The genetics of non-host resistance to the lettuce pathogen Bremia lactucae in Lactuca saligna

Plants are continuously exposed to a wide variety of pathogens. However, all plant species are non-hosts for the majority of the potential plant pathogens. The genetic dissection of non-host resistance is hampered by the lack of segregating population from crosses between host and non-host species, since hardly any non-host is crossable with a host. We have studied the non-host resistance in Lactuca saligna (wild lettuce) to lettuce downy mildew ( Bremia lactucae ). L. saligna is one of the few examples of a non-host species that is crossable with a related host species, L. sativa (lettuce). Based on this interspecific cross, segregating populations have been developed for genetical analysis of the non-host resistance. To map the resistance, we have used two strategies in which we make use of DNA markers to genotype plants. As no accurate linkage map was available for lettuce, we started with the construction of a linkage map of L. saligna´L. sativa . In Chapter 2, the development of an integrated linkage map, based on two populations, is described. To acquire DNA markers, AFLP analyses have been performed on the F 2 populations of the crosses L. saligna CGN 5271 ´L. sativa Olof and L. saligna CGN 11341´L. sativa Norden. Based on these AFLP analyses the polymorphism rate between L. saligna and L. sativa is estimated to be 81%. A linkage map was constructed that comprises 12 SSRs and 476 AFLP markers over 854 cM in nine linkage groups (n=9). Since the markers are randomly spread over all chromosomes, we assume this map is an accurate representative of both parental genomes and very useful for Marker Assisted Selection.The first mapping strategy for downy mildew resistance is described in Chapter 3. In that study, we have performed a QTL analysis on 126 F 2 plants of a cross between the resistant L. saligna CGN 5271 and the susceptible L. sativa Olof. For this QTL analysis all 126 F 2 plants have been tested for resistance in four disease tests with two complex Bremia races (NL14 and NL16). The F 2 population showed a wide and continuous range of resistance levels from completely resistant to completely susceptible. Evidence is presented for a quantitative resistance against both Bremia races as well as for a race-specific resistance against Bremia race NL16 and not against NL14. These disease test data sets have been combined with DNA marker data of all 126 F 2 plants that had already been obtained for the construction of the linkage map. QTL mapping revealed a qualitative gene ( R39 ) explaining the race-specific resistance and three QTLs ( RBQ1 , RBQ2 and RBQ3 ) explaining the quantitative resistance. The qualitative gene R39 is a dominant gene that gives nearly complete resistance to race NL16 in L. saligna CGN 5271 and therefore it shows features similar to Dm genes (dominant race specific genes that give a complete resistance to d owny m ildew). The three QTLs explain 51% of the quantitative resistance against NL14, which indicates that probably not all QTLs have been detected in this F 2 population.In addition to this rather classical F 2 mapping strategy, we have performed an alternative mapping strategy based on the development and characterization of a set of Backcross Inbred Lines (BILs). These BILs are genetically nearly completely like L. sativa but contain a single chromosome substitution segment of L. saligna CGN 5271 (Chapter 4). Starting from an F 1 plant, BILs have been developed by four to five generations of backcrosses and one generation of selfing. All backcrosses from F 1 to BC 4 were made randomly without intentional selection. Marker Assisted Selection was started in the BC 4 generation. Finally, a set of 29 lines was obtained that covers 95% of the L. saligna genome, comprising 16 lines with a single homozygous introgression (BILs), one line with two homozygous introgressions, five lines with heterozygous single introgressions and seven lines with two or more heterozygous introgressions. Several chromosome regions showed severe distorted segregation in the F 2 population. Based on segregation ratios in backcross lines, we were able to explain distorted segregations of three chromosome regions observed in the F 2 population by genetic loci that are involved in pollen- or egg cell fitness.When seed of the first developed BILs was available, a disease test had been set up to test if the BILs, which carried QTLs as identified in the F 2 population, showed enhanced levels of quantitative resistance indeed. Nine BILs (or nearly-BILs) have been tested for resistance to Bremia race NL16. They covered together 31% of the L. saligna parental genome. Two resistance loci detected in the F 2 population ( R39 and RBQ3 ) have been confirmed in the disease test on the BILs. R39 is a dominant gene, which gives a complete resistance against Bremia race NL16. RBQ3 reduces the infection severity of the susceptible L. sativa by 49% ten days post inoculation. The quantitative effects from the resistance genes in these BILs were higher than expected from the F 2 mapping results. No conclusive comparisons of RBQ2 could have been made, as the introgression in the backcross line was not homozygous. RBQ1 has not been tested. Most exciting, the BIL method revealed a new resistance locus on Chromosome 8 with a 77% reduction on the infection severity compared to the susceptible control ten days post inoculation. We conclude that the BIL mapping method can reveal new QTLs unnoticed in the F 2 mapping method and it enables a quantification of the resistance gene effect in a L. sativa background.To extend our knowledge about the non-host resistance of L. saligna to Bremia , we have compared the genetics of non-host resistance to Bremia in L. saligna CGN 5271 with another accession L. saligna CGN 11341. The two accessions show a 39% AFLP polymorphism rate. We have analyzed the non-host resistance of L. saligna CGN 11341 by disease tests and DNA marker analyses on an F 2 and BC 1 population. Disease tests with Bremia races NL14 and NL16 showed a wide range of infection severity scores from resistant to susceptible to both races. The majority of plants had a similar resistance level to both Bremia races. These findings imply that the resistance of L. saligna is quantitatively expressed and is probably race non-specific. A few F 2 and BC 1 plants were completely resistant against Bremia race NL16 and rather susceptible to race NL14. QTL mapping revealed that a major resistance gene that was located on Chromosome 9 explains this race-specific resistance. This gene is designated R39b , as it may be different from R39 .No additional QTLs have been detected in this small F 2 population (n= 54). However, F 2 plants with L. saligna CGN 11341 alleles at loci of RBQ1 , RBQ2 , RBQ3 and RBQ4 mapped in L. saligna CGN 5271, were more resistant than F 2 plants with L. sativa alleles at these loci. In conclusion, we state that it is very likely that the same genes explain the resistances to Bremia in both L. saligna accessions. A backcross program for a set of Backcross Inbred Lines (BIL) that cover R39b and loci for putative QTLs, is in progress.In the last chapter of this thesis the basic results of the study have been discussed. We adduce that non-host resistance in L. saligna is not explained by accumulation of race-specific major resistance genes ( Dm genes) but by a resistance mechanism based on QTLs. Further, we have made a comparison for efficiency of four breeding methods to introgress the resistance genes from L. saligna . Based on this study, we conclude that twice as many resistance genes are introgressed when Marker Assisted Selection is used. Finally, several recommendations concerning research on non-host resistance and the applications of Backcross Inbred Lines have been suggested.