|Home | About | Journals | Submit | Contact Us | Français|
Chordopoxviruses of the subfamily Chordopoxvirinae, family Poxviridae, infect vertebrates and consist of at least eight genera with broad host ranges. For most chordopoxviruses, the number of viral genes and their relative order are highly conserved in the central region. The GC content of chordopoxvirus genomes, however, evolved into two distinct types: those with genome GC content of more than 60% and those with a content of less than 40% GC. Two standard PCR assays were developed to identify chordopoxviruses based on whether the target virus has a low or high GC content. In design of the assays, the genus Avipoxvirus, which encodes major rearrangements of gene clusters, was excluded. These pan-pox assays amplify DNA from more than 150 different isolates and strains, including from primary clinical materials, from all seven targeted genera of chordopoxviruses and four unclassified new poxvirus species. The pan-pox assays represent an important advance for the screening and diagnosis of human and animal poxvirus infections, and the technology used is accessible to many laboratories worldwide.
The poxviruses (family Poxviridae) cause illness characterized by generalized or localized cutaneous lesions, and most member viruses have broad host ranges. The overall broad host range of this family is demonstrated by the two subfamilies of the Poxviridae. The subfamily Entomopoxvirinae infects insects, and the subfamily Chordopoxvirinae infects vertebrates; the latter consists of eight genera, and other “unclassified chordopoxviruses.” The classified genera are Orthopoxvirus, Parapoxvirus, Avipoxvirus, Capripoxvirus, Leporipoxvirus, Suipoxvirus, Molluscipoxvirus, and Yatapoxvirus.
Poxvirus infections of humans, cattle, sheep, goats, companion animals, birds, and zoo animals have been reported worldwide (2) but in general represent an underappreciated cause of health care utilization. Within the United States, until recently the international classification of diseases (ICD) codes were not available for many of these virus infections. Poxvirus infections can be clinically confused with other cutaneous disease, and other poxviruses are emerging or reemerging infections in various parts of the world. The majority of human poxvirus infections are zoonotic. Poxvirus infections are also a significant burden to agricultural communities. Capripoxvirus infections of ruminants cause significant morbidity and mortality. Parapoxvirus infections of ruminants and their handlers are endemic in the United States and worldwide; infected humans often turn to veterinarians for diagnostic assistance (14). Human monkeypox (an orthopoxvirus) is an emerging infection with smallpox-like characteristics that was introduced into the United States in 2003 via global commerce in the exotic pet animal trade. The evolution of monkeypox virus and its two major clades in Africa represents an incompletely understood emerging infectious risk. Smallpox, caused by variola virus, an eradicated disease, remains a significant biothreat agent of international concern and the subject of World Health Organization-approved research activities. A recent study on health care utilization for the other nonzoonotic human poxvirus infections (21) began to define the burden (22/10,000 health care visits) of molluscum contagiosum infections in the community. Over the past 5 years, an increasing number of potential poxvirus therapies have been studied. Better diagnostics targeting poxviruses as the cause of cutaneous infections will decrease the inappropriate treatment of these infections. For instance, parapoxvirus infections are often misdiagnosed as cutaneous anthrax, which unnecessarily contributes to overuse of antibacterial agents. By using effective therapies and prevention measures, after appropriate diagnosis, the interhuman transmissibility of poxvirus infections will be reduced, and days lost from work will be reduced.
Poxviruses represent one of largest viruses known and replicate in the cytoplasm of the infected cell and encode most enzymes for their life cycle. The genomes of poxviruses are a linear double-stranded DNA genome in the range of 134 kb (Parapoxvirus) to 330 kb (Avipoxvirus) and encode more than 130 genes. The central region, comprised of nearly 100 genes which encode viral gene expression, DNA replication, and virion formation, has a structural arrangement that is conserved in most chordopoxviruses. Between genera of chordopoxviruses, host specificity and genome sequence have diverged. Interestingly, the chordopoxviruses have two distinct types of genome based on GC content: parapoxviruses, molluscipoxviruses, and crocodilepox virus (an unclassified poxvirus most similar to molluscipoxvirus) have high GC contents (>60%); the genomes of the other six genera of chordopoxvirus have a low GC (30 to 40%) content. The evolutionary factors for this divergence are unknown.
The preponderance of data suggests that although the poxviruses can readily recombine under tissue culture conditions (23), the genomes are quite stable in evolutionary time (15). To develop a PCR assay, or assays, which could easily be used to screen for the presence of a poxvirus in a clinical sample, we developed two PCR signatures, one that would be expected to amplify nucleic acid from poxviruses with high GC content (high-GC PCR) while the other would be expected to amplify nucleic acid from most poxviruses with low GC content (low-GC PCR).
