Sequence-typing of Xylella fastidiosa: new perspective for MLST analysis by nanopore sequencing

Multi-locus sequence typing (MLST) is a widely used methodology for typing strains of Xylella fastidiosa (Xf) (Yuan et al., 2010; EPPO 7/24(4)). This approach is based on the sequence analyses of 7 housekeeping genes, on which the allelic profile and the sequence type (ST) is determined.  Given the difficulties to isolate and culture Xf, the MLST analysis is often performed on total plant DNA instead of using bacterial DNA recovered from cultured isolates. In some cases, this may result in weak amplifications or PCR failures. In order to overcome this issue and speed up the sequencing step we tested the Oxford nanopore technology as an alternative approach to Sanger sequencing. To this end, we used the Oxford nanopore MinION device for sequencing the PCR products obtained from 4 olives and from 3 samples of Dodonaea purpurea. Previous analyses had identified in these infected sources only alleles corresponding to the ST53 genotype. PCR barcoding Nanopore kit SQK PBK004 was used to sequence the 7 amplicons. Reads having passed the EPI2ME quality filter were launched in the MLST 2.0 pipeline (Larsen et al., 2012). Results showed that in all samples, the expected allele for each PCR product was obtained, i.e. confirming the occurrence of the ST53 genotype in all tested samples. Indeed, for some genes multiple alleles were identified. However, the occurrence of these additional alleles was not confirmed when the corresponding PCR products were cloned and a large number of clones sequenced. Overall, although this technology suffers from high error rates, its use for MLST analysis proved to be cost effective and time saving, allowing to gather a rapid snapshot of the Xf genotype associated to a given host plant/outbreak.