Can a single genomic difference result in a better biocontrol agent?

Abstract

The filamentous fungus Trichoderma atroviride is used widely for biological control of major plant diseases. Biocontrol activity is linked to a wide range of parameters and we have limited knowledge of what specific attributes make certain strains particularly effective as a biocontrol agent (BCA). In this study, we report on molecular differences between two closely related T. atroviride isolates and the implications they might have on biocontrol activity. Two T. atroviride isolates (LU132 and LU140) were isolated from the same onion paddock in New Zealand as potential BCAs against onion white rot. The isolates were phenotypically distinct and extensive studies found that LU132 was an excellent BCA and achieved better control than LU140 of a variety of plant pathogens. Direct comparison of a wide range of phenotypic characteristics of the two isolates found that the biggest difference between the isolates is LU132’s faster growth rate. Attempts were made to design an isolate-specific molecular marker for LU132, but it was impossible to distinguish it from LU140, suggesting high genetic similarity. Sequencing of the genomes of both isolates revealed only 2 single nucleotide polymorphisms (SNPs). The SNPs were associated with 3 genes similar to: 1) a hypothetical gene, conserved in fungi; 2) predicted Small EDRK-rich Factor H4F5 (SERF), conserved in animals and fungi; 3) Proliferating Cell Nuclear Antigen (PCNA), conserved in eukaryotes and archaea. No expression differences of these 3 genes could be identified but cDNA sequencing revealed a gene annotation error in SERF. While the published annotated genome of T. atroviride describes SERF as having 2 exons and one intron, it actually has an additional exon in LU132 and LU140. One of the SNPs is located in the 3rd exon of SERF where it changes the amino acid sequence in LU132. Bioinformatic analysis suggests that the amino acid change creates a MAPK docking motif on the LU132 SERF protein that is not present in the LU140 protein. To study the function of the gene it was knocked-out in both isolates via Agrobacterium-mediated transformation and the mutants were subjected to phenotypic examinations. We will present first results from the gene function analysis and discuss potential implications for biocontrol activity in this important species.