The practical development and application of molecular markers in barley breeding has been examined, using barley leaf rust (Puccinia hordei) resistance genes as a model system. Molecular markers can be powerful selection tools to improve plant breeding efficiency. However, to process large breeding populations with confidence, marker systems must be robust and reliable. This requires the development of practical marker systems and their validation to demonstrate the transferability of specific markers. Of particular importance is the ability to process many assays at relatively low cost, requiring rapid and high-throughput DNA extraction and marker assay systems.
Initial experiments optimised DNA extraction and polymerase chain reaction (PCR) protocols for use with barley. A modified extraction protocol using a CTAB-based buffer produced adequate yields of good quality DNA from seedling
leaf tissue. An alternative protocol with a different buffering system was used in later experiments. More samples could be processed with the latter system, however lower yields of lesser quality DNA were produced.
Titration experiments using PCR and random oligonucleotide primers demonstrated that Mg++ concentration and the balance of template and primer DNA were key factors for consistent amplification of random amplified polymorphic DNA (RAPD) products from barley. Validation experiments demonstrated the effectiveness of the protocols and RAPD data were used to resolve issues of genotype identification and demonstrate non-selective marker uses.
An F2 population from a cross between Q21861, a barley accession with multiple disease resistances, and Galleon, a rust susceptible cultivar, was screened for seedling resistance to barley leaf rust and stem rust (P. graminis f sp.
tritici). F3 progeny lines were also screened for leaf rust reaction. Segregation patterns supported the presence in Q21861 of one dominant and one recessive stem rust resistance gene, likely to be Rpg1 and rpg4, respectively. A single dominant leaf rust resistance gene from Q21861 was designated RphQ.
Preliminary RAPD-PCR screening of the parents indicated sufficient polymorphism for marker development. Bulk segregant analysis (BSA) of 8-plant DNA bulks from resistant and susceptible F2 plants produced a 2.7 kb polymorphic marker using the Operon Inc. RAPD primer OPU-02. Analysis of individual DNA samples indicated linkage with RphQ at a distance of 12.0 cM. The marker was designated OPU-022700.
A backcross derived breeding population (TG306-5*2/Q21861) made up of three
BC1F2 derived families was selected to validate OPU-022700. It was chosen as representative material in which DNA markers might be used for pyramiding leaf rust resistance genes. The population was expected to segregate for RphQ, Rph12 and possibly Rph4.
A series of experiments using marker assays and phenotyping with leaf rust pathotypes virulent to Rph12, Rph4 and Rph2 confirmed linkage between OPU-022700 and RphQ in the validation population. However, OPU-022700 could not be reliably assayed in a second laboratory. Linkages between RphQ and the RAPD markers OPAH-03600 and OPI-O42200, identified in North America as linked to leaf rust resistance in Q21861, were also
validated. However, OPI-042200 proved to be unreliable when assayed in our laboratory.
Candidate simple sequence repeat (SSR) markers in the vicinity of Rph2 were identified from consensus maps and assessed for parental polymorphism. Linkage between RphQ and the markers Bmag387 and Bmac96 was established, placing the Q21861 resistance gene on chromosome 5H and in the vicinity of Rph2. Combined marker and phenotypic data from these experiments with additional phenotypic data from Rph2 virulent leaf rust pathotypes supported a North American conclusion that RphQ was probably allelic to Rph2.
Candidate SSR markers for Rph12 (chromosome 5HL) and Rph4 (chromosome IHS) were identified. Two markers on 5HL, GMS61 and GMS27, showed linkage to each other but not
Rph12, possibly because of the small size of the single BC1F2 derived BC1F4 family segregating for that gene. However, GMS61 and GMS27 showed common inheritance with Rph12 in the three BC-derived families, indicating potential association. All candidate SSR markers for Rph4 (1HS) were monomorphic, however the gene's presence in the BC-derived population was confirmed by phenotypic assessment.
The information generated through the experiments was useful at the practical plant breeding level. The marker data confirmed distorted inheritance of chromosome 5H regions in the validation population, which could have confused interpretation of Q2186rs resistance genetics. The data were also used to determine the chromosomal order of 10 loci and construct graphical genotypes of chromosome 5H in the backcross derived
breeding lines. This enabled decisive interpretation of the inheritance of 5H segments in the population. This conclusion was supported by use of the data to accurately predict rust resistance patterns in a set of BC1F7 lines.