In this study, we describe the design and validation of the two new pan-chordopoxvirus standard PCR assays using over 150 chordopoxvirus isolates. We also describe the use of these new PCR assays to diagnose and discover previously unknown poxviruses as the cause of infections based on the sequence information from the resultant PCR amplicons.
The origins of the viruses and methods for preparing DNA from purified virions or infected cell cultures are described elsewhere (5-7, 13, 16, 18, 20, 22). DNA from the CDC poxvirus collection was used for assay development. In the initial validation of the low-GC PCR assay, we used 10 Eurasian or Old World orthopoxviruses (OPV) including a monkeypox virus (MPXV_US03_39), a cowpox virus (CPXV_BRT), a variola virus (VARV_BSH75), a laboratory vaccinia virus (VACV_WR), a rabbit poxvirus (RPXV), an ectromelia virus (ETCV_MOS), a camelpox virus (CMLV_V78903), a taterapoxvirus (TATV_DAH68), and two recent vaccinia virus human clinical samples (VACV_07065 and VACV_07070) (1); three North American OPV including a racoonpox virus (RACV_MD61), a volepox virus (VPXV_CA85), and a skunkpox virus (SKPV_WA78); two leporipoxviruses, myxoma virus (MYXV_Lau), rabbit fibroma virus (RFV_KAS); two yatapoxviruses, Yaba monkey tumor virus (YMTV_V83) and tanapox virus (TANV_04) (3); three unclassified poxviruses (Cotia virus from South America [8, 17], the clinical isolate NY_014 from New York state, and a North American deer poxvirus DPXV_V89); and a nucleic acid sample extracted from a fox squirrel body lesion (POX_08040).
The development of the high-GC PCR assay utilized nucleic acid extracted from clinical specimens of parapoxviruses including two orf viruses (ORFV) from Missouri (ORFV_06044 and ORFV_08041); four pseudocowpox viruses (PCPV) from Missouri (PCPV_06025), West Virginia (PCPV_08024), and Bangladesh (PCPV_07012 and PCPV 07013); two bovine papular stomatitis viruses (BPSV) from Bangladesh (BPSV_07005) and Washington (BPSV_07058); and molluscum contagiosum virus (MOCV) (MOCV_08031).
A multiple sequence alignment program MAFFT (11) was used for the alignment of poxvirus genome sequences. The conserved sequences from multiple genome sequence alignments were studied manually for the design of pan-poxvirus PCR primers. The sequence editing and alignment from high- and low-GC PCR amplicons used the DNASTAR Lasergene, version 8, software package (DNASTAR, Inc, Madison, WI) and BioEdit software (http://www.mbio.ncsu.edu/BioEdit/bioedit.html).
To infer the phylogeny of clinical poxvirus specimens, the Bayesian analysis software package BEAST, BEAUti, and Tracer (4) were used. The analyses run a Markov chain Monte Carlo (MCMC) chain length of 5,000,000 to 10,000,000 (until an EES value of >100) using the HKY (Hasegawa-Kishino-Yano) nucleotide substitution model with gamma (categories 4) distribution and a proportion of invariant sites with a fixed mean substitution rate at 1.0 and sampling of every 1,000 states.
Viral genomes from genera with high-GC content were aligned separately from those with low-GC content group. The primer sequences were selected based on the genome sequence alignment results obtained with strains of low-GC content poxviruses (Table (Table1).1). The conserved sequences were screened and analyzed by considering the sequence specificity, annealing temperature, and potential secondary structure formation with the assistance of Primer Express (version 1.5; Applied Biosystems) software. The selected forward primer was 5′-ACA CCA AAA ACT CAT ATA ACT TCT (insulin metalloproteinase-like protein gene), and the reverse primer was CCT ATT TTA CTC CTT AGT AAA TGA T (intracellular mature virion [IMV] membrane protein gene). The amplicon size from the low-GC PCR assay is about 220 bp (bp). The annealing temperature for the PCR assay is 50°C as the length of conserved sequences is restricted.
The primer design for the high-GC PCR assay was based on the available genome sequences of poxvirus strains with high GC content. The selection criterion used for the primer selection was similar to that of the low-GC PCR assay. The selected forward primer was CAT CCC CAA GGA GAC CAA CGA G, and the reverse primer was TCC TCG TCG CCG TCG AAG TC (both primers target an RNA polymerase subunit gene). The amplicon size from the high-GC PCR assay is about 630 bp. The PCR assay has a predicted annealing temperature of 65°C.
