Characterisation of Potential Fungal Disease Resistance Genes in Banana

Bananas are an extremely important crop, serving as both a staple food in developing countries and as a dessert fruit in Western society. Two of the most devastating pathogens currently affecting both commercial and subsistence banana production are Fusarium oxysporum (Foc; causal agent of Fusarium wilt) and Mycosphaerella species (causal agent of black and yellow Sigatoka). Conventional breeding programs designed to improve the disease resistance characteristics of the commercially elite Cavendish cultivar have, thus far, been largely unsuccessful. Genetic engineering is now regarded as the most promising method to generate enhanced disease resistance in banana. In other crops and model species, strategies to enhance disease resistance have included the transgenic expression of defense-related genes such as; disease resistance genes (R genes), downstream signaling genes (eg. NPR1, non-pathogenesis related) and antimicrobial peptides (AMPs). The overall aims of this research were to amplify and compare the nucleotide binding site (NBS) domains of potential disease resistance genes from disease resistant and disease susceptible banana cultivars. To isolate and compare complete R gene sequences from these cultivars. To generate transgenic Lady Finger banana plants expressing the D4E1 antimicrobial peptide under the control of two different promoters and finally to assess extracts from these plants for their ability to inhibit the growth of Foc Race1. Using degenerate primers, the NBS domains of six resistance gene candidate (RGC) sequences were amplified from the disease resistant cultivar Calcutta 4 (C4) and the disease susceptible cultivar Cavendish (Cav). The RGC 1, 2, 5 and 6 sequences showed similarity to previously characterized R gene sequences isolated from monocotyledonous plant species, while RGCs 3 and 4 showed similarity to R genes which form part of the Fusarium wilt resistance locus isolated from the dicotyledonous species, Lycopersicon esculentum; as well as other monocotyledonous R genes. RGCs 1-4 and 6 were present and transcriptionally active in both C4 and Cav, whereas RGC-5 was present in Cav only and was not transcribed. The transcripts could not be detected by Northern analysis, which is consistent with previous reports that R genes are constitutively transcriptionally active at only low levels. The NBS domains of RGCs 1-6 showed less than 65% similarity (amino acid level) to one another but when each individual RGC isolated from the C4 and Cav gDNA and cDNA templates was compared the sequences showed greater than 97% similarity (amino acid level). Comparative sequence analysis revealed amino acid positions that were consistently different between the C4 and Cav clones. Southern analysis revealed that RGC 1-5 were present in both the C4 and Cav genomes in only low copy number (1-2 gene copies with 1-3 alleles), whereas RGC-6 showed high copy number in both cultivars. Complete RGC sequences were subsequently amplified by RNA-ligase-mediated (RLM) -RACE and 3'-RACE using specific primers designed to each of the RGC 1-4 NBS domains. Amplicons for each RGC were assembled to form potentially complete RGC sequences. Analysis of the sequences revealed the presence of coiled coil (CC) motifs in two of the amino terminal sequences while leucine rich repeats (LRRs) were identified at the carboxy terminal of all sequences. Multiple 3'-RACE products were amplified for each RGC sequence. Although the polyadenylated products were of different lengths, the sequences were greater than 98% identical at the amino acid level (except an RGC 3 clone which was 91-95% identical to the other RGC 3 clones due to a 37 amino acid deletion). Specific primers used to amplify each complete RGC sequence from both C4 and Cav DNA revealed that: RGC 1 (3.53 kbp) could be amplified from both C4 and Cav; RGCs 2 (2.99 kbp) and 4 (4.44 kbp) could be amplified from only Cav, however, the proposed truncations of these sequences (RGC 2: 1.3 kbp, RGC 4: 2.8 kbp and 2.9 kbp) could be amplified from both cultivars; RGC 3 (4.57 kbp) could not be amplified from either C4 or Cav, however, the three shorter sequences (1.96 kbp, 1.34 kbp and 1.28 kbp) could be amplified from both templates. The functional significance of the truncated sequences is currently unknown, however, truncated sequences have been detected in a number of R gene families isolated from other crops. No major sequence differences, such as deletions/insertions or early stop codons, were identified between the RGC sequences amplified from C4 as compared to Cav (greater than 91% amino acid similarity) and no sequence was identified as being present in the susceptible but absent from the resistant cultivar. However, comparative analysis of multiple clones isolated from C4 and Cav did reveal amino acid residues that were consistently different between the two cultivars. These differences may result in differing resistance capabilities, functional genomics studies would need to be undertaken to determine this. It has been proposed that CC-NBS-LRR type R genes employ NDR1/HIN1-like (NHL) proteins, after pathogen invasion is detected, in the signaling process that ultimately leads to the elaboration of a defense response. A NHL partial sequence (420 bp) was amplified from the C4 banana cultivar. The complete sequence of this gene (termed NHL-1) was isolated using RLM and 3'-RACE technologies (576 bp and 535 bp amplicons, respectively) and subsequently the 1.106 kbp sequence was PCR amplified from both the C4 and Cav cultivars. The banana NHL-1 gene contained conserved motifs/domains previously identified within other NHL-type gene sequences. These included a signal peptide motif, a transmembrane domain and three previously identified conserved motifs. Based on current research into NHL type genes, the banana NHL-1 sequence may not be useful as a transgene to enhance disease resistance in elite cultivars. However, it potentially plays an important role in the defense response signal transduction pathway and therefore will further our understanding of plant-pathogen interactions in banana. Transgenic Lady Finger banana plants expressing the D4E1 antimicrobial peptide under the control of either the maize polyubiquitin (Ubi) or banana bunchy top virus (BBTV) DNA-6 (Bt6.1) promoters were generated. These plants were subsequently assessed for the ability of their crude protein extracts to inhibit the germination of Fusarium oxysporum f.sp. cubense Race1 conidia in vitro. These anti-fungal bioassays revealed that fungal colony growth was reduced by 37-100% using extracts from the pUbi-D4E1 transgenic lines and 89-99% using extracts from the pBt6.1-D4E1 transgenic lines. The transgenic lines are currently undergoing multiplication in preparation for glasshouse and small plant challenge trials for resistance to Fusarium wilt. These preliminary results suggest that D4E1 may be useful in enhancing disease resistance in banana