Development of a microarray for potyvirus detection and identification

Abstract

Potyvirus is the largest and one of the most economically important of the virus genera infecting plants. The complexities of potyvirus identification resulting from many different species, mixed infections, emerging new viruses, new hosts, and new vectors, etc., often requires the use of multiple detection methods which is time consumable and costly. Therefore an assay that can test for a range of potyviruses simultaneously, with good specificity and sensitivity, is desirable. This study looked at the feasibility of producing an oligonucleotide microarray for detection and identification of potyviruses at both species and strain level. Thirty plant samples with suspected potyvirus infections were collected from field and research laboratories in New Zealand and partial NIb gene, complete CP gene and 3΄UTR were sequenced. Twelve definitive potyviruses, one tentative potyvirus, one non-potyvirus, and one novel potyviruslike sequence were identified, six of which were first records for New Zealand. Sequence analysis showed that NIb and CP genes and the 3΄UTR contained both conserved and variable sequences which were used to design both species and strain level specific probes. Four Potyvirus species were chosen for a “proof of concept” study and probes were designed using two different software programs (ROSO and CAG software). A total of eighty five probes including 33 perfect-match and 52 mismatch probes were selected to represent the four targeted potyviruses. Each probe was synthesized with spacers of either 6 or 12 poly-cytosine or poly-thymine at the 5΄ terminus. Arrays showed high specificity to the targets when tested using nineteen different geographically diverse potyvirus isolates representing the four target species, four distinct but closely related New Zealand potyviruses, and four healthy plant species. Factors affecting the hybridization efficiency, e.g. the size of the target fragments, secondary structure of probes and targets, label type, strandedness, Tm and GC content of probes, were also investigated. The approaches and protocols developed in this study should form a useful basis for developing other potyviruses arrays and the results also provide useful insights into issues of generic interest for the development of arrays for detecting other pathogens.