Analysis of non-TIR NBS-LRR resistance gene analogs in Musa acuminata Colla: isolation, RFLP marker development, and physical mapping

Abstract

Background: Many commercial banana varieties lack sources of resistance to pests and diseases, as a consequence of sterility and narrow genetic background. Fertile wild relatives, by contrast, possess greater variability and represent potential sources of disease resistance genes (R-genes). The largest known family of plant R-genes encode proteins with nucleotide-binding site (NBS) and C-terminal leucine-rich repeat (LRR) domains. Conserved motifs in such genes in diverse plant species offer a means for isolation of candidate genes in banana which may be involved in plant defence.

Results: A computational strategy was developed for unbiased conserved motif discovery in NBS and LRR domains in R-genes and homologues in monocotyledonous plant species. Degenerate PCR primers targeting conserved motifs were tested on the wild cultivar Musa acuminata subsp. burmannicoides, var. Calcutta 4, which is resistant to a number of fungal pathogens and nematodes. One hundred and seventy four resistance gene analogs (RGAs) were amplified and assembled into 52 contiguous sequences. Motifs present were typical of the non-TIR NBS-LRR RGA subfamily. A phylogenetic analysis of deduced amino-acid sequences for 33 RGAs with contiguous open reading frames (ORFs), together with RGAs from Arabidopsis thaliana and Oryza sativa, grouped most Musa RGAs within monocotyledon-specific clades. RFLP-RGA markers were developed, with 12 displaying distinct polymorphisms in parentals and F1 progeny of a diploid M. acuminata mapping population. Eighty eight BAC clones were identified in M. acuminata Calcutta 4, M. acuminata Grande Naine, and M. balbisiana Pisang Klutuk Wulung BAC libraries when hybridized to two RGA probes. Multiple copy RGAs were common within BAC clones, potentially representing variation reservoirs for evolution of new R-gene specificities.

Conclusion: This is the first large scale analysis of NBS-LRR RGAs in M. acuminata Calcutta 4. Contig sequences were deposited in GenBank and assigned numbers ER935972 - ER936023. RGA sequences and isolated BACs are a valuable resource for R-gene discovery, and in future applications will provide insight into the organization and evolution of NBS-LRR R-genes in the Musa A and B genome. The developed RFLP-RGA markers are applicable for genetic map development and marker assisted selection for defined traits such as pest and disease resistance.

Figure 1

Computational protocol for primer design targeting motifs in non-TIR NBS and LRR domains in monocotyledons.

Figure 2

Bayesian phylogenetic analysis of NBS-LRR amino acid sequences from M. acuminata Calcutta 4, O. sativa and A. thaliana. The majority rule consensus tree was derived from analysis of a common NBS region between the kinase 2 and GLPL motifs, and included 33 M. acuminata Calcutta 4 sequences, together with 21 representative non-TIR NBS-LRR domain sequences from A. thaliana and 43 from O. sativa. Clade numbers are included to facilitate discussion of data. All additional information for Musa tree sequences are summarised in Table 3. The branch lengths are proportional to the average number of amino acid substitutions per site, as indicated by the scale.

Figure 3

Multiple loci polymorphisms observed in M. acuminata parentals with RGA genetic markers. Polymorphisms were observed in DraI, HindIII, and EcoRV-digested genomic DNA from M. acuminata spp. microcarpa genetic map parentals Borneo and Pisang Lilin, following hybridization of Southern blots with RGA probes MaRGA08 (panel A) and MaRGA37 (panel B).

Figure 4

Segregation of polymorphic bands in a subset of M. acuminata mapping population F1 progeny. Hybrization of RGA probes MaRGA12 (Panel A) and MaRGA37 (Panel B) onto parentals and F1 progeny. P1: Pisang Lilin; P2: Borneo; and lanes 1 to 28: individual F1 plants. Segregating bands selected for mapping from P1 and P2 are indicated by black and white arrowheads, respectively.

Figure 5

RFLPs (A) and re-validated BAC clones (B) in M. balbisiana, identified with probe MaRGA08. Panel A shows a fingerprinting gel of BAC clones digested with EcoRI, stained with ethidium bromide. Panel B shows results of hybridization of the Southern blot from panel A with radiolabelled probe MaRGA08.

Figure 6

Conserved motifs in non-TIR NBS and LRR domains targeted using degenerate RGA primers. The arrows show primer positions, with tips indicating the 3' primer end. The scheme is not to scale.