neighbor-joining, maximum likelihood or maximum parsimony with 7

bootstrap values calculated from 1000 replicates. Accession numbers of reference 8

sequences used are as follows: HEV-A (AY177911, AY69748-AY69761); HEV-B 9

(AF029859, AF083069, AF85363, AF105342, AF114383, AF231763, AF233852, 10

AF241360, AF311939, AF317694, AF524867, AJ493062, AY167105, AY302539-11

AY302560, AY556057, AY556070); HEV-C; (AB205396, AF081308, AF081310, 12

AF205396, AF465511-AF465515, AF499635-AF499643, AF546702, AY876912, 13

AY876913); HEV-D (AY426531, DQ201177, EF107097). 14

Patient demographics and sample processing. All samples previously tested negative 15

for poliovirus by PCR of cytopathogenic cell culture supernatants in L20M and RD cell 16

lines (ATCC). Patient ages averaged 52 months with a range of 1 month to 14.5 years 17

old. The patients included 24 boys and 11 girls. Control samples were collected from 18

healthy contacts of children with AFP and are denoted with “C” in tables and figures, i.e. 19

5006C. 20

21

Results: 22

Sampling: 23


.org/D

ownloaded from

http://jvi.asm.org/

Sequence-independent viral nucleic acid amplification was performed on 35 fecal 1

samples from children diagnosed with AFP within two weeks prior to sampling. Clinical 2

outcomes of the AFP cases ranged from full recovery to death (Table 1). Viral particles 3

were partially purified and their nucleic acid randomly amplified and sub-cloned (see 4

materials and methods). Between 35 and 240 sub-cloned fragments were sequenced from 5

each sample. 454 pyrosequencing also was performed on a subset of 10 samples, yielding 6

between 3,715 and 25,516 high-quality sequences per sample. The average sequence 7

length for sub-cloned products was 432bp, while 454 pyrosequencing yielded an average 8

of 201bp. Unique sequences (singlets) or contigs were then analyzed by tBLASTx 9

against the GenBank nr database. Sequences with E-values ≤10-3

were categorized 10

according to their closest BLAST alignment. A total of 29% and 51% of sequences could 11

not be classified using Sanger and 454 pyrosequencing, respectively, a range similar to 12

levels from previous metagenomic studies of stool samples (7, 8, 12, 14, 21, 40). Both 13

Sanger and 454 pyrosequencing yielded ~23% eukaryotic viral sequences (Figure 1A, 14

1B). 15

Diversity of Viruses and Genetic Composition 16

Multiple previously characterized and divergent novel viruses were detected. The 17

majority of the known viral sequences were ssRNA viruses belonging to the order 18

Picornavirales. We also detected dsRNA (picobirnavirus and rotavirus), ssDNA (TTV 19

anelloviruses, adeno-associated virus), and dsDNA viruses (adenovirus) (Figure 1). 20

Highly divergent viruses with amino acid sequence identity less than 55% were denoted 21

as “–like” according to the closest tBLASTx match (e.g. circovirus-like or dicistrovirus-22

like). Viral detection using limited Sanger sequencing and 454 pyrosequencing were 23


.org/D

ownloaded from

http://jvi.asm.org/

compared in 10 samples (Table 1). Identification of co-infection increased from an 1

average of 1.4 eukaryotic viruses detected per sample (range 0-4) using Sanger 2

sequencing, to 2.6 (range 0-6) eukaryotic viruses for 454 pyrosequencing. The fraction 3

of each viral genome recovered increased from an average of 19% using Sanger 4

sequencing to 41% using 454 pyrosequencing (Fig. 2A, B). 454 pyrosequencing enabled 5

near complete genome sequencing of several HEV strains, a novel nodavirus-like virus 6

(in preparation), and a novel dicistrovirus-like virus (in preparation). For these near 7

complete genomes, the average depth of sequencing was 90-100X, however it was highly 8

variable (2X-378X) and diminished near regions of known complex secondary RNA 9

structure (Fig 2C, D). Regions with over-represented sequencing depth were often 10

preceded by stretches of sequence with 4/4 or 4/5 matches to the 3’-end of the primer 11

preceding the random “N-octamer” used in amplification (Fig 2 and data not shown). 12

