Prediction of NB‐LRR resistance genes based on full‐length sequence homology. (18th April 2022)

• Record Type: Journal Article
• Authors: Andolfo, Giuseppe; Dohm, Juliane C.; Himmelbauer, Heinz
• Is Part Of: Plant journal. Volume 110:Number 6(2022)
• Page Start: 1592
• Page End: 1602
• Publication Date: 2022-04-18
• Subjects: plant breeding -- disease resistance genes -- genome annotation -- repeat masking -- gene prediction -- Beta vulgaris -- Solanum lycopersicum -- Cucurbita; Plant molecular biology -- Periodicals; Plant cells and tissues -- Periodicals; Botany -- Periodicals; 580
• DOI: 10.1111/tpj.15756
• Languages: English
• ISSNs: 0960-7412

Abstract

SUMMARY: The activation of plant immunity is mediated by resistance (R)‐gene receptors, also known as nucleotide‐binding leucine‐rich repeat (NB‐LRR) genes, which in turn trigger the authentic defense response. R‐gene identification is a crucial goal for both classic and modern plant breeding strategies for disease resistance. The conventional method identifies NB‐LRR genes using a protein motif/domain‐based search (PDS) within an automatically predicted gene set of the respective genome assembly. PDS proved to be imprecise since repeat masking prior to automatic genome annotation unwittingly prevented comprehensive NB‐LRR gene detection. Furthermore, R‐genes have diversified in a species‐specific manner, so that NB‐LRR gene identification cannot be universally standardized. Here, we present the full‐length Homology‐based R‐gene Prediction (HRP) method for the comprehensive identification and annotation of a genome's R‐gene repertoire. Our method has substantially addressed the complex genomic organization of tomato ( Solanum lycopersicum ) NB‐LRR gene loci, proving to be more performant than the well‐established RenSeq approach. HRP efficiency was also tested on three differently assembled and annotated Beta sp. genomes. Indeed, HRP identified up to 45% more full‐length NB‐LRR genes compared to previous approaches. HRP also turned out to be a more refined strategy for R‐gene allele mining, testified by the identification of hitherto undiscovered Fom‐2 homologs in five Cucurbita sp. genomes. In summary, our high‐performance method for full‐length NB‐LRR gene discovery will propel the identification of novel R‐genes towards development of improved cultivars. Significance Statement: De novo prediction of plant disease resistance (R)‐genes is difficult due to their intrinsic features. We introduce Homology‐based R‐gene Prediction, a novel strategy for the comprehensive prediction of R‐genes and their alleles, outperforming previous approaches of R‐gene discovery.