Linkage map in kiwifruit: the choice of mapping populations

A genetic map of kiwifruit (Actinidia spp.) was constructed using microsatellite and AFLP markers and the pseudo-test-cross mapping strategy. Some 105 microsatellite loci, of the 251 isolated from A. chinensis, segregated in the interspecific cross A. chinensis x A. callosa in a 1:1 ratio or in a ratio that could mimic the classical backcross, and were mapped using 94 individuals of the progeny. AFLP markers were also produced using MseI and EcoRI restriction enzymes and 15 primer combinations. Two linkage maps were produced, one for each parent. Some 225 markers segregated from the female parent (86 SSR and 139 AFLP) and 168 markers/loci (35 SSR, 132 AFLP and the sex determinant) from the male parent. Disappointedly only a few SSR markers segregated from the male parent either because of lack of amplification (18%) or because of homozygosity (48%). As a result, the male map had large gaps and only 14 linkage groups were recognised has holologous in the two maps. A test of polymorphism was therefore carried out on four genotypes of A. chinensis (the SSR source species) and four species (A. callosa, A. eriantha, A. latifolia, and A. polygama) using 30 randomly selected SSRs markers. Some 25/30 SSR were successfully amplified in A. chinensis and 14 to 15 of them were heterozygous; 17 to 22 SSR were amplified in the remaining species and 4 to 11 of them were heterozygous. The analysis of allelic combination at the 25 SSR for all combinations of genotype pairs revealed that 79% of SSR could be mapped in A. chinensis A. chinensis pairs, 66% in A. chinensis other, and 45% in other other, where ‘other’ means any species different from A. chinensis. A similar tendency (35% vs. 19% vs. 9%) was found in the percentage of anchor loci, which is loci that map in both parents and could be used to find homologous linkage groups. These results prove that an intra-specific cross between A. chinensis genotypes should allow mapping the highest number of SSRs and should yield the highest number of anchor markers. More general aspects dealing with the selection of parents for the pseudo-test-cross and linkage maps are discussed in the paper.