Some inter-and intraspecific crosses may result in reduced viability or sterility in the offspring, often due to genetic incompatibilities resulting from interactions between two or more loci. Hybrid necrosis is a postzygotic genetic incompatibility that is phenotypically manifested as necrotic lesions on the plant. We observed hybrid necrosis in interspecific lettuce (Lactuca sativa and Lactuca saligna) hybrids that correlated with resistance to downy mildew. Segregation analysis revealed a specific allelic combination at two interacting loci to be responsible. The allelic interaction had two consequences: (1) a quantitative temperature-dependent autoimmunity reaction leading to necrotic lesions, lethality, and quantitative resistance to an otherwise virulent race of Bremia lactucae; and (2) a qualitative temperature-independent racespecific resistance to an avirulent race of B. lactucae. We demonstrated by transient expression and silencing experiments that one of the two interacting genes was Rin4. In Arabidopsis thaliana, RIN4 is known to interact with multiple R gene products, and their interactions result in hypersensitive resistance to Pseudomonas syringae. Site-directed mutation studies on the necrosis-eliciting allele of Rin4 in lettuce showed that three residues were critical for hybrid necrosis.

The nonhost resistance of wild lettuce (Lactuca saligna) to downy mildew (Bremia lactucae) is based on at least 15 quantitative trait loci (QTL), each effective at one or more plant developmental stages. We used QTL pyramiding (stacking) to determine how many of these QTL from L. saligna are sufficient to impart complete resistance towards B. lactucae to cultivated lettuce, L. sativa. The alleles of four of the most promising QTL, rbq4, rbq5, rbq6+11, and rbq7 are effective at both the young and adult plant stages. Lines with these four QTL in all possible combinations were generated by crossing the respective backcross inbred lines (BIL). Using the 11 resulting lines (combiBIL), we determined that combinations of three QTL, rbq4, rbq5, and rbq6+11, led to increased levels of resistance; however, one QTL, rbq7, did not add to the resistance level when combined with the other QTL. One line, tripleBIL268, which contains the three QTL rbq4, rbq5, and rbq6+11, was completely resistant to B. lactucae at the young plant stage. This suggests that these three QTL are sufficient to confer the complete resistance of the nonhost L. saligna and any additional QTL in L. saligna are redundant. Histological analysis of B. lactucae infection in L. saligna, the BIL, and the combiBIL 48 h after inoculation revealed different microscopical phenotypes of resistance. The QTL differed with respect to the stage of the infection process with which they interfered.

In a stacking study of eight resistance QTLs in lettuce against downy mildew, only three out of ten double combinations showed an increased resistance effect under field conditions. Complete race nonspecific resistance to lettuce downy mildew, as observed for the nonhost wild lettuce species Lactuca saligna, is desired in lettuce cultivation. Genetic dissection of L. saligna's complete resistance has revealed several quantitative loci (QTL) for resistance with field infection reductions of 30-50 %. To test the effect of stacking these QTL, we analyzed interactions between homozygous L. saligna CGN05271 chromosome segments introgressed into the genetic background of L. sativa cv. Olof. Eight different backcross inbred lines (BILs) with single introgressions of 30-70 cM and selected predominately for quantitative resistance in field situations were intercrossed. Ten developed homozygous lines with stacked introgression segments (double combinations) were evaluated for resistance in the field. Seven double combinations showed a similar infection as the individual most resistant parental BIL, revealing epistatic interactions with 'less-than-additive' effects. Three double combinations showed an increased resistance level compared to their parental BILs and their interactions were additive, 'less-than-additive' epistatic and 'more-than-additive' epistatic, respectively. The additive interaction reduced field infection by 73 %. The double combination with a 'more-than-additive' epistatic effect, derived from a combination between a susceptible and a resistant BIL with 0 and 30 % infection reduction, respectively, showed an average field infection reduction of 52 %. For the latter line, an attempt to genetically dissect its underlying epistatic loci by substitution mapping did not result in smaller mapping intervals as none of the 22 substitution lines reached a similar high resistance level. Implications for breeding and the inheritance of L. saligna's complete resistance are discussed.

Three regions with quantitative resistance to downy mildew of non-host and wild lettuce species, Lactuca saligna , disintegrate into seventeen sub-QTLs with plant-stage-dependent effects, reducing or even promoting the infection. Previous studies on the genetic dissection of the complete resistance of wild lettuce, Lactuca saligna, to downy mildew revealed 15 introgression regions that conferred plant stage dependent quantitative resistances (QTLs). Three backcross inbred lines (BILs), carrying an individual 30-50 cM long introgression segment from L. saligna in a cultivated lettuce, L. sativa, background, reduced infection by 60-70 % at young plant stage and by 30-50 % at adult plant stage in field situations. We studied these three quantitative resistances in order to narrow down their mapping interval and determine their number of loci, either single or multiple. We performed recombinant screenings and developed near isogenic lines (NILs) with smaller overlapping L. saligna introgressions (substitution mapping). In segregating introgression line populations, recombination was suppressed up to 17-fold compared to the original L. saligna × L. sativa F 2 population. Recombination suppression depended on the chromosome region and was stronger suppressed at the smallest introgression lengths. Disease evaluation of the NILs revealed that the resistance of all three BILs was not explained by a single locus but by multiple sub-QTLs. The 17 L. saligna-derived sub-QTLs had a smaller and plant stage dependent resistance effect, some segments reducing; others even promoting downy mildew infection. Implications for lettuce breeding are outlined.