Prediction and Characterization of RXLR Effectors inPythiumSpecies

AbstractBeing widely existed in oomycetes, the RXLR effector features conserved RXLR-dEER motifs in its N terminal. Every knownPhytophthoraorHyaloperonosporapathogen harbors hundreds of RXLRs. InPythiumspecies, however, none of the RXLR effectors has been characterized yet. Here, we developed a stringent method forde novoidentification of RXLRs and characterized 359 putative RXLR effectors from nine testedPythiumspecies. Phylogenetic analysis revealed a single superfamily formed by all oomycetous RXLRs, suggesting they descent from a common ancestor. RXLR effectors fromPythiumandPhytophthoraspecies exhibited similar sequence features, protein structures and genome locations. In particular, the mosquito biological agentP. guiyangensecontains a significantly larger RXLR repertoire than the other eightPythiumspecies examined, which may result from gene duplication and genome rearrangement events as indicated by synteny analysis. Expression pattern analysis of RXLR-encoding genes in the plant pathogenP. ultimumdetected transcripts from the vast majority of predictedRXLRswith some of them being induced at infection stages. One such RXLRs showed necrosis-inducing activity. Furthermore, all predictedRXLRswere cloned from two biocontrol agentsP. oligandrumandP. periplocum. Three of them were found to encode effectors inducing defense response inNicotiana benthamiana. Taken together, our findings represent the first complete synopsis ofPythiumRXLR effectors, which provides critical clues on their evolutionary patterns as well as the mechanisms of their interactions with diverse hosts.Author summaryPathogens from thePythiumgenus are widespread across multiple ecological niches. Most of them are soilborne plant pathogens whereas others cause infectious diseases in mammals. SomePythiumspecies can be used as biocontrol agents for plant diseases or mosquito management. Despite that phylogenetically close oomycete pathogens secrete RXLR effectors to enable infection, no RXLR protein was previously characterized in anyPythiumspecies. Here we developed a stringent method to predictPythiumRXLR effectors and compared them with known RXLRs from other species. All oomycetous RXLRs form a huge superfamily, which indicates they may share a common ancestor. Our sequence analysis results suggest that the expansion of RXLR repertoire results from gene duplication and genome recombination events. We further demonstrated that most predictedPythium RXLRscan be transcribed and some of them encode effectors exhibiting pathogenic or defense-inducing activities. This work expands our understanding of RXLR evolution in oomycetes in general, and provides novel insights into the molecular interactions betweenPythiumpathogens and their diverse hosts.