Phylogenetic and evolutionary analysis of NBS-encoding genes in Rosaceae fruit crops

Abstract

Phylogenetic relationships of the nucleotide binding site (NBS)-encoding resistance gene homologues (RGHs) among 12 species in five genera of Rosaceae fruit crops were evaluated. A total of 228 Rosaceous RGHs were deeply separated into two distinct clades, designated as TIR (sequences within this clade containing a Toll Interleukin-1 Receptor domain) and NonTIR (sequences lacking a TIR domain). Most Rosaceous RGH genes were phylogenetically distinct from Arabidopsis, Rice or Pine genes, except for a few Rosaceous members which grouped closely with Arabidopsis genes. Within Rosaceae, sequences from multiple species were often phylogenetically clustered together, forming heterogenous groups, however, apple- and chestnut rose-specific groups really exist. Gene duplication followed by sequence divergence were proposed as the mode for the evolution of a large number of distantly or closely related RGH genes in Rosaceae, and this mode may play a role in the generation of new resistance specificity. Positively selected sites within NBS-coding region were detected and thus nucleotide variation within NBS domain may function in determining disease resistance specificity. This study also discusses the synteny of a genomic region that encompass powdery mildew resistance locus among Malus, Prunus and Rosa, which may have potential use for fruit tree disease breeding and important gene cloning.

Introduction

Plants defend themselves against pathogens in complex ways, one of which is the gene-for-gene manner where resistance (R) genes perform a pivotal role in pathogen detection and defense-response initiation. Numerous R genes have been cloned from a variety of plant species (Hammond-Kosack and Jones, 1997, Dangl and Jones, 2001). The largest class encodes a nucleotide-binding site (NBS) and leucine-rich repeats (LRR) domain. The NBS is a signaling domain which is responsible for ATP/GTP binding and hydrolysis in defense system (Belkhadir et al., 2004). It is highly conserved, containing several strictly ordered motifs, i.e. P-loop, kinase-2, kinase-3a, and GLPL motifs. With degenerate primers targeted to the conserved motifs, a large number of NBS-encoding sequences with homology to R genes, so-called “R gene homologues” (RGHs), have been isolated from different plant species (Kanazin et al., 1996, Leister et al., 1996, Shen et al., 1998, Yu et al., 1996, Collins et al., 1998, Zhu et al., 2002, Ashfield et al., 2003). In fact, RGHs are abundant in plant genomes, comprising an estimated 2% of the Arabidopsis genome (Meyers et al., 2003). The large number of RGH sequences provides a facile system to study the evolutionary biology and genetics of NBS-encoding gene family. Meyers and colleagues (1999) proposed that RGHs phylogenetically fall into two distinct groups, i.e. TIR and NonTIR subfamilies. Different evolutionary histories were thereafter suggested for the TIR and NonTIR subfamilies during the long period evolution of plant–pathogen interaction (Cannon et al., 2002). Moreover, genetic analysis has shown that RGHs tend to occur in clusters, and often map to major resistance genes or quantitative trait loci (QTL) (Michelmore and Meyers, 1998, Young, 2000). These clusters may provide information about the organization and evolution of R genes and RGHs in plant genomes (Grube et al., 2000, Pan et al., 2000a, Pan et al., 2000b, Bai et al., 2002, Meyers et al., 2003). Furthermore, these findings have raised an interesting question of whether RGH gene clusters and their genetically linked R loci exhibited conservation among related plant species. If there is strong synteny and collinearity, R loci in less studied crop species which confer resistance to identical or related pathogens should be located at corresponding positions in the genomes of model plant species. Comparative genomics provides a powerful tool to resolve this problem (Gale and Devos, 1998). R genes and RGHs in two families, Poceae and Solanaceae, have been extensively studied. In cereal genomes, Leister et al. (1998) observed rapid re-organization of RGH clusters and poor synteny of R loci between related species. However, within Solanaceae family, RGHs and R loci appeared to be positionally well conserved (Pan et al., 2000a). Further research into this area, including studies from other plant families, is required.

Rosaceae is an economically important family throughout the world, including the most important fruit-producing crops such as apples (Malus), pears (Pyrus), strawberries (Frageria) and stone fruits (Prunus; peachs, plums, apricots, etc.), as well as other valuable ornamental plants including roses (Rosa). Most of these species are woody perennials with a long intergeneration time and large plant canopy, which makes the classic genetic analysis difficult and therefore their genomes are poorly known (Dirlewanger et al., 2004). More and more researchers began to realize the importance of molecular genetics and that the genomic information which governs important traits will have great potential use in Rosaceous crops breeding. Great progress has been made during the last decade in Rosaceae with a focus on peach and apple. For example, both Europe and US initiated “Prunus Genome Project” and “Apple Genome Project”. Vast useful information has been achieved, and a Rosaceae database (www.genome.clemson.edu/gdr) has been established. However, information is still scarce on the other less studied Rosaceous crops, of which many may have great benefits to human health. Chestnut rose (Rosa roxburghii Tratt), for instance, is a rare fruit crop in Southwest China, and has recently been labeled as one of the three promising new fruit crops in China (Wen and Deng, 2005) due to its fruits having both high content of vitamin C (2000–3000 mg/100 g FW) and high levels of superoxide dismutase (SOD) activity (Ma et al., 1997). Unfortunately, chestnut rose is severely affected by powdery mildew, which is one of the most damaging diseases of Rosaceous fruit crops worldwide. So far the powdery mildew resistance loci have been mapped in apple (Calenge et al., 2005), peach (Dirlewanger et al., 1996, Dirlewanger et al., 2004, Quarta et al., 2000, Dettori et al., 2001, Lalli et al., 2005), and rose (Linde et al., 2004). If there is synteny and collinearity between different species within Rosaceae for the genomic region encompassing powdery mildew resistance loci already mapped in species such as apple and peach, such genetic information could be cross-utilized to facillitate mapping and future cloning of powdery mildew R genes in less studied Rosaceous fruit crops such as chestnut rose.

In this study, 228 RGH genes in Rosaceae genome were collected and used for evolutionary analysis. The phylogeny was comparatively analyzed both on species and genus level, suggesting that gene duplication followed by divergence may have occurred after the radiation of species or genus. Positively selected sites within NBS-coding region were also detected and NBS domain is believed to function in determining disease resistance specificity. This study also discusses the synteny of a genomic region that encompass powdery mildew resistance locus among Malus, Prunus, and Rosa.