DNA sequencing reveals false positives during the detection of aster yellows phytoplasmas in leafhoppers

During the summer of 2001 and 2002, 850 and 865 carrot plants and 926 and 2584 leafhoppers associated with aster yellow (AY)-type disease were collected from five fields in Manitoba, Canada. DNA was extracted from 999 individual leafhoppers and 381 leaf tissues from both apparently healthy and AY-like infected carrot plants. All DNA samples were examined by PCR for the presence of phytoplasmas using three universal primer pairs P1/P6, P1/P7 and R16F2n/R2 derived from phytoplasma rDNA sequences. DNA amplification with these three primer pairs generated the expected amplification products of 1.7, 1.5 and 1.2 kb, respectively. Diluted PCR products obtained using universal primer pair P1/P6 were nested with R16F2n/R16R2n. The latter set of primers amplified DNA samples from 92 carrot plants and 83 leafhopper samples. In order to assess the diversity among insect and plant phytoplasmas, nested PCR products from all 92 carrot and 83 leafhopper samples were subjected to RFLP analysis using restriction endonucleases KpnI, MseI, and HhaI. This RFLP analysis showed similar patterns among carrot and leafhopper samples. Phytoplasmas detected in most samples belonged to the subgroup 16Sr-IA. To understand why the R16F2n/R16R2n of primers did not amplify the PCR product obtained from the first PCR in the remaining samples, four PCR products of P1/P6 from two plants (representing 16Sr-IA and 16Sr-IB) and two leafhopper samples that did not amplify with nested PCR, were used for DNA sequencing. 1 The BLAST analysis of the obtained sequences showed that the PCR amplicons from the two carrot samples precisely matched with the GenBank sequences of known phytoplasmas. Alignments of these two sequences have shown very slight variations (transition/transversion ratio mean of 0.539) that would correspond to the minor differences at the 16S level between the 16Sr-IA and 16Sr-IB phytoplasma subgroups. The sequences of PCR products obtained from the two insect samples had similarity (>98%) with the sequences of phytoplasma in carrot except that their length differed from the carrot samples by 6 bp. They actually matched bacterial sequences from the GenBank, indicating that a single PCR using P1/P6 was not enough to detect phytoplasma in leafhoppers.