Healthy contact control samples 13

Stool from 6 healthy children with family contact to an AFP case also were analyzed 14

using Sanger sequencing. Unexpectedly the fraction of viral/total sequences more than 15

doubled from 23% to 49% (Figure 1C). HEV-C composed the majority (60%) of the 16

viral clones sequenced and was observed in a higher frequency in healthy contacts (6/6) 17

compared to children with AFP (6/35) (p-value < 0.001). Different genotypes of HEV-C 18

were found, including coxsackievirus A14, A13, A24 and poliovirus (virus Sabin 2 19

vaccine strain). While several coxsackieviruses have been shown to correlate with 20

enteric diseases, CoxA14, CoxA13 and CoxA24 have yet to be associated with illness 21

(28). The poliovirus sequences identified here are likely derived from ingestion of oral 22

vaccine, as Pakistani children routinely received 10 or more doses. 23


.org/D

ownloaded from

http://jvi.asm.org/

Viral populations in each patient 1

Within each patient the percentages of viral and non-viral sequences varied greatly, 2

consistent with previously published data (12). Eukaryotic viruses were detected in the 3

stool of 29/35 (83%) AFP cases. Based on tBLASTx classification, sequences from 7 4

viral families were detected plus 4 novel virus groups. 17/29 of the virus positive cases 5

contained at least one human enterovirus (HEV), of which 5 samples appeared to be co-6

infected with two clearly distinguishable HEV species (Fig. 3, 4). Targeted amplification 7

from stool samples using pan-HEV primers (38) indicated 23/35 tested patients were 8

positive for HEV-infection 17 of which were also detected using viral metagenomics. 9

Other known enteric viruses observed included adenovirus, picobirnavirus, rotavirus, 10

Aichi virus, parechovirus, and rhinovirus, as well as assorted plant viruses of the genus 11

Partitiviridae and Tobamoviridae (Fig 3, 4, Table 1). In addition to previously identified 12

viruses, several potentially novel viruses were observed in multiple patients. These 13

viruses included one with weak identity (<35% amino acid identity) to the insect virus 14

family dicistroviridae (a member of the order Picornavirales) (detected in 3 AFP 15

patients), viruses belonging to a novel candidate picornavirus genus named Cosavirus 16

(present in 8) (19), a novel circovirus-like virus most closely related to porcine circovirus 17

(<55% amino acid identity, present in a single patient), a novel human bocavirus (<80% 18

amino acid identity, present in a single patient) (17) and a virus displaying weak amino 19

acid identity (<33%) to a fish nodavirus (present in a single patient) (Fig. 3, 4). 20

The classifications of sequences obtained in 10 individuals analyzed by both 21

Sanger and 454 pyrosequencing were compared (Fig. 4). In the majority of cases the 22

same viruses were found by both methods, although in 4/10 cases more viruses were 23


.org/D

ownloaded from

http://jvi.asm.org/

found using pyrosequencing (5192, 6178, 5550, 6572). The proportion of viral/non-viral 1

and the detection of specific viral species within the same sample was generally 2

comparable by both methods with some exceptions (Nodavirus-like/Cosaviruses ratio in 3

AF1449 and Adenoviruses/other viruses in 5550). 4

Potentially new, divergent HEV serotypes 5

All HEV singlet sequences or contigs were examined individually using BLASTx and 6

BLASTn similarity searches and by phylogenetic analysis (data not shown). Based on 7

these analyses, 3 samples were identified as containing the most divergent HEV, thereby 8

being candidates for new HEV serotypes. HEV variants showing the same antibody 9

neutralization profiles (i.e. belonging to the same serotype) have previously been shown 10

to encode VP1 with >88% amino acid or >75% nucleotide identity to other members of 11

the same serotype (24). Degenerate PCR primers flanking the VP1 gene were used to 12

amplify VP1 sequences in the three samples with divergent HEV. VP1 from sample 13

5034, 5044C, and 5048C exhibited 78%, 75%, and 75% nucleotide identity and 88%, 14

87%, and 83% amino acid identity, respectively, to the closest HEV sequence available in 15

GenBank. Sequences from 5034, a patient sample, and 5048C, a healthy contact sample, 16

exhibited 98% nucleotide identity to each other. 5034 and 5048C samples were collected 17

within a 3 month period of each other in 2007, both from the Punjab province of 18

