Multiplex primer prediction software for divergent targets

Gardner, Shea N.; Hiddessen, Amy L.; Williams, Peter L.; Hara, Christine; Wagner, Mark C.; Colston, Bill W.

In some applications, such as clinical or biodetection applications, the exogenous nucleic acids from other eukaryotic sources, notably human sources, may be present in varying and unknown concentrations. To test the effects of background human genomic DNA on the performance (amplification specificity) of the Poxviridae 16-plex primer set, we conducted PCR reactions across a series of mass ratios of vaccinia Lister DNA:human genomic DNA at 1 : 1, 10 : 1, 100 : 1, 1000 : 1 and 10000 : 1. These correspond to approximate copy number ratios of vaccinia : human genomes of 1.3 Â 10 4 : 0.82, 1.3 Â 10 4 : 0.082, 1.3 Â 10 4 : 0.0082, 1.3 Â 10 4 : 0.00082 and 1.3 Â 10 4 : 0.000082, using genome sizes of 1.9 Â 10 5 bp and 3 Â 10 9 bp for vaccinia and human DNA, respectively. For context, there are 6.58 pg or two copies of the genome in one human cell. Our algorithm predicted 233 amplicons between 50 and 1000 bp from the human genome. In experiments, this would appear as a 'smear' on the agarose gel, which indeed was observed for the mass ratios of 1 : 1 and 10 : 1 (Figure 7 , lanes 6 and 7, respectively). However, even at a ratio of 10 : 1 vaccinia:human DNA (lane 7), the 617-bp vaccinia amplicon is clearly visible on the gel, despite the numerous nonspecific amplicons. At ratios of 100 : 1, 1000 : 1 and 10000 : 1 (lanes 8, 9 and 10, respectively, Figure 7) , the smear is drastically reduced to nonexistent (at the resolution of the agarose gel), and the vaccinia amplicon is readily visible. For comparison, we tested the 16-plex primers against the same mass of human DNA in the absence of vaccinia DNA at 2.7 pg (the 1 : 1 ratio mass), 0.027 pg (100 : 1 mass) and 2.7 Â 10 À4 pg (10000 : 1 mass) (lanes 12, 13, and 14, respectively, Figure 7) . The data show a similar smear at 2.7 pg as was observed when vaccinia DNA was present (lane 6). However, the low-intensity 617-bp amplicon from vaccinia is visible in lane 6 (with vaccinia) while a similar amplicon is not present in lane 12 (without vaccina). These results provide evidence that amplification with these family level primer sets will depend on viral titers in the actual sample, and that there are cases where amplification will either not be possible or require additional sample purification steps. However, as discussed further in the next section, specific amplification with a short-primer multiplex could be used for selective enrichment of viral targets by generating amplicons with known sequence, which may be feasible for viral detection if combined with a probebased amplicon detection method such as TaqMan Õ or Luminex bead based suspension arrays (http://www .luminexcorp.com/). We compared recently published conserved Orthopoxvirus primers (36) to the 148 Poxviridae genomes, and computational predictions suggest that 67 of the available genomes might not be amplified by the two primers they designed for the Orthopox genus. These included a number of monkeypox, ectromelia, several vaccinia and a couple of variola strains. However, in permissive hybridization conditions it is possible that primers would anneal despite mismatches to target, allowing more of the targets to be amplified. Four conserved Orthopoxvirus primers from an earlier publication (37) , before many of the Poxviridae genomes were available, do not match 53 genomes, including a number of monkeypox, ectromelia, camelpox and one variola minor genome. The primers in (37) included inosine bases, which we replaced with each possibl