The validation PCR assays were set up as follows. PCR mixtures contained ~10 to 100 ng of viral DNA and a 20 μM primer pair in 50 μl of a solution of 50 mM Tris-HCl buffer (pH 9.2); 16 mM (NH4)2SO4; 2.25 mM MgCl2; 2% (vol/vol) dimethyl sulfoxide; 0.1% (vol/vol) detergent Tween 20; 350 μM (each) dATP, dCTP, dGTP, and dTTP; and 2 units of the DNA polymerases Taq and Pwo provided in the Expand Long Template PCR Kit (Roche Molecular Biologicals, Indianapolis, IN). Using a mixture of proofreading DNA polymerases reduces the potential sequence error from PCR amplification. For PCR amplification, we used a Model 9700 thermocycler (Perkin-Elmer Cetus, Boston, MA) programmed as follows: after 2 min at 92°C for DNA denaturing, reaction mixtures were thermocycled 10 times through successive denaturing (92°C for 10 s), annealing (50°C or 65°C for 30 s), and elongation (68°C for 1 min) steps and then through 20 cycles of denaturing, annealing, and elongation in which each successive elongation step added 2 s. Four microliters was inspected for the amplicon size by electrophoresis in a 4% E-gel (Invitrogen-Novex, Carlsbad, CA) run for 30 min (preloaded ethidium bromide). PCR products were stored at 4°C.
A four-base-cutter restriction endonuclease, TaqI (5′-TCGA-3′; New England Biolabs, Beverly, MA), was initially used to generate restriction fragment length polymorphism (RFLP) profiles for differentiation: the PCR amplicons were digested with 2 units of TaqI in 10-μl reaction mixtures incubated at 65°C for a minimum of 1 h. Other frequent-cutter restriction enzymes could be used to generate RFLP patterns for DNA which has no TaqI restriction site. To visualize the restriction fragments, 5 μl of each digest was separated by a 4% E-gel (Invitrogen-Novex, Carlsbad, CA) run for 30 min (preloaded ethidium bromide). The Fluor-S system (Bio-Rad, Hercules, CA) was used to create gel digital photo images in tagged image file format (TIFF).
The low- and high-GC PCR amplicons were sequenced after PCR product cleanup using ExoSAP-IT for incubation with the PCR product at 37°C for 15 min, followed by incubation at 80°C for 15 min to inactivate the enzymes (USB, Cleveland, OH). The purified products (0.2 to 0.5 μg) were sequenced in separate 10-μl reaction mixtures containing 1 μM (each) oligonucleotide primer pair and 4 μl of BigDye Terminator, version 3.1, cycle sequencing RR-100 reagent (Applied Biosystems, Foster City, CA). Mixtures were thermocycled 35 times through 96°C for 10 s, 50°C for 5 s, and 60°C for 1 min. The reaction products were clarified using a DyeEx 2.0 Spin kit (QIAGENE, Valencia, CA). Then, 20 μl of Hi-Di Formamide (Applied Biosystems, Foster City, CA) was added to 10 μl of purified sequencing reaction product. Samples were loaded in a capillary electrophoresis system (Model 3130X1 Genetic Analyzer, Applied Biosystems, Foster City, CA).
The new sequences were deposited in the GenBank, National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/). The low-GC content sequences were deposited under the following accession numbers: SKPV_WA78, GQ923123; VPXV_CA85, GQ923124; RACV_MD61, GQ923125; DPXV_V89, GQ923126; 2001_960, GQ923127 (Pox_WA01960); POX_OH08040, GQ923128; Cotia virus, GQ923129; NY_014, GQ923130; VACV_07065, GQ923131; VACV_07070, GQ923132. The high-GC content sequences were deposited under the following accession numbers: PCPV_06025, GQ902049; PCPV_08024, GQ902050; PCPV_07012, GQ902051; PCPV 07013, GQ902052; BPSV_07058, GQ902053; BPSV_07005, GQ902054; ORFV_06044, GQ902055; ORFV_08041, GQ902056; MOCV_08031, GQ902057.