Pakistan. Phylogenetic analyses of amplified VP1 nucleotide sequences show that 5034 19

and 5048C are deeply rooted CoxA24 serotypes, while 5044C may be divergent enough 20

from pre-existing genotypes to qualify as a prototype of a new HEV-C genotype (Fig. 5). 21

22

Discussion 23


.org/D

ownloaded from

http://jvi.asm.org/

In this study, we examined stool samples from 35 South Asian children with AFP 1

and 6 healthy contact cases for the presence of viral nucleic acids. Viruses were detected 2

in 29/35 stool samples from children with AFP and in all 6 healthy contacts. We also 3

report on the use of 454 pyrosequencing for viral metagenomics of human feces and a 4

comparison of this technique to traditional shotgun sub-cloning and sequencing. 5

Pyrosequencing provided superior genomic coverage (41% vs. 19% average coverage of 6

detected viral genomes) and more sensitive viral detection (average 2.6 vs. 1.4 viruses per 7

patient) (data not shown). Consistent with previous reports (39), the shorter reads 8

associated with pyrosequencing nearly doubled the number of “unclassifiable” sequences 9

when compared to Sanger sequencing (51% vs. 29%, respectively) (Fig. 1A, 1B). It 10

appears this increase in unclassifiable sequences is primarily due to misclassification of 11

diverse bacterial sequences, corresponding to a reduction of bacterial sequences from 12

28% by Sanger sequencing to 7.9% by 454/pyrosequencing, while viral sequences 13

accounted for ~23% of the total, regardless of sequencing method. Despite the problems 14

associated with shorter sequence reads, pyrosequencing was superior in both viral 15

detection and coverage within the 10 patients tested for approximately the same financial 16

cost. As pyrosequencing technology improves to generate longer sequence reads it 17

should therefore supplant Sanger shotgun sequencing as the preferred method of viral 18

discovery. 19

Prior viral metagenomic studies of feces utilized Sanger shotgun sequencing of 20

532 (8), 4,600 (12) and 36,769 plasmid subclones (40) at the cost of analyzing fewer 21

samples (1, 12, and 3, respectively). Two of these studies detected primarily plant 22

viruses (40) or bacteriophage (8) (in this case likely the results of focusing on dsDNA 23


.org/D

ownloaded from

http://jvi.asm.org/

viruses), while the third study, on diarrhea samples, detected known viral pathogens as 1

well as sequences divergent enough to potentially belong to two new viral species 2

(astrovirus and nodavirus) (12, 13). In our study, the most common plant virus detected 3

was pepper mild mottle virus, which also has been reported in high frequency in North 4

America and Singapore human stool samples (40). In addition to bacteriophage and plant 5

viruses, we also detected known pathogenic enteric viruses including rotavirus, 6

adenovirus, picobirnavirus, and numerous members of the Picornaviridae family 7

including parechovirus, Aichi virus, rhinovirus, cardioviruses, and species A-C of the 8

human Enterovirus genus, as well as several new viral species. The high proportion of 9

healthy children with viruses in their stools (6/6) (Table 1) underlines the often 10

asymptomatic nature of many enteric viral infections whose clinical outcomes are likely 11

dictated by a combination of viral and host genetics, active and passive (i.e. maternal 12

antibodies) immunity, and overall health (25). 13

Specific nested pan-enterovirus PCR primers detected HEV infection in 23/35 of 14

the AFP cases (19), while 17/35 AFP samples exhibited at least one HEV sequence using 15

viral metagenomics. Both metagenomic analysis and pan-HEV PCR detected HEV 16

infection in all 6 healthy contacts. This correlation was less pronounced for members of 17

the new candidate picornavirus genus (Cosaviruses) as these sequences were only found 18

in 9/41 AFP cases and healthy contacts by shotgun sequencing compared to 19/41 using 19

nested PCR (19). Similarly, human cardiovirus, initially detected in patients 5152 and 20

5003 was found in two additional patients (1449, 5003) upon specific screening with 21

nested PCR (see accompanying article). It is possible that the viral loads of cosaviruses 22

in stool is generally lower than that of HEV, thereby making detection less likely using 23


.org/D

ownloaded from

http://jvi.asm.org/

limited shotgun sequencing. Indeed, in depth 454 sequencing of 8,276 clones in patient 1

5550 and 25,516 clones in patient 6572 revealed the presence of cosaviruses missed by 2

