INFERRING GENOMIC RECOMBINATION FOR IMPROVED DISEASE CONTROL: INSIGHTS FROM X. PERFORANS

Xanthomonas perforans (Xp) is one of four species causing bacterial spot disease of tomato and pepper, with X euvesicatoria (Xe), X vesicatoria and X gardneri. Xp is more closely related to Xe and are found worldwide. Durable management against this disease has been elusive, with some studies showing evidence for breakdown of resistance within a single season even with gene pyramiding. Based on single and multilocus studies, recombination between Xp and Xe has been suggested but the rate of recombination depends on the genes studied. In a recent multilocus study, bacterial spot strains from Nigeria were identified as Xp based on the highly conserved pathogenicity-related hrpB gene, but were more closely related to Xe based on six housekeeping genes. In this study, our objective was to unravel the identity of these strains and determine the genomic impact of recombination on Xp, particularly on proteins important for pathogenicity such as secretion systems and effectors. We sequenced the genome of a representative strain and compared it with 35 previously sequenced X. perforans genomes. In total, our dataset included one strain each from Nigeria and Italy, and 33 strains from the United States. We used these data to infer a core genome maximum likelihood phylogeny, which was used for recombination analysis using the ClonalFrameML software. We also compared these genomes with all previously sequenced Xe strains. Our analysis identified the Nigerian and Italian strain as lineages of Xp, with the former being more closely related to Xe strains than all other Xp strains. The number and impact of recombination events differed among lineages. While the relative rate of recombination to mutation was about 8.5 for the dataset, the number of recombination events in the Nigerian and Italian lineages were respectively 465 and 111. Recombination events among the US strains vary, with highest of 113. Recombination changed gene content in the secretion systems. The most striking variation among strains was found in the type four secretion system; the Nigerian strain had two T4SS gene systems, the Vir and Dot/Icm systems-which are important in lateral gene transfers between bacteria and delivery of effectors into hosts. While there were differences in copy number, all other Xp strains had only the Vir system. Furthermore, comparison of type three effectors showed that the Nigerian and Italian strains uniquely had avrHah1, XopAQ and PTP effectors. Additionally, the Nigerian strain was the only one to lack XopJ4. Effectors are often the bacterial targets for breeding efforts and XopJ4 was recently suggested to be a potentially stable target. Taken together, our results showed that different lineages of Xp have evolved independently due to recombination affecting different locations in the genome and at varying rates. The result is lineage-specific secretion system and effector profiles. Our analysis indicates that genome characterization will be necessary for effective resistance breeding against X. perforans.