A new multiple sequence alignment program, MAFFT, was used to align poxvirus genome sequences (10). MAFFT constructed initial alignments by the progressive method and finished alignments using iterative refinement methods, which improved the speed of the alignment of large, complex genome sequences of multiple genera of poxviruses. The genome sequence alignment of each genus used is depicted in Fig. Fig.1.1. The entomopoxvirus genome sequences were not included in the analysis as they diverged significantly from chordopoxviruses, with considerable gene order rearrangement. Within the chordopoxvirus subfamily, avipoxviruses have the largest genome and are more diverged from other genera of chordopoxviruses. In addition, avipoxvirus genomes contain multiple rearrangements in comparison to other chordopoxviruses, which introduced numerous gaps in the genome sequence alignments, reducing the quality of the alignment results. Thus, for the low-GC PCR assay design, we did not include the avipoxvirus genome.
The genome sequence alignments were manually studied to identify suitable primer targets. The low-GC PCR assay target selected spans a portion of insulin metalloproteinase-like protein (forward primer, G1L ortholog) to the IMV membrane protein gene (reverse primer, G3L ortholog) and generates an amplicon about 230 bp in size (Fig. (Fig.1A).1A). The primer annealing site sequences are conserved except for a single nucleotide variation in the SWPV_NEB and RFV_KAS forward primer annealing sites. Neither single nucleotide polymorphism (SNP) is expected to affect the PCR amplification of these DNA. The BLAST analysis of low-GC primer sequences with Avipoxvirus genomes suggests that the low-GC PCR could not amplify avipoxvirus DNA as the primer annealing target sequences have limited homology, and the target size changed significantly. Genome sequences with high-GC content are limited; most sequences are from the genus of parapoxvirus, but one genome is from the genus of Molluscipoxvirus and an unclassified poxvirus, crocodilepox virus. The high-GC PCR genome target selected was one of the RNA polymerase subunit genes (VAC-COP J6R ortholog); the primer sequences are conserved, and the primers are predicted to generate an amplicon about 630 bp in size (Fig. (Fig.1B1B).
The initial validation used purified poxvirus DNAs. As expected, the primers do not amplify the high-GC poxvirus DNAs. The low-GC PCR assay amplified an appropriately sized amplicon from all the poxvirus isolates or clinical materials with known low-GC content tested (Fig. (Fig.2A).2A). This included 10 Eurasian OPV, three North American OPV, two leporipoxviruses, two yatapoxviruses, three unclassified poxviruses, and a new poxvirus isolate from a fox squirrel lesion. Concordant with the predicted amplicon sequence alignments, the amplicon sizes are similar among the isolates in different genera (Fig. (Fig.2A).2A). The RFLP analysis of the amplicons with TaqI restriction endonuclease digestion provided an additional, convenient way to differentiate some genera or species. All Eurasian OPV yield an identical double-banded (about 127 and 107 nucleotides in size) RFLP pattern; North American OPV also yield double bands, but the sizes of these bands (about 170 and 64 nucleotides in size) are different from those of the Eurasian OPV. Of note, the newly discovered poxvirus NY_014 (CDC personal communication) and a deerpox virus DPXV_V89 from North America have an RFLP pattern similar to that of North American OPV. The two yatapoxviruses and the unclassified Cotia virus are not cleaved by TaqI. TaqI cleavage of the amplicon produced from the two leporipoxvirus DNAs produced multiple small fragments; TaqI cleavage of the amplicon generated from the fox squirrel poxvirus lesion also produced multiple bands, but the banding pattern sizes were not identical to those derived from the two leporipoxviruses, myxoma virus and rabbit fibroma virus.
The high-GC PCR assay was designed to amplify the known high-GC content poxviruses: parapoxvirus, molluscum contagiosum virus, and crocodilepox virus strain Zimbabwe. Neither the parapoxviruses nor molluscum contagiosum virus is easily propagated in immortalized cell culture systems. For the validation of the high-GC PCR assay, we directly used nucleic acid from parapoxvirus and molluscum contagiosum virus clinical material for the PCR amplification and sequencing confirmation. Two ORFV, four PCPV, two BPSV, and one molluscum contagiosum virus DNA were amplified (Fig. (Fig.2B).2B). The corresponding TaqI RFLP patterns are unique for each species and agree with the predicted RFLP patterns predicted from sequences (unpublished results). The RFLP patterns of BPSV isolates are similar but different from the pattern of MOCV_08031, as predicted by sequences. In comparison, the predicted TaqI RFLP pattern of MOCV_08031 shows variation from that of MOCV_T1. The sequences show that MOCV_08031 is about 97% identical to MOCV type I.