Sanger sequencing. These results indicate that while a wide range of distinct viruses 3

(belonging to different and in some cases new viral species) can be detected using low-4

level Sanger subclone sequencing, the very high sensitivity of nested PCR stills allows 5

more cases of presumably low-level infections with known viruses to be detected. 6

We detected at least 5 novel viruses: a new human bocavirus (17), a new 7

Picornaviridae genus (19), a new circovirus (in preparation), new nodavirus (in 8

preparation), and new discistroviruses (in preparation). Sequences from divergent viruses 9

that may represent new genotypes of enteroviruses, parechovirus (22), cardioviruses (see 10

accompanying article), and picobirnaviruses (not shown) were also found. The viral 11

sequences comprising a novel nodavirus were clearly distinguishable from the nodavirus 12

sequences recently generated from diarrhea samples, overall exhibiting less than 41% 13

amino acid identity (12). The most diverse viral sequences detected belonged to the 14

dicistrovirus-like category in which polymerase and other enzymatic regions exhibited 15

less than 35% amino acid identity to dicistroviruses currently in GenBank. Dicistrovirus-16

like sequences were detected in three patients, two of which, 6178 and 6344, were co-17

infected with the members of the new Picornaviridae genus, Cosavirus. The 18

dicistrovirus-like sequences exhibited 70-75% nucleotide identity to each other, a level of 19

divergence seen between different species of dicistroviruses. 20

It remains to be determined which of these novel viruses are capable of replication 21

in the human gut, as it is conceivable that some were consumed and their nucleic acid 22

traveled through the digestive tract intact, as attested to by the detection of nucleic acids 23


.org/D

ownloaded from

http://jvi.asm.org/

from plant viruses which have also been shown to remain infectious (40). Nodaviruses 1

are small single-stranded, bipartite RNA viruses that to date have only been shown to 2

naturally infect insects and fish. Nodaviruses previously have been detected in human 3

stool (12) and are semi-permissive for replication in mammalian tissues (5, 11). 4

Similarly, dicistroviruses have been shown to replicate and be pathogenic in insects (9, 5

36). The internal ribosomal entry site between the two cistronic segments can act as a 6

powerful promoter in mammalian cells (27). However reports of viral replication within 7

mammalian cell lines are contradictory; one group has shown replication of a 8

dicistrovirus, Taura syndrome virus (TSV), in human cell lines (4), while another has 9

failed to reproduce TSV growth in mammalian cell lines (26). Unlike nodaviruses and 10

dicistrovirus, circovirus replication in mammals is well established for both the 11

pathogenic porcine circovirus 2, and the non-pathogenic porcine circovirus 1 (2, 23, 35). 12

Whether or not the circovirus seen in patient 5006 represents the first human circovirus, 13

or a circovirus from ingested meat remains to be determined. In vitro replication as well 14

as serological and larger epidemiological studies will be necessary to determine the 15

tropism and pathogenic potential of these novel viruses. 16

Three of the 35 AFP cases were fatal: child 5550 exhibited the highest level of co-17

infection in which 6 distinguishable viruses were observed (adenovirus, cosavirus, HEV-18

B, HEV-C, rhinovirus, and cucumber mosaic virus), patient 2296 was co-infected with a 19

HEV-B and a cosavirus, and patient 6178 exhibited co-infection with dicistrovirus and 20

cosavirus. Stool from patient 6178 was likely to contain a high titer of dicistrovirus 21

based on the large fraction of sequence from this virus derived by random amplification 22

(Fig. 2 B, D) and dilutional end-point PCR which indicated a viral load of greater than 23


.org/D

ownloaded from

http://jvi.asm.org/

106

copies per ml of stool supernatant (data not shown). While cosavirus was present in 1

all 3 fatal cases, the difference was not statistically significant when cosavirus prevalence 2

was compared in all AFP patients and healthy controls (19). Since even clearly 3

pathogenic picornaviruses, such as poliovirus, typically produce no clinical 4

manifestations in 99 to 99.9 % of infections (25), failure to detect a significant 5

association with disease in this small cohort does not absolve these viruses of possible 6

pathogenic roles. 7

In summary, we have used limited Sanger sequencing of stool samples from 8

children with AFP to detect both known and novel viruses. By increasing the depth of 9

