The purpose of the research described in this thesis was to identify and develop molecular markers linked to specific genes conferring resistance to the sunflower fungal pathogen,
Puccinia helianthi Schw., to facilitate the production of simflower germplasm with pyramided rust resistance genes. To achieve this goal, the research required the identification of the molecular markers, localisation of the markers to a genetic linkage map to understand the genetic interactions between the genes of interest, and to evaluate the markers in a gene pyramiding situation.
F2 populations, segregating for resistance to specific races of simflower rust, were supplied in which the genetics of the resistance had been characterised. In most cases, resistance was inherited in a single dominant fashion. Arbitrary primed PCR technology (RAPDs and DAFs), together with a pooling strategy (bulked segregant analysis), identified 16 DNA markers linked to nine specific rust resistance genes
of sunflower. Most of the markers were closely linked (<5.0 cM) to the respective rust resistance genes, while other markers were only associated at a distance of >15 cM.
Seven of the RAPD loci closely linked to the rust resistance genes were converted into SCAR (Sequence Characterised Amplified Region) loci: SCTO6950, SCX20660, SCO4950, SCX10400, SCP04300, SCR111500, and SCL20400, representing the rust resistance genes R1, RAdv, R4/RP1, R2, R5 and RSx53 respectively.
Dominant polymorphisms were detected for the majority of the SCAR loci amplified. Only the SCAR markers SCOO4950 and SCL20400 revealed different sized fragments in each parent to give co-dominant marker loci.
Co-localisation of the molecular markers to an existing genetic linkage map of sunflower demonstrated that two rust resistance gene clusters, on independent linkage groups, exist: (1) Sunflower linkage group 8 containing the genes R1, R2, and RAH52; and (2) Sunflower linkage group 13 containing the genes R4, RP1, RAdv, as well as additional genes not yet mapped but located to this region by their
associations with other genes within this cluster.
Consistent amplification of the R1 gene marker in lines not previously thought to contain the R1 gene suggested evidence of a repeat sequence randomly distributed throughout the sunflower genome, while comparison of the 660bp RAdv marker sequence with known sequences from a gene database, identified significant amino acid similarity between the marker sequence and the conserved motif sequences, 'CX2CX4HX5C' and 'VLFDSGA', commonly represented in refrofransposons. Evidence of sequence duplication and the presence of a transposable element associated with rust resistance provides important
information about the genomic regions that contain the resistance genes and the surrounding areas.
Evaluation of the SCAR markers in a wide range of sunflower lines, including breeding lines, rust differentials, inbreds, hybrid cultivars and wild accessions, proved that the SCARs were successful at detecting individual rust resistance loci. Few lines contained resistance genes at the R2 and R5 loci, indicating limited use of these lines in breeding programs even though the resistance exhibited by these genes is very effective to the current spectrum of virulent rust races. In contrast, the R4 locus was widely evident in the large selection of germplasm screened, supporting the wide spread usage of this locus
in commercial hybrids in Australia in the past decade.
The reliability of the SCAR markers for detecting individual resistance loci within gene pyramiding experiments was investigated using five crosses, containing different pairwise combinations of the rust resistance genes, RAdv, R2, R4, R5 and RSx53, at the F1 stage and in two two-gene crosses (R4//RAdv, and RAdv//RSx53) at F2 stage. In each experiment, the markers were successful. The markers failed to amplify in
only a few of the F1 and F2 individuals screened. The dominant nature of most of the SCAR markers will limit their usefulness in the gene pyramiding program due to their inability to differentiate homozygous resistant individuals from heterozygous individuals. Further work to identify co-dominant markers, such as SSRs or RFLPs will help alleviate this situation.
Research in this thesis contributes a library of molecular markers linked to important rust resistance genes, a valuable resource that can be integrated into a sunflower gene pyramiding program to facilitate selection of lines with specific genes. Additional information regarding the genetic linkages between the rust resistance genes, sequence duplications and refrofransposons located
around rust resistance loci contribute towards a better understanding of the possible genetic makeup and evolution of rust resistance.