Bacterial sequences interfering in detection of phytoplasma by PCR using primers derived from the ribosomal RNA operon

Effective detection of pathogenic phytoplasma in deciduous fruit tree budwood is critical for investigating their epidemiology and for quarantine of these agents. The polymerase chain reaction (PCR) is the method of choice for rapid sensitive detection of phytoplasma. For PCR detection of most phytoplasmas, universal primers have been selected from the highly conserved 16S rRNA gene. Such primers are believed to detect the majority of known fruit tree phytoplasmas. To detect phytoplasmas by PCR, we initially used the primer pair P1/Tint, located in the 16S and tRNA spacer region of the tRNA operon. In 1999, 350 samples of dormant budwood of stone and pome fruit selections from around the world were tested for phytoplasma by PCR. Surprised by the high number of positive samples in PCR, we re-tested all of the positive samples with the different pair of universal primers rU3/fU5. Most of the samples did not yield amplification products with the second pair of primers. We searched for a possible source of contaminating DNA producing false signals in PCR and found that washing and sterilizing the bark prior to sample preparation eliminated many positive results. We cloned two of the PCR products generated with the P1/Tint primers and found that their sequences had high homology (over 90%) to known bacterial sequences. We isolated bacteria from dormant bark establishing individual colonies on nonselective media; several of these bacterial colonies yielded PCR amplification products of the expected size with the P1/Tint primers. By fatty acid analysis, three of these bacterial isolates were identified as members of the Staphylococcus genus; one isolate was identified as S. sciuri. A species of Pseudomonas isolated from apple rootstock independent of the current experiment also yielded products with the P1/Tint primer pair. The primer pair rU3/fU5 did not amplify DNA from any of these bacterial colonies. This primer pair, however, does sometimes yield amplification products from other DNA of unknown origin, which have electrophoretic mobility similar to that of the phytoplasma product. Extensive analysis of the phytoplasma and bacterial sequences currently available in data base helped us to develop new universal primer pair P1694/Pc399 which did not amplify bacterial or other non-phytoplasma DNAs. Our results suggest that reliable detection of phytoplasma by PCR is complex and may require the use of more than one pair of primers.