In view of modern developments in the technologies available for breeding potatoes for resistance to virus diseases, it is timely to review the host major genes that confer resistance, in Solanum species, to potato viruses X, Y, A and V (the viruses for which the resistance genes have been most extensively studied). Over the course of 60 years, many such genes in Solanum species have been characterized: a comprehensive list is presented. Inheritance studies are reviewed, including linkage studies and molecular mapping, and the positions of resistance genes mapped so far are listed. It is apparent from recent research that disease resistance genes are often clustered in particular regions of the chromosomes; the signi®cance of these resistance gene clusters is discussed. The information presented will be useful for potato breeding, and for genetic and mapping studies and gene cloning.

Tetraploid cultivated potato (Solanum tuberosum) is the World's fourth most important crop and has been subjected to much breeding eort, including the incorporation of resistance to viruses. Several new approaches, ideas and technologies have emerged recently that could aect the future direction of virus resistance breeding. Thus, there are new opportunities to harness molecular techniques in the form of linked molecular markers to speed up and simplify selection of host resistance genes. The practical application of pathogen-derived transgenic resistance has arrived with the ®rst release of GM potatoes engineered for virus resistance in the USA. Recently, a cloned host virus resistance gene from potato has been shown to be eective when inserted into a potato cultivar lacking the gene. These and other developments oer great opportunities for improving virus resistance, and it is timely to consider these advances and consider the future direction of resistance breeding in potato. We review the sources of available resistance, conventional breeding methods, marker-assisted selection, somaclonal variation, pathogen-derived and other transgenic resistance, and transformation with cloned host genes. The relative merits of the dierent methods are discussed, and the likely direction of future developments is considered.

The R10 and R11 late blight differentials of Black (tetraploid clones 3681ad1 and 5008ab6) were crossed with the susceptible potato (Solanum tuberosum) cultivar Maris Piper and the progeny were assessed for blight resistance in a whole plant glasshouse test using race 1,2,3,4,6,7 of Phytophthora infestans. The disease scores for the R10 population displayed a continuous distribution whereas the progeny in the R11 population could be categorised as resistant or susceptible. A bulk segregant analysis using amplified fragment length polymorphism assays was done on the ten most resistant and ten most susceptible progeny in each population and two closely linked markers were found to be associated with resistance. R11 mapped to 8.5 cM from marker PAG/MAAG_172.3 and R10 mapped as a quantitative trait locus in which marker PAC/MATC_264.1 explained 56.9% of the variation in disease scores. The results were consistent with R10 and R11 being allelic versions of genes at the R3 locus on chromosome 11. The implications are discussed for mapping R-genes which fail to give complete immunity to a pathogen.

Twelve potato clones were exposed to infection by aphids with potato leafroll luteovirus (PLRV) in three field trials in order to assess their resistance to infection. Up to 92% of the plants of some clones became infected, although other clones were relatively resistant to infection and one clone remained virusfree in all three trials. The resistance of the same 12 clones to PLRV multiplication was assessed in glasshouse-grown plants: plants were graft-inoculated and their daughter tubers were used to grow plants with secondary infection. High concentrations of PLRV were found in some clones (c. 1700 ng/g leaf) while in others much less virus accumulated (as little as 60 ng/g leaf). However, clones in which little virus accumulated were not necessarily those which were most resistant to infection in the field, and there was no association between the two types of resistance. Nevertheless, both types of resistance were found in some clones. The clone G8107(1), which remained virus-free in all the field exposure trials, was also the most resistant to PLRV multiplication. The combination of these two types of resistance in cultivars should help to eliminate the spread of PLRV in crops.

In an experimental breeding scheme to improve late blight (Phytophthora infestans) and white potato cyst nematode (Globodera pallida) resistance of tetraploid potato over three generations of crossing and selection, 15 clones survived the final selection, and these were derived from 15 great-grandparents. There was no direct selection for resistance to Potato virus Y (PVY), but 14 out of the 15 greatgrandparents were resistant to PVY and three had extreme resistance. Thirteen of the 15 descendants had PVY resistance and one extreme resistance. This was within the range expected for a random (unselected) sample from the genotypes of the greatgrandparents. Hence, we found no evidence for any positive or negative association between PVY resistance and the attributes selected. The conclusion is that laborious selection is not required in every generation when many parents have PVY resistance, including some with more than one copy of a PVY resistance gene or resistance at more than one locus. However, in the future, determining the major virus resistance genes present in potential parents in each generation using diagnostic molecular markers would prevent susceptible × susceptible crosses being made and maximise the number of resistant × resistant ones.