Detection of North American orthopoxviruses by real time-PCR

Amino acid sequences of the G9R gene were aligned using the ClustalW accessory in BioEdit (Figure 1) and translated back into the original nucleotide sequence for phylogenetic analysis. A heuristic tree search was carried out in PAUP* 4.0b10 over 1000 replicates.

Bootstrap values were calculated from 1000 bootstrap replicates of 1000 random addition sequence replicates each. One of two most parsimonious trees is illustrated in Figure 2.

We designed a real time PCR (rt-PCR) assay based on TaqMan^® chemistry and technology using the VPXV 086 gene, SKPV 086 gene and RCNV 070 gene (ortholog of the VACV Copenhagen G9R). The targeted gene encodes for the myristylated protein, which is an essential component for cell membrane fusion and entry[18]. The orthologous genes for all viruses included in this study were aligned using BioEdit to pick the target area on the basis of conservation and GC content. The primers G9R-NA OPXV forward (5'YGG-ACC-RGG-AGG-TCT-TTC-TGC-ATT 3'), G9R-NA OPXV reverse (5'TCT-GGC-CAA-CAT-GAT-TCT-AAT-ACT-GCR-TC 3'), and G9R-NA OPXV probe (5'FAM AG-GGA-ACG-VTA-YAA-TGG-YAC-YGC-TCC-CAA-YTG-CTG-TCG-CAC-TTT-BHQ1 3') were designated to amplify a 156 bp fragment from a highly conserved region of the gene (Figure 1). Primers and probe were synthesized in the Biotechnology Core facility (CDC, Atlanta, GA), utilizing standard phosphoramidite chemistry. The nomenclature for the degenerated positions in the primers and probe designed was based on IUPAC. The probe contained the fluorophore (FAM) at the 5' end and the black hole quencher at the 3' end. DNA amplification was carried out using default rt-PCR thermal conditions for the ABI7900 (Applied Biosystem, Foster City, Ca), briefly: one cycle of 95°C for 10 minutes, followed by 40 cycles of 95°C for 15 seconds and 60°C for 1 minute. PCR amplification is based on fluorescent emission after annealing/elongation (60°C). The assay was performed in a final volume of 25 ul per well containing 12.5 ul of TaqMan^® Universal PCR Master Mix (Applied Biosystem, New Jersey, USA), 4.5 ul of deionized distilled water, 1 ul of each forward and reverse primer at a concentration of 20 uM, 1 ul of probe at a concentration of 10 uM, and 5 ul of template DNA.

A total of thirteen chordopoxviruses, which comprised seven old world OPXVs, three NA OPXVs, and three from different genera, were included to test the analytical specificity of the assay (Table 1). The latter three included one Parapoxvirus, one Leporipoxvirus and one Yatapoxvirus. All viruses in the panel (with the exception of the squirrel fibroma virus, which was a crude cell culture extract), were propagated via cell culture and purified by two sucrose cushions[19‐21]. Total viral DNA was extracted using Qiagen tissue kits on the BioRobot^® EZ1 workstation, according to the manufacturer's instructions, and later quantified by spectrophotometry using Qubit™ Quantitation Platform by Invitrogen™. All viruses in the panel were serially diluted with deionized distilled water until obtaining appropriate working concentrations. To determine analytical specificity of our assay, the total DNA concentration was adjusted to 2 ng per 5 ul for all the thirteen viruses in the panel, and the rt-PCR assay was run in triplicate for each sample.

Analytical sensitivity was determined by using DNA of the three NA OPXVs at seven different concentrations, 40 fg, 4 fg, 2 fg, 1 fg, 0.5 fg, 0.25 fg, and .125 fg/ul. The rt-PCR for each concentration was run in 24 replicates. A statistical probit analysis, a non-linear regression model, using commercial software SAS 9.2 (SAS Institute, Cary, NC, USA) was done to determine the detection limit of each assay with 95% confidence.

Peromyscus californicus (Peromyscus Genetic Stock Center, University of South Carolina) were infected intranasally with 10 ul of volepox virus at a concentration of 1.6 × 10³ PFU or mock infected with 10 ul of phosphate-buffered saline (PBS). Pre-screening of the animals included evaluation of anti-VPXV and anti-VACV immunoglobulin type G by modified ELISA with 1:100 dilution of animal serum, and orthopoxviral DNA in EDTA blood, oral swab, ocular swab and anal swabs by rt-PCR as previously described[22, 23]. Specimens taken from live animals (blood, oral swab, ocular swab, and anal swab), and during necropsy (spleen, gonad, kidney, submandibular lymph node, supradrenal gland, brain, liver and lung) were performed/collected in accordance with CDC Institutional Animal Care and Use Committee (IACUC) guidelines under the approved protocol 2126 CARMOUC. In some animals, additional organs were taken at necropsy if lesions or signs of disease were observed (skin, tongue, bladder, intestine and/or pancreas) (Table 2). A total of 187 samples, from different tissues and bodily fluids, were collected from animals with clear signs of disease (for example: conjunctivitis, ruffled hair, skin lesions, crusty noses, apathy and/or swollen faces) on days 6, 7 and 8 post-infection as well as 45 tissue samples and bodily fluids from healthy, never infected Peromyscus californicus. Swabs were hydrated for 5 minutes with 400 ul of PBS, and transferred to SETS tubes (Swab Extraction Tube System, Roche, Catalog No. 3315568). The SETS tubes were centrifuged in the VWR Galaxy Mini centrifuge (6000 rpm) for 1 minute to collect the eluant for DNA extraction and cell culture. The tissue samples were suspended in 1 ml of PBS and a sterile SPEX bead was added. The GenoGrinder 2000 (SPEX Sample Prep) was used following the manufacturer's instructions to triturate the tissue for later cell culture and DNA extraction using the Qiagen tissue kits on the BioRobot^® EZ1 workstation, according to the manufacturer's instructions.

