Search

Map-based cloning of the avirulence gene AvrLm1 of Leptosphaeria maculans was initiated utilizing a genetic map of the fungus and a BAC library constructed from an AvrLm1 isolate. Seven polymorphic DNA markers closely linked to AvrLm1 were identified. Of these, two were shown to border the locus on its 5′ end and were present, with size polymorphism, in both the virulent and the avirulent isolates. In contrast, three markers, J19-1.1, J53-1.3 (in coupling phase with avirulence), and Vir1 (in repulsion phase with avirulence), cosegregated with AvrLm1 in 312 progeny from five in vitro crosses. J19-1.1 and J53-1.3 were never amplified in the virulent parents or progeny, whereas Vir1 was never amplified in the avirulent parents or progeny. J19-1.1 and J53-1.3 were shown to be separated by 40 kb within a 184-kb BAC contig. In addition, the 1.6-cM genetic distance between J53-1.3 and the nearest recombinant marker corresponded to a 121-kb physical distance. When analyzing a European Union-wide collection of 192 isolates, J53-1.3, J19-1.1, and Vir1 were found to be closely associated with the AvrLm1 locus. The results of polymerase chain reaction amplification with primers for the three markers were in accordance with the interaction phenotype for 92.2% (J53-1.3), 90.6% (J19-1.1), and 88.0% (Vir1) of the isolates. In addition, genome organization of the AvrLm1 region was highly conserved in field isolates, because 89.1% of the avirulent isolates and 79.0% of the virulent isolates showed the same association of markers as that of the parents of in vitro crosses. The large-scale analysis of field isolates with markers originating from the genetic map therefore confirms (i) the physical proximity between the markers and the target locus and (ii) that AvrLm1 is located in (or close to) a recombination-deficient genome region. As a consequence, map-based markers provided us with high-quality markers for an overview of the occurrence of race “AvrLm1” at the field scale. These data were used to propose hypotheses on evolution towards virulence in field isolates.

Field experiments were conducted in Versailles, France, to assess blackleg resistance of Brassica napus cultivars Quinta and Glacier under natural infection conditions. Blackleg disease severity was assessed twice during growth of B. napus. Quinta resistance was highly expressed as only 13% to 18% of the plants exhibited leaf symptoms in December, whereas Glacier and other cultivars displayed more than 80% of infected plants. In June (harvest), 70% (first year) to 41.5% (second year) of Quinta plants were canker-free. In contrast, Glacier was as infected as the susceptible control cultivars, with more than 88% of plants displaying canker. The Leptosphaeria maculans population structure was examined in parallel. Based on soluble protein patterns, 9% of the 299 fungal isolates collected were characterized as Tox0 species, and belonged to the NA1 sub-group. All but two of Tox0 isolates were isolated from atypical dark necrotic leaf lesions, mainly occurring on Quinta. In contrast, the Tox+ isolates were recovered from typical leaf lesions. Following a cotyledon inoculation test on the differential set Westar, Quinta and Glacier, 92 to 95% of Tox+ isolates collected on susceptible cultivars were characterized as PG3 isolates, i.e. avirulent on Quinta. The remaining Tox+ isolates belong to PG4, i.e. virulent on the three cultivars. No PG2 isolate, i.e. avirulent on both Quinta and Glacier, was identified in the sampling. The present study suggests that specific resistance expressed at the cotyledon level can be efficient under field conditions where the corresponding avirulent races of the pathogen are prevalent.

Leptosphaeria maculans causes blackleg of oilseed rape. Gene-for-gene interactions between race PG3 and Brassica napus cv. Quinta were related to interaction between the fungal avirulence (Avr) gene AvrLm1 and the corresponding resistance gene Rlm1. AvrLm1 isolates were aviru-lent on cvs. Doublol, Vivol, Columbus, and Capitol, and no recombinant phenotypes were observed in the progeny of two AvrLm1 × avrLm1 crosses, suggesting that all of these cultivars may possess Rlm1 or genes displaying the same recognition spectrum, or that a cluster of Avr genes is present at the Avrlm1 locus. In one cross, segregation distortion was observed at the AvrLm1 locus that could be explained by interaction between AvrLm1 and one unlinked deleterious gene, termed Del1. Incompatibility toward cvs. Jet Neuf and Darmor.bzh was governed by a single gene, unlinked to AvrLm1 or Del1. This avirulence gene was termed AvrLm4. Preliminary plant genetic analysis suggested the occurrence of a corresponding dominant resistance gene, termed Rlm4, present in the Quinta line analyzed and linked to Rlm1.

Specificity of interaction at the cotyledon stage was recently demonstrated between the blackleg pathogen, Leptosphaeria maculans, and Brassica napus. Three pathogenicity groups were distinguished, PG2 avirulent towards 'Quinta' and 'Glacier', PG3 avirulent towards 'Quinta', and PG4 virulent on the two cultivars. The genetic control of the interactions was investigated on both the pathogen and the plant. Tetrad analysis was performed following PG3 × PG4 and PG2 × PG4 crosses. 'Quinta' and 'Glacier' were crossed with the susceptible winter oilseed rape cultivar 'Score'. The analysis of F 1 , F 2 and testcross populations suggested that the incompatible interaction between 'Quinta' and PG3 isolates is conditioned by the presence of the dominant single resistance allele Rlm1 in 'Quinta' and the matching avirulence gene Avr Lm1 in L. maculans. Race-specific resistance of 'Glacier' to PG2 isolates was conditioned by the matching gene pair Rlm2/Avr Lm2. Finally, the data suggest that two avirulence genes matching two dominant loci control the 'Quinta'-PG2 interaction. The consequences of the occurrence of race-specific resistance in B. napus are discussed with respect to future breeding for blackleg resistance.

All Leptosphaeria maculans field isolates displayed a minichromosome (MC) clearly separated from the overall electrokaryotype following pulsed-field gel electrophoresis. MCs exhibited a length polymorphism ranging from 650 to 950 kb. Tetrad analyses revealed the parental inheritance of MC length polymorphism (50% of the tetrads) or else the generation of novel-sized MCs (27%), which suggested that recombination occurred between MCs. Nineteen percent of the tetrads displayed a lack of the MC band in the electrokaryotype for one or two of the four resulting genotypes. Crosses between isolates carrying or lacking MCs revealed non-Mendelian segregation and suggested that some isolates could display at least two copies of the MC. Only repeated sequences hybridising to all chromosomes were isolated from the MC. Finally, saprophytic or parasitic fitness was not modified when isolates apparently lacked the MC. All these data suggested that the L. maculans MC behaves like a 'B' chromosome.

Specific interactions of the fungal pathogen Leptosphaeria maculans with Brassica napus cultivars were observed when the cultivars were inoculated with isolates belonging to pathogenicity groups (PG) PG3 and PG4. PG3 isolates induced resistance responses on cotyledons or leaves of cv. Quinta, whereas PG4 isolates caused sporulating lesions on this cultivar. In contrast, both pathotypes caused disease symptoms on cvs. Westar and Glacier. The genetic basis of cultivar specificity was studied using tetrad analysis after in vitro crosses between one PG3 and one PG4 isolate. For the genetic study, the use of random amplified polymorphic DNA (RAPD) as genetic markers was assessed. Of 61 primers, 10 generated reproducible polymorphisms. Of these, 9 generated 18 RAPD markers displaying a 2 :2 segregation ratio within the 10 analyzed tetrads. A 2 :2 segregation ratio for avirulence/virulence to cv. Quinta also was observed in the progeny. Consequently, the single genetic locus controlling cultivar specificity on Quinta was considered the first avirulence gene described in L. maculans and was designated AvrLM1.