Abstract
BackgroundBiotrophic fungal plant pathogens cause billions of dollars in losses to North American crops annually. The model for functional investigation of these fungi is Ustilago maydis. Its 20.5 Mb annotated genome sequence has been an excellent resource for investigating biotrophic plant pathogenesis. Expressed-sequence tag libraries and microarray hybridizations have provided insight regarding the type of transcripts produced by U. maydis but these analyses were not comprehensive and there were insufficient data for transcriptome comparison to other smut fungi. To improve transcriptome annotation and enable comparative analyses, comprehensive strand-specific RNA-seq was performed on cell-types of three related smut species: U. maydis (common smut of corn), Ustilago hordei (covered smut of barley), and Sporisorium reilianum (head smut of corn).ResultsIn total, >1 billion paired-end sequence reads were obtained from haploid cell, dikaryon and teliospore RNA of U. maydis, haploid cell RNA of U. hordei, and haploid and dikaryon cell RNA of S. reilianum. The sequences were assembled into transfrags using Trinity, and updated gene models were created using PASA and categorized with Cufflinks Cuffcompare. Representative genes that were predicted for the first time with these RNA-seq analyses and genes with novel annotation features were independently assessed by reverse transcriptase PCR. The analyses indicate hundreds more predicted proteins, relative to the previous genome annotation, could be produced by U. maydis from altered transcript forms, and that the number of non-coding RNAs produced, including transcribed intergenic sequences and natural antisense transcripts, approximately equals the number of mRNAs. This high representation of non-coding RNAs appears to be a conserved feature of the smut fungi regardless of whether they have RNA interference machinery. Approximately 50% of the identified NATs were conserved among the smut fungi.ConclusionsOverall, these analyses revealed: 1) smut genomes encode a number of transcriptional units that is twice the number of annotated protein-coding genes, 2) a small number of intergenic transcripts may encode proteins with characteristics of fungal effectors, 3) the vast majority of intergenic and antisense transcripts do not contain ORFs, 4) a large proportion of the identified antisense transcripts were detected at orthologous loci among the smut fungi, and 5) there is an enrichment of functional categories among orthologous loci that suggests antisense RNAs could have a genome-wide, non-RNAi-mediated, influence on gene expression in smut fungi.
Highlights
Biotrophic fungal plant pathogens cause billions of dollars in losses to North American crops annually
It is unlikely that this is the only form of variation leading to these differences; here we focus on deep transcriptome analysis of the best characterized smut fungus, U. maydis, while augmenting the information obtained through comparative transcriptome analyses
In this study, we present deep RNA-seq and genome annotation updates for three closely related smut fungi: U. maydis, which causes common smut of corn, U. hordei, the causal agent of covered smut of barley, and S. reilianum, which causes head smut of corn
Summary
Biotrophic fungal plant pathogens cause billions of dollars in losses to North American crops annually. Several studies in the past 2 years alone have employed high-throughput and massively parallel RNA-seq approaches to update current genome annotations of phytopathogenic filamentous fungi including those of Botrytis cinerea (504 new gene models; [2]); Colletotrichum graminicola (906 new and 819 updated gene models; [3]) and Fusarium graminearum (412 new and 1529 updated gene models; [4]). These studies and others have been instrumental in correcting annotation errors, gathering information on untranslated regions (UTRs) and alternative splice sites, as well as identifying novel protein-coding genes and new isoforms, contributing to our understanding of pathogenicity determinants in phytopathogenic fungi. The deep transcriptome analysis presented here involved U. maydis, U. hordei, and Sporisorium reilianum and was carried out with a focus on comparative analyses which expanded the genome annotations and identified conserved natural antisense transcripts
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