the nucleic acid sampling using 454 pyrosequencing, we also were able to detect more 10

viruses likely present at lower viral loads in the samples. These studies provide a 11

framework for further studies that can be applied to numerous cases of AFP reported by 12

the Global Polio Laboratory Network (700,000 cases since 1997), in which only ~6.5% 13

cases have been attributed to polio and 15-30% to non-polio enteroviruses (1). PCR 14

studies of stool and tissues from different age and geographic groups, both with and 15

without disease symptoms, as well as serological tests will be required to determine the 16

epidemiology and pathogenicity of these new viruses. The presence of numerous known 17

and new viruses in stool samples from developing countries, a likely result of poor 18

sanitary conditions and frequent enteric infections, also indicate that such samples 19

provide readily accessible material for further viral discovery. 20

21

Acknowledgments: We thank Dr Michael P. Busch and Blood Systems Research 22

Institute for sustained support and NHLBI grant R01HL083254 to ELD. We also thank 23


.org/D

ownloaded from

http://jvi.asm.org/

Hope Biswas for statistical assistance and Shahzad Shaukat, Salmaan Sharif, Muhammad 1

Masroor Alam, and Mehar Angez for sample collection and polioviruses testing. 2

1. 2007. Laboratory surveillance for wild and vaccine-derived polioviruses--3

worldwide, January 2006-June 2007. MMWR Morb Mortal Wkly Rep. 56:965-9. 4

2. Allan, G. M., F. McNeilly, S. Kennedy, B. Daft, E. G. Clarke, J. A. Ellis, D. 5

M. Haines, B. M. Meehan, and B. M. Adair. 1998. Isolation of porcine 6

circovirus-like viruses from pigs with a wasting disease in the USA and Europe. J 7

Vet Diagn Invest. 10:3-10. 8

3. Allander, T., S. U. Emerson, R. E. Engle, R. H. Purcell, and J. Bukh. 2001. A 9

virus discovery method incorporating DNase treatment and its application to the 10

identification of two bovine parvovirus species. Proc Natl Acad Sci U S A. 11

98:11609-14. Epub 2001 Sep 18. 12

4. Audelo-del-Valle, J., O. Clement-Mellado, A. Magana-Hernandez, A. Flisser, 13

F. Montiel-Aguirre, and B. Briseno-Garcia. 2003. Infection of cultured human 14

and monkey cell lines with extract of penaeid shrimp infected with Taura 15

syndrome virus. Emerg Infect Dis. 9:265-6. 16

5. Ball, L. A. 1992. Cellular expression of a functional nodavirus RNA replicon 17

from vaccinia virus vectors. J Virol. 66:2335-45. 18

6. Bolanaki, E., C. Kottaridi, E. Dedepsidis, Z. Kyriakopoulou, V. Pliaka, A. 19

Pratti, S. Levidiotou-Stefanou, and P. Markoulatos. 2008. Direct extraction 20

and molecular characterization of enteroviruses genomes from human faecal 21

samples. Mol Cell Probes. 22:156-61. Epub 2008 Feb 14. 22

7. Breitbart, M., M. Haynes, S. Kelley, F. Angly, R. A. Edwards, B. Felts, J. M. 23

Mahaffy, J. Mueller, J. Nulton, S. Rayhawk, B. Rodriguez-Brito, P. Salamon, 24 and F. Rohwer. 2008. Viral diversity and dynamics in an infant gut. Res 25

Microbiol. 159:367-73. Epub 2008 May 1. 26

8. Breitbart, M., I. Hewson, B. Felts, J. M. Mahaffy, J. Nulton, P. Salamon, and 27

F. Rohwer. 2003. Metagenomic analyses of an uncultured viral community from 28

human feces. J Bacteriol. 185:6220-3. 29

9. Cox-Foster, D. L., S. Conlan, E. C. Holmes, G. Palacios, J. D. Evans, N. A. 30

Moran, P. L. Quan, T. Briese, M. Hornig, D. M. Geiser, V. Martinson, D. 31

vanEngelsdorp, A. L. Kalkstein, A. Drysdale, J. Hui, J. Zhai, L. Cui, S. K. 32 Hutchison, J. F. Simons, M. Egholm, J. S. Pettis, and W. I. Lipkin. 2007. A 33

metagenomic survey of microbes in honey bee colony collapse disorder. Science. 34

