2.2.76
MARKERS FOR DETECTING SOMATIC HYBRIDS OF PHYTOPHTHORA INFESTANS: AFLP VERSUS RG57

AI PURVIS, DS SHAW and SJ ASSINDER

School of Biological Sciences, University of Wales, Bangor, Gwynedd LL57 2UW, UK

Background and objectives
Genotypic diversity within populations of the late-blight fungus, Phytophthora infestans has increased recently in the UK and other countries following migration within diseased tubers from Mexico in 1976. Although the A2 mating type was first introduced at this time, its frequency in the UK has remained low; it has been detected in only 5-10% of sampled sites (4% of isolates) over the past five years, although where it occurs, its frequency may exceed 50%. The opportunity for sexual reproduction to generate variation is thus strictly limited. A further explanation for the continued generation of new genotypes is somatic hybridisation. Early work suggested that somatic recombination could occur in vitro and on host tissue [1] but this could not be confirmed without molecular data. We are investigating the possibility that somatic recombination, as recombination of chromosomal genes following heterokaryon formation or re-assortment of nuclear and mitochondrial genes, occurs both in vitro and in the field. To this end we are investigating with traditional RFLPs and with AFLPs the levels of variation detectable within a group of 98 isolates and aim to demonstrate recombination of variable bands between suitable isolates through the parasexual cycle. The results of the initial study will be presented.

Materials and methods
Isolates were selected on the basis of their DNA fingerprint pattern using RFLP probe RG57 and mitochondrial haplotype. A range of isolates from around the UK with different RG57 patterns and mitochondrial haplotypes were selected. AFLP fingerprinting was performed on the selected isolates as described in [2]. Two primer combinations (E19-M40 and E19-M16) were used for the selective amplification. Polymorphic bands were scored as presence (1) or absence (0) for each sample. NTSYS-PC was used to construct two UPGMA trees with all individuals; one based on the total RG57 data and another from the polymorphic bands from the AFLP data.

Results and conclusions
Nine bands out of 25 were polymorphic within the sample of 98 isolates using probe RG57. Using two pairs of AFLP primers, approximately 150 bands were amplified within the total sample and 24 of these were polymorphic.

The RG57 tree splits the sample on the basis of their mitochondrial type. The majority of samples with type Ia mitochondria appeared in one clade and those with type IIa clustered tightly in another. Two loci seem to be important in determining this division. The same pattern is recreated with the AFLP data but an additional level of resolution can be seen as a separate clade consisting of the A2 isolates. The split dividing mating type and that dividing mitochondrial type are each dependent on seven loci from the group of 24.

Seventeen isolates of a common RG57 genotype (genotype 39) have been collected from around Britain (North England, Scotland and East Anglia). With AFLPs, 12 of these were grouped together in three closely related clusters. However, the remaining five had completely different AFLP patterns from the main body of the genotype 39 group, questioning the assumption that isolates with the same RG57 fingerprint belong to a single clonal lineage. A total of 10 isolates with an appropriate number of polymorphic bands have been selected from the AFLP study to further investigate the inheritance of bands through parasexual recombination.

References
1. Malcolmson J, 1970. Nature 225, 971-72.
2.Vos P, et. al. 1995. Nucleic Acids Research 23, 4407-4414.