Diversity and evolution of resistance genes in tuber-bearing Solanum species

Abstract

Potato (Solanum tuberosum L.) is a crop with a large secondary gene pool, which contains many important traits that can be exploited in breeding programs. As late blight is one of the biggest problems in potato growing areas, the crop needs a large number of applications of fungicides to be able to grow in north–western Europe. There is a strong focus on resistance breeding. This thesis describes the use of nucleotide binding site (NBS) profiling to study Solanum systematics and to identify resistance gene markers, which might be applied by the breeding companies. It also studies in depth the diversity and evolution of two late blight resistance (R) genes Rpi-blb1 and Rpi-blb2 in a wide range of Solanum species. Chapter 2 evaluates the potential of NBS profiling for phylogeny reconstruction in a set of over 100 genebank accessions, representing 47 tuber-bearing Solanum species. Results from NBS profiling are compared with those from amplified fragment length polymorphism (AFLP). Cladistic and phenetic analyses show that the two techniques deliver trees with a similar topology and resolution, indicating that NBS profiling can be an alternative for phylogeny reconstruction. No clear effects of targeting resistance genes are observed in the NBS profiling tree. Within the group of tuber-bearing Solanum species, 91% of the intraspecific fragments from the co-migrating bands have sequence identity higher than 95%, indicating that homoplasy is limited (Chapter 3). Chapter 3 presents the changes in genetic diversity at resistance gene loci in a set of 456 European potato cultivars during the last 70 - 80 years. The genetic diversity at these loci increased slightly, which most likely reflects the breeding efforts to introgress resistances from wild species into cultivated potato. Several candidate R-gene markers are identified by linking NBS profiling markers with pedigree and phenotypic data of the cultivars. As homoplasy in NBS profiling markers is low, the markers could also be linked to tuber-bearing Solanum species that had contributed the resistance marker to the set of cultivars. One of the markers identified is very likely introgressed from Solanum vernei, as indicated by the presence of the marker in S. vernei accessions and in cultivars that have S. vernei in their pedigree. The marker also correlates to the resistance data from the cultivars involved. Chapter 4 describes the allele mining of two late blight R-genes, Rpi-blb1 and Rpi-blb2, originally derived from S. bulbocastanum. It also analyzes the structure of the cluster that contains Rpi-blb1, by determining the presence or absence of the genes flanking Rpi-blb1 (RGA1-blb and RGA3-blb). A wide range of Solanum species was screened for the presence of RGA1-blb and it was found to be present and highly conserved not only in all the tested tuber-bearing Solanum species but also in the non-tuber-bearing species S. etuberosum, S. fernandezianum and S. palustre, suggesting that RGA1-blb was already present before the divergence of tuber-bearing and non-tuber-bearing Solanum species. The allele frequency of RGA3-blb is, however, much lower. Highly conserved Rpi-blb1 (>99.5%) homologues are discovered not only in S. bulbocastanum but also in Solanum stoloniferum, a distinct tetraploid species from the series Longipedicellata. A number of dominant R-genes (Rpi-sto1, Rpi-plt1, Rpi-pta1 and Rpi-pta2) are identified in several F1 populations, derived from the relevant late blight resistant parental genotypes harboring the Rpi-blb1 homologue. Furthermore, Rpi-sto1 and Rpi-plt1 reside at the same position on chromosome VIII as Rpi-blb1. We propose that the above four genes share the same ancestry with Rpi-blb1 from S. bulbocastanum. Segregation data also indicates that an additional unknown late blight resistance gene is present in three of the four segregating populations. In contrast to Rpi-blb1, Rpi-blb2 is not detected in the examined set of material. Allele frequency and allelic diversity of Rpi-blb1 and Rpi-blb2 is analyzed in accessions from S. bulbocastanum and the closely related species S. cardiophyllum (Chapter 5). Highly conserved Rpi-blb1 alleles are found in 24 Mexican accessions, but not in material originating from Guatemala. Sequence analysis of a randomly selected set of genotypes reveals 19 Rpi-blb1 haplotypes. Our results confirm that Rpi-blbl belongs to the type II class of resistance genes that evolve slowly (Chapter 4 and 5). Sequences of all putative susceptible Rpi-blb1 are identical, suggesting that a single mutation event generates this allele. Rpi-blb2 is present in only eight S. bulbocastanum accessions but not in other wild species examined. This, taken together with the fact that all the Rpi-blb2 alleles examined are identical, suggests that Rpi-blb2 has evolved recently (Chapter 4 and 5). Chapter 6 discusses findings obtained from this study in a context of systematics and evolution. The Rpi-blb1 gene is originally discovered and cloned from S. bulbocastanum, a species that cannot be crossed with the cultivated potato S. tuberosum directly. Our study shows that functional homologs of Rpi-blb1 are also present in S. stoloniferum, a species that can be crossed with cultivated potato directly. So, the Rpi-sto1 gene from S. stoloniferum should be easier to introduce into cultivated potato than the Rpi-blb1 gene from S. bulbocastanum. We anticipate that for other resistance genes present in primitive species, a similar situation may exist, i. e. homologs being also present in more advanced species that can be more easily used for breeding. Therefore, before starting a potato breeding program in a species that does not allow an immediate cross with cultivated potato, evaluation of directly crossable germplasm for the presence of that gene may speed up the breeding program and save valuable time and money.