Conflicting results obtained by RAPD-PCR and large-subunit rDNA sequences in determining and comparing yeast strains isolated from flowers: a comparison of two methods.

Sixty-six yeast strains isolated from the nectar of various plant species in Central Europe were characterized by randomly amplified polymorphic DNA PCR (RAPD-PCR) and by sequencing of the variable D1/D2 domain of large-subunit (26S) rDNA. The usefulness of both methods for the determination and comparison of unknown ascomycetous and basidiomycetous yeast strains was compared and evaluated. The reproducibility of RAPD-PCR was shown to be low and the information obtained by this method was clearly not as precise as that obtained from sequence analysis. Numerous imponderables make RAPD-PCR analysis unreliable, at least as a means of identifying yeasts in ecological studies. The lack of standard protocols for RAPD-PCR analysis and the absence of a general database of banding patterns made it impossible to identify unknown yeast strains or to recognize new species. In contrast to RAPD-PCR, sequence analysis of the D1/D2 domain was found to be a fast and reliable method for the rapid identification of yeast species and was also shown to be an invaluable tool for the discovery of new species.