VALIDATION OF YEAST IDENTIFICATION BY IN SILICO RFLP

摘要

Reliability, high reproducibility and high throughput are advantages of genetic identification methods over conventional methods based on phenotypic characteristics. In our work, we showed that the PCR-RFLP of internal transcribed spacer (ITS) of ribosomal DNA (rDNA) method can be a very powerful identification or validation method, if it is combined with computer-based restriction analysis of sequences stored in broadening databases. We isolated 197 yeast samples from grape berries and fermenting grape must and identified 13 species. For identification conventional methods, D1/D2 sequencing and PCR-RFLP of ITS regions were used. Because in our case, conventional methods were found insufficient for reliable and fast identification, in silico RFLP was applied; using appropriate software, enzyme restriction patterns on the agarose gel were simulated and compared with the restriction pattern obtained with PCR-RFLP. With this in silico approach we managed to identify or validate the identification of all 13 species. Our results show that natural isolates of ascomycetous or basidiomycetous yeasts can be rapidly (in 3 days) and reliably identified by in silico RFLP. In silico RFLP will be highly useful in the identification process of yeasts that could be reliably identified by conventional methods. For its speed, simplicity, high reproducibility and high throughput if compared to conventional identification methods, the method will be chosen when rapid validation or identification of yeast samples is needed. Another application, determining the enzymes, which distinguishes one species from another using RFLP, is obvious. Moreover, a database of ITS sequences will have great value at phylogenetic classification of new undescribed species.We believe that a program such as CandID with restriction databases of other yeast species can enrich the microbiological practice in the future.