318:283-7. Epub 2007 Sep 6. 35

10. da Silva, E. E., M. T. Winkler, and M. A. Pallansch. 1996. Role of enterovirus 36

71 in acute flaccid paralysis after the eradication of poliovirus in Brazil. Emerg 37

Infect Dis. 2:231-3. 38

11. Delsert, C., N. Morin, and M. Comps. 1997. Fish nodavirus lytic cycle and 39

semipermissive expression in mammalian and fish cell cultures. J Virol. 71:5673-40

7. 41

12. Finkbeiner, S. R., A. F. Allred, P. I. Tarr, E. J. Klein, C. D. Kirkwood, and D. 42

Wang. 2008. Metagenomic analysis of human diarrhea: viral detection and 43

discovery. PLoS Pathog. 4:e1000011. 44


.org/D

ownloaded from

http://jvi.asm.org/

13. Finkbeiner, S. R., C. D. Kirkwood, and D. Wang. 2008. Complete genome 1

sequence of a highly divergent astrovirus isolated from a child with acute 2

diarrhea. Virol J. 5:117. 3

14. Gill, S. R., M. Pop, R. T. Deboy, P. B. Eckburg, P. J. Turnbaugh, B. S. 4

Samuel, J. I. Gordon, D. A. Relman, C. M. Fraser-Liggett, and K. E. Nelson. 5 2006. Metagenomic analysis of the human distal gut microbiome. Science. 6

312:1355-9. 7

15. Hayward, J. C., S. M. Gillespie, K. M. Kaplan, R. Packer, M. Pallansch, S. 8

Plotkin, and L. B. Schonberger. 1989. Outbreak of poliomyelitis-like paralysis 9

associated with enterovirus 71. Pediatr Infect Dis J. 8:611-6. 10

16. Jean, C. M., S. Honarmand, J. K. Louie, and C. A. Glaser. 2007. Risk factors 11

for West Nile virus neuroinvasive disease, California, 2005. Emerg Infect Dis. 12

13:1918-20. 13

17. Kapoor, A., E. Slikas, P. Simmonds, T. Chieochansin, A. Naeem, S. Shaukat, 14

M. M. Alam, S. Sharif, M. Angez, S. Zaidi, and E. Delwart. 2009. A Newly 15

Identified Bocavirus Species in Human Stool. J Infect Dis. 199:196-200. 16

18. Kapoor, A., J. Victoria, P. Simmonds, C. Wang, R. W. Shafer, R. Nims, O. 17

Nielsen, and E. Delwart. 2008. A highly divergent picornavirus in a marine 18

mammal. J Virol 82:311-320. 19

19. Kapoor, A., J. G. Victoria, P. Simmonds, T. Chieochansin, E. Slikas, A. 20

Naeem, S. Shaukat, S. Salmaan, A. M. Muhammad, M. Angez, C. Wang, R. 21 W. Shafer, Z. Sohail, and E. L. Delwart. 2009. A highly diversified and 22

prevalent new picornavirus genus in the stool of South Asian children. Proc Natl 23

Acad Sci U S A. In Press. 24

20. Kistler, A. L., D. R. Webster, S. Rouskin, V. Magrini, J. J. Credle, D. P. 25

Schnurr, H. A. Boushey, E. R. Mardis, H. Li, and J. L. DeRisi. 2007. Genome-26

wide diversity and selective pressure in the human rhinovirus. Virol J. 4:40. 27

21. Ley, R. E., M. Hamady, C. Lozupone, P. J. Turnbaugh, R. R. Ramey, J. S. 28

Bircher, M. L. Schlegel, T. A. Tucker, M. D. Schrenzel, R. Knight, and J. I. 29 Gordon. 2008. Evolution of mammals and their gut microbes. Science. 320:1647-30

