PhD Exit Seminar for Renan Nascimento de Almeida (Guttman Lab)

Identifying host-adaptive genomic signatures in the Pseudomonas syringae

Abstract

Pseudomonas syringae is a diverse bacterial species complex capable of causing disease on various crop and wild plant species. Despite the relatively high number of affected crops, an understanding of the genetic basis allowing P. syringae to promote disease on many of these hosts is lacking. My goal was to identify genomic signatures from P. syringae genomes associated with adaptation to distinct types of hosts. I hypothesized that such signatures were overrepresented in pathogens from a given host, and that they could be extracted when compared to genome data from non-pathogens. Using statistical approaches such as GWAS and machine learning, I identified five plasmid-borne type III secreted effectors (T3SEs) and 32 other plasmid-borne genes strongly associated with isolation from diseased coffee plants. We then showed that isolates lacking this plasmid exhibit reduced growth on coffee leaves, suggesting that these genes play an important role on coffee colonization by P. syringae. Furthermore, I employed machine learning techniques to investigate if we could use genomic data to predict P. syringae virulence on bean and found that our model could predict virulence with high accuracy (mean absolute error = 0.05). I functionally validated the model by predicting virulence for 16 strains and found that 15 (94%) had virulence levels within the bounds of estimated predictions, demonstrating the power of machine learning for predicting host-specific adaptation. Finally, I also identified specific T3SE gain and loss events potentially associated with bean adaptation and showed that the loss of the HopAZ1 T3SE contributed to this process. Its presence reduced the growth of the strong kidney bean pathogen P. syringae pathovar phaseolicola 1448A on bean. Altogether, my thesis shows that pathogen genomes contain a written history of past evolutionary events associated with adaptation to specific hosts. Such signatures can be unraveled by applying statistical approaches to next-generation sequencing data.

—————————————————————-

Join Zoom Meeting

Wednesday, November 10th, 2021 at 10:00 am

https://utoronto.zoom.us/j/89254487823

Meeting ID: 892 5448 7823

Host: David Guttman (david.guttman@utoronto.ca)

—————————————————————-