BSC-40 cell monolayers (African green monkey kidney cell line) were inoculated with 10-fold dilutions of sonicated tissue extract or swab elute. Infected cells were incubated at 36°C in a 6% CO₂ atmosphere in semi solid medium (RPMI + 1% Carboxymethylcellulose). Cell infection was monitored microscopically by observation of OPXV specific cytopathic effect (CPE). At 48 h post inoculation, cells were stained with crystal violet to reveal plaques to determine the viral titer (PFU/ml). A total of 187 putative positive, and 45 negative samples were examined for the presence of viable virus by cell culture isolation. These samples were also used for determining analytical sensitivity and specificity of the rt-PCR.

The orthopoxviruses are all very closely related[11]. Yet the NA OPXVs cluster into their own phylogenetic subclade, as shown in the alignment of the myristylated protein ortholog (Figure 2). Our NA OPXV rt-PCR assay was validated using at least one representative strain of each species of OPXV as well as one strain from the genera Parapoxvirus, Leporipoxvirus, and Yatapoxvirus (Table 1). The assay consistently detected all three NA OPXVs at 2 ng total DNA (equivalent to > 8-9 million virus genomes), with threshold cycle (Ct) values ranging from 16 to 17(Table 1). The purity of these viral preparations was verified using a real-time PCR assay to detect contaminating cellular DNA (RNase P), which did not cross-react with any of the NA OPXVs DNA preparations (data not shown). The NA OPXV rt-PCR assay demonstrated specificity for NA OPXVs, not reacting with the other seven OPXVs or three non-orthopoxvirus members of Chordopoxviridae when tested at similar concentrations (Table 1).

To determine the analytical sensitivity of the assay for each of the NA OPXVs, 24 replicates of seven different concentrations were conducted with the NA OPXV rt-PCR assay. The assay was least sensitive for RCNV; DNA was detected in all 24 replicates down to 10 fg of DNA (Table 3). For 5 fg of total DNA, 20 out of 24 replicates yielded positive results, and for 2.5 fg, five out of 24 replicates were detected as positive. The assay showed higher sensitivity for SKPV DNA, with positive results in all 24 replicates down to 2.5 fg of total DNA (Table 3). At 1.25 fg, 18 of 24 replicates were detected as positive, and at 0.625 fg, eleven of 24 replicates were positive. The assay proved its highest sensitivity for VPXV DNA. All 24 replicates at concentrations as low as 2.5 fg were detected (Table 3). At 1.25 fg of total VPXV DNA, the assay detects 23 of the 24 replicates, and at 0.625 fg, 19 of the replicates were positive. The average Ct values from all positive samples are shown in Table 3. Based on RNaseP assay results, we can confidently calculate the number of viral genomes detected based on mass, since there was no detectable cellular DNA contamination in our preparations. The statistical analysis (probit) indicated that the assay is capable of detecting with 95% confidence, 1.1 fg of VPXV DNA (~5 genomes), 1.99 fg of SKPV DNA (~8 genomes) and 6.4 fg of RCNV total DNA (~29 genomes) under the cycling conditions, primers and probes described (Figure 3).

Analysis of all samples (blood, oral swab, ocular swab and anal swabs) taken prior to infection tested negative for OPXV DNA by rt-PCR, and anti-VPXV and anti-VACV IgG by modified ELISA[22, 23](data not included in the assay evaluation). One hundred and eighty four out of 187 specimens collected at days 6, 7 and 8 post-infection from different tissues and bodily fluids generated Ct values in the positive range (between 15 and 40) for our NA OPXV rt-PCR assay, indicating the presence of DNA specific for a NA OPXV (Table 2). Only three anal swabs out of 13 taken were negative, and may contain levels of viral DNA under the detection threshold for our assay (1.1 fg). All 45 samples from the naïve (PBS mock infected) animals were negative by our G9R-NA OPXV rt-PCR assay (Table 2).

One hundred and seventy-eight samples out of a total of 184 rt-PCR positive samples demonstrated detectable CPE in a single passage on BSC-40 cells. All these samples had a Ct values of 37 or below which, based on the quantification from the standard curve, would equate to a minimum of 45 genomes of VPXV DNA. Six samples out of 184 rt-PCR positives with Ct values of 38 and above were discordant with cell culture viability results; in these samples no viable virus was detected and the estimated number of genomes is less than 45. All 45 samples from naïve animals were confirmed negative by both rt-PCR and cell culture. When comparing cell culture with rt-PCR results, samples with Ct values equal or lower than 37 were positive for VPXV growth, and samples with Ct values 38 and above did not show evidence of viable VPXV.