51. Epub 2008 May 22. 31

22. Li, L., J. Victoria, A. Kapoor, A. Naeem, S. Shaukat, S. Sharif, M. Masroor, 32

M. Angez, S. Zaidi, and E. Delwart. 2009. Genomic characterization of a novel 33

human parechovirus type. Emerg Infect Dis. [Epub ahead of print]. 34

23. Morozov, I., T. Sirinarumitr, S. D. Sorden, P. G. Halbur, M. K. Morgan, K. 35

J. Yoon, and P. S. Paul. 1998. Detection of a novel strain of porcine circovirus in 36

pigs with postweaning multisystemic wasting syndrome. J Clin Microbiol. 37

36:2535-41. 38

24. Oberste, M. S., K. Maher, D. R. Kilpatrick, and M. A. Pallansch. 1999. 39

Molecular evolution of the human enteroviruses: correlation of serotype with VP1 40

sequence and application to picornavirus classification. J Virol. 73:1941-8. 41

25. Pallansch, M., and R. P. Roos. 2001. Enteroviruses: polioviruses, 42

coxsackieviruese, echoviruses, and newer enteroviruses. Lippincott Williams and 43

Wilkins, Philadelphia, PA. 44


.org/D

ownloaded from

http://jvi.asm.org/

26. Pantoja, C. R., S. A. Navarro, J. Naranjo, D. V. Lightner, and C. P. Gerba. 1

2004. Nonsusceptibility of primate cells to Taura syndrome virus. Emerg Infect 2

Dis. 10:2106-12. 3

27. Pfingsten, J. S., and J. S. Kieft. 2008. RNA structure-based ribosome 4

recruitment: lessons from the Dicistroviridae intergenic region IRESes. Rna. 5

14:1255-63. Epub 2008 May 30. 6

28. Pulli, T., P. Koskimies, and T. Hyypia. 1995. Molecular comparison of 7

coxsackie A virus serotypes. Virology. 212:30-8. 8

29. Saad, M., S. Youssef, D. Kirschke, M. Shubair, D. Haddadin, J. Myers, and 9

J. Moorman. 2005. Acute flaccid paralysis: the spectrum of a newly recognized 10

complication of West Nile virus infection. J Infect. 51:120-7. Epub 2004 Nov 6. 11

30. Saeed, M., S. Z. Zaidi, A. Naeem, M. Masroor, S. Sharif, S. Shaukat, M. 12

Angez, and A. Khan. 2007. Epidemiology and clinical findings associated with 13

enteroviral acute flaccid paralysis in Pakistan. BMC Infect Dis. 7:6. 14

31. Shoja, Z. O., H. Tabatabie, S. Shahmahmoudi, and R. Nategh. 2007. 15

Comparison of cell culture with RT-PCR for enterovirus detection in stool 16

specimens from patients with acute flaccid paralysis. J Clin Lab Anal. 21:232-6. 17

32. Singh, S. S., S. P. Manimunda, A. P. Sugunan, Sahina, and P. Vijayachari. 18

2008. Four cases of acute flaccid paralysis associated with chikungunya virus 19

infection. Epidemiol Infect. 136:1277-80. 20

33. Solomon, T., and H. Willison. 2003. Infectious causes of acute flaccid paralysis. 21

Curr Opin Infect Dis. 16:375-81. 22

34. Stang, A., K. Korn, O. Wildner, and K. Uberla. 2005. Characterization of virus 23

isolates by particle-associated nucleic acid PCR. J Clin Microbiol. 43:716-20. 24

35. Tischer, I., W. Mields, D. Wolff, M. Vagt, and W. Griem. 1986. Studies on 25

epidemiology and pathogenicity of porcine circovirus. Arch Virol. 91:271-6. 26

36. Van Munster, M., A. M. Dullemans, M. Verbeek, J. F. Van Den Heuvel, A. 27

Clerivet, and F. Van Der Wilk. 2002. Sequence analysis and genomic 28

organization of Aphid lethal paralysis virus: a new member of the family 29

Dicistroviridae. J Gen Virol. 83:3131-8. 30

37. Victoria, J. G., A. Kapoor, K. Dupuis, D. P. Schnurr, and E. L. Delwart. 31

2008. Rapid identification of known and new RNA viruses from animal tissues. 32

PLoS Pathog. 4:e1000163. 33

38. Welch, J. B., K. McGowan, B. Searle, J. Gillon, L. M. Jarvis, and P. 34

Simmonds. 2001. Detection of enterovirus viraemia in blood donors. Vox Sang. 35

80:211-5. 36

39. Wommack, K. E., J. Bhavsar, and J. Ravel. 2008. Metagenomics: read length 37

matters. Appl Environ Microbiol. 74:1453-63. Epub 2008 Jan 11. 38

40. Zhang, T., M. Breitbart, W. H. Lee, J. Q. Run, C. L. Wei, S. W. Soh, M. L. 39

Hibberd, E. T. Liu, F. Rohwer, and Y. Ruan. 2006. RNA viral community in 40

human feces: prevalence of plant pathogenic viruses. PLoS Biol. 4:e3. 41

42

43

44


.org/D

ownloaded from

http://jvi.asm.org/

FIGURE LEGENDS 1

2

Figure 1. Sequence classification and distribution. Left panel: Sequences were 3

classified as either eukaryotic, phage, bacterial, or viruses based on the best tBLASTx 4