The low- and high-GC PCR assays were further tested with 146 poxviruses including nucleic acids directly extracted from one swinepox virus, one crocodilepox virus, and clinical materials from one molluscum contagiosum virus (Table (Table2).2). The low-GC PCR assay was able to amplify orthopoxviruses (including 8 camelpox, 78 cowpox, 17 monkeypox, 8 mousepox, 14 vaccinia, and 1 raccoonpox virus isolates), capripoxviruses (including 1 lumpy skin disease virus and 1 sheeppox virus), suipoxvirus (including 1 swinepox virus), yatapoxviruses (including 1 tanapox virus), and 1 Yaba-like disease virus.
The high-GC PCR assay was able to amplify DNA from 13 parapoxvirus isolates including, 9 orf viruses, 2 bovine papular stomatitis viruses, 1 pseudocowpox virus, and 1 camel parapoxvirus species and DNA from 1 molluscum contagiosum virus and 1 crocodilepox virus. Both low-GC and high-GC PCR assays worked well with poxvirus DNA from cell cultures and clinical samples. For the high-GC PCR assay, the DNA was extracted from clinical samples as high-GC poxviruses are difficult to grow in cell culture systems.
The low- and high-GC PCR assays have been used successfully to aid in the identification of suspected poxviruses in a variety of animal-derived rash specimens. The newly discovered poxvirus NY_014, which caused a progressive panniculitis in an immune-suppressed patient, was amplified by the low-GC PCR. A DNA sample (POX_08040) extracted from a fox squirrel (Sciurus niger) with multiple nodular lesions was tested to rule out the possible infection with the oral rabies vaccine (a live vaccinia-rabies glycoprotein recombinant virus) (9). The CDC standard OPV generic and vaccinia-specific real-time PCR assays did not amplify DNA derived from the sample (16); that the low-GC PCR amplified the DNA and sequence from the amplicon shows that the infection was caused by a leporipoxvirus-like poxvirus (Fig. (Fig.3A),3A), perhaps related to the squirrel fibroma virus which was first reported by King et al. over 35 years ago (12).The low-GC PCR also identified a new poxvirus (POX_01960) from nucleic acid extracted from formalin-fixed tissue of a rat skin lesion; the rat was a contact of a child with a febrile rash illness in the northwestern United States (Fig. (Fig.3A)3A) (CDC personal communication). The amplicon-derived sequence did not have significant identity to any known poxvirus nucleotide sequences in a GenBank search (http://blast.ncbi.nlm.nih.gov). This may support the phylogram, which suggests that sample POX_01960 contains a virus which is a member of a new genus of chordopoxvirus as it forms a distinct clade.
MOCV_08031 was obtained from a specimen initially suspected to be a herpes virus infection. The high-GC PCR assay amplified the clinical sample, and the sequences of the amplified amplicon identified it as a molluscum contagiosum-like virus (unpublished results). The MOCV has at least three subtypes (19), and the sequence of the amplicon derived from the clinical sample MOCV 08031 is 3% divergent based on nucleic acid content relative to MOCV subtype I (Fig. (Fig.3B);3B); this may indicate that it is another subtype. The high-GC PCR assay amplified three PCPV isolates and two BPSV isolates which were initially diagnosed by PCPV and BPSV real-time PCR assays, respectively (unpublished results). Comparative sequence analysis of the amplicons generated by the high-GC PCR assay was concordant with the prior real-time PCR assay results and additionally indicated that PCPV and BPSV contain subclades, as does ORFV (Fig. (Fig.3B3B).
The designs of the low-GC and high-GC pan-poxvirus assays were based on the available chordopoxvirus genome sequences. Avipoxviruses have greatly divergent genome sequences and need a separate diagnostic PCR assay. The high-GC PCR assay was generated with limited but widely diverged genome sequences and could be further modified when more high-GC genome sequences become available. Our initial test showed that the high-GC PCR did not amplify a sealpox virus DNA.
The new low-GC and high-GC PCR assays significantly increased our ability to diagnose unknown rash illnesses as poxvirus in origin and gave us preliminary information about additional, novel poxviruses. Broader use of these assays in screening for unknown vesiculo-pustular illness may increase our awareness of poxvirus infections in humans and nonhuman animals worldwide. The assays can be used to monitor zoo animals, domestic animals, or wild animals for mild rash illnesses. We expect many new poxviruses can be discovered by the screening of rash-derived specimens using these pan-pox assays.
We thank M. Buettner (Oberschleissheim) for providing parapox virus DNA and W. Eichhorn (Munich) for providing clinical material of swinepox virus. The skilled assistance of L. Dobrzykowski is greatly acknowledged.
The findings and conclusions in this report are those of the author(s) and do not necessarily represent the official position of the Centers for Disease Control and Prevention/the Agency for Toxic Substances and Disease Registry.
Published ahead of print on 11 November 2009.