Characterisation of the Rhynchosporium commune - barley interaction

Abstract

Rhynchosporium commune is the causal agent of barley scald disease. Although scald is a significant issue for commercial barley growers, the molecular mechanisms underpinning the disease are poorly understood. To address this, I generated a high-quality assembly of the Australian R. commune strain WAI453 using long-read sequencing data. The resulting 57.76 Mb assembled genome contained 20 gapless nuclear contigs. To complement this assembly, RNA-seq datasets were generated from in vitro and in planta samples and used to facilitate the genome annotation and differential gene expression (DGE) analysis. The genome annotation was generated by the combination of the two de novo gene predictors CodingQuarry and BRAKER with RNA-seq reads as the helper for annotation generating high quality of genome annotation with total 13,726 predicted genes. In the DGE analysis, there were 1061 upregulated genes during in planta growth compared to 418 genes being upregulated during in vitro growth. In that DGE dataset, the NIP2 and NIP2-like protein (NLP) effector genes were highly expressed during in planta growth with NIP2.1 and NLP3 having highest expression level. Population genetics was performed to analyse the diversity of the NIP2 genes in the global population of R. commune and its sister species. Only a single haplotype of NIP2.3 was present in 99.48% of the tested 191 R. commune global isolates. In addition, NIP2.6 was only present in the R. commune but not present in its sister species. These results suggest NIP2.1, NLP3, NIP2.3, and NIP2.6 are important for R. commune. Functional studies of NIP2.1 was then performed by heterologously producing the protein in Escherichia coli using the CydisCo system. Soluble protein was generated and correct folding was confirmed by mass spectrometry and circular dichroism spectroscopy. Infiltration assays of the purified protein into barley leaves failed to induce cell death nor suppress a reactive oxygen species (ROS) burst suggesting this protein has a different function for R. commune during plant infection. Further analysis of the predicted NIP2.1 structure revealed a structural similarity to other proteins containing an RNA binding domain in the Protein Data Bank suggesting NIP2.1 might have an RNA binding function. Further analysis of the DGE dataset also revealed that four secondary metabolite gene clusters were upregulated during disease. One of the clusters, PKS4, contained a predicted cluster-specific transcription factor gene WAI453.tig_09.16 that was highly expressed during in planta growth compared to in vitro. Given its strong expression during disease, the function of the PKS4 cluster was examined by silencing and overexpressing WAI453.tig_09.16 in R. commune. Mutant strain of R. commune expressing reduced levels of WAI453.tig_09.16 gene was less virulent than the R. commune-wild type implying a role for the PKS4 cluster in disease. In addition, an analysis of the culture filtrates of the WAI453.tig_09.16-overexpression mutant grown in vitro showed the organic and water phases induced necrosis on barley leaves. Five hybrid terpene-polyketide compounds linked to PKS4 cluster, named rhynchospene A-E, were discovered from organic phase WAI453.tig_09.16-overexpression mutant culture filtrate. The Rhynchospene B induced necrosis in barley leaves at 200 ppm. In contrast to the activity identified in the organic phase, the necrosis-inducing activity from water phase of the WAI453.tig_09.16-overexpression mutant culture filtrate was found to be proteinaceous. Bottom-up proteomics was performed on the partially resolved culture filtrate and a predicted effector, WAI453.tig_02.762, was highly abundant in the culture filtrate of the WAI453.tig_09.16-overexpression mutant compared to the filtrates from the WT strain. Overall, these results advance our understanding about virulence factors of R. commune and provide a valuable community resource for further exploration of this important pathogen.