score (E-value: <10-3

). Unclassified sequences are those which had E-values greater than 5

10-3

by tBLASTx, BLASTn, and BLASTx analysis. Sequences which did not fall into set 6

categories were denoted as “other” and included fungus and plasmid vector sequences. 7

Right panel: The subset of sequences classified as viral broken down by viral 8

classification. (A) Sanger sequencing clones for all 35 patients. (B) Sequences from a 9

subset of 10 patients generated by 454/pyrosequencing. (C) Sanger sequencing clones 10

from stool of 6 healthy contacts of AFP cases. 11

12

Figure 2. Genomic coverage and depth of sequencing. (A) Lines below the graphical 13

depiction of a generic picornavirus genome represent location of HEV-B, HEV-C, or 14

HCoSV (cosaviruses) singlets or contigs acquired from different AFP cases by either 15

Sanger sequencing (S) or 454 pyrosequencing (454). (B) Lines below the graphical 16

depiction of a generic dicistrovirus genome represent location of dicistrovirus singlets or 17

contigs acquired from two AFP cases by either Sanger sequencing or 454 18

pyrosequencing. (C) Depth of sequencing for each nucleotide position is shown for 454 19

pyrosequenced HEV-B in patient 5550 and (D) dicistrovirus in patient 6178. 20

21

Figure 3. Sequence classification and distribution per patient. Left: Sequences 22

generated by subcloning and Sanger sequencing from individual patients in which 23


.org/D

ownloaded from

http://jvi.asm.org/

eukaryotic viral sequences were detected categorized into bacterial (B), unclassified (U), 1

phage (P), eukaryotic (E), viral (V) or other (O). Right: Characterization of viral 2

sequence by viral family or species. Patient 5727 which had a single Tobacco green 3

mosaic virus sequence is not shown. Abbreviations: adeno-associated virus (AAV), 4

cucumber mosaic virus (CMV), pepper mottle virus (PepMoV). 5

6

Figure 4. Comparison of sequences obtained by Sanger and pyrosequencing. Direct 7

comparison of Sanger sequenced products (top) and 454 pyrosequencing products 8

(bottom). Pie charts are labeled in same manner as Figure 2. For patients 6278 and 6341 9

no recognized viral sequences were detected by either 454 pyrosequencing or Sanger 10

sequencing (data not shown). Abbreviations: adeno-associated virus (AAV), cucumber 11

mosaic virus (CMV), pepper mottle virus (PepMoV). 12

13

14

Figure 5. VP1 phylogenetic analysis samples containing divergent HEV sequences. 15

Unrooted neighbor-joining phylogenetic relationships based on alignment of VP1 16

nucleotide sequences for 5044C, 5034 and 5048C. Filled symbols represent VP1 17

sequences amplified from AFP patients or healthy contacts of AFP patients. Open 18

symbols represent the corresponding closest BLASTn match in GenBank. Collapsed 19

branches represent at least 3 representative sequences from HEV species (HEV-A, HEV-20

B, and HEV-D) or serotype (i.e. CoxA21). 21

22


.org/D

ownloaded from

http://jvi.asm.org/


.org/D

ownloaded from

http://jvi.asm.org/


.org/D

ownloaded from

http://jvi.asm.org/


.org/D

ownloaded from

http://jvi.asm.org/


.org/D

ownloaded from

http://jvi.asm.org/


.org/D

ownloaded from

http://jvi.asm.org/


.org/D

ownloaded from

http://jvi.asm.org/

Documents

JVI Accepts, published online ahead of print on 11 …jvi.asm.org/content/early/2009/02/11/JVI.02301-08.full.pdf6 CGA CTC GTA ACA GG, 454-H CGT CCA GGC ACA ATC CAG TC, 454-I CCG 7