Genetic relatedness among Pseudomonas avellanae, P-syringae pv. theae and P.s. pv. actinidiae, and their identification

A total of 37 strains of Pseudomonas avellanae, P. syringae pv. theae and P.s. pv. actinidiae, including pathotype and reference strains, obtained from all the countries where these pathogens have been reported, were compared by means of ARDRA, repetitive PCR using ERIC, BOX and REP primer sets, whole-cell protein analysis, biochemical and nutritional tests, and pathogenicity tests. P. syringae pathovar type strains representing six genomospecies sensu Gardan et al. (1999), were also included for comparison in UPGMA cluster analysis of repetitive PCR data and SDS-PAGE of protein extracts. Among the 12 endonucleases used in ARDRA, only Tru 9I differentiated P. avellanae from P.s. pv. theae and P.s. pv. actinidiae. UPGMA cluster analysis of repetitive PCR genomic fingerprints showed 65% similarity between P.s. pv. theae and P. avellanae and 50% between the latter species and P.s. pv. actinidiae. Strains of P.s. pv. actinidiae could be grouped according to their geographic origin. Similar results were obtained with SDS-PAGE cluster analysis. PCR amplification using primers PAV 1 and PAV 22 that were developed to detect P. avellanae in apparently healthy and visibly infected hazelnut specimens yielded a band of 762 bp from all strains of P. avellanae, P.s. pv. theae and P.s. pv. actinidiae. All strains lacked the syrB gene. Based on these data, we suggest that P.s. pv. actinidiae should be included in the genomospecies 8 together with P. avellanae and P.s. pv. theae. Selected biochemical and nutritional tests could differentiate these groups of strains. Pathogenicity tests clearly indicated that each group is specifically pathogenic only on the host plant species from which it was originally isolated.