Department of Plant Breeding, Wageningen University, 6708 PB Wageningen, THE NETHERLANDS.
Tomatoes are part of the Solanaceae family which comprises over 3000 species. Around half of this family is represented by the genus Solanum which includes amongst others tomato, aubergine and potato. Resources to study tomato plants are developed worldwide to elucidate the relationship between genotypes and phenotypes and this made tomato a suitable model to study crop plants.
Since a few years, genotyping is no longer perceived as a major bottle neck in plant molecular breeding and this also holds for tomato. Genome (re)sequencing projects are delivering large numbers of informative markers such as single nucleotide polymorphisms (SNP). We used a set of 5528 SNPs to evaluate the tomato germplasm. Genotyping different tomato samples allowed the evaluation of the level of heterozygosity and the number of introgressions in commercial varieties. We found relatively more differences between cherry and round/beef tomatoes on chromosomes 4, 5 and 12. Furthermore, we identified a set of markers suitable to differentiate S. lycopersicum var. Moneymaker from all other wild relatives we evaluated. The SNPs can be used for genotyping, identification of varieties, comparison of genetic and physical maps and to confirm (dis)similarities (Chapter 2).
A part of tomato research aims at determining how to expand genetic diversity in the existing tomato crop; this can be done by incorporating useful traits found in wild germplasm. In this thesis we focused on exploring the variation between accessions of wild relatives of the species in the subsection Lycopersicon. Especially, on the species Solanum pimpinellifolium since it represents a good source to explore variation for quality traits that can be incorporated into cultivated tomatoes. A recombinant inbred line (RIL) population was developed from a cross between an accession of this wild relative and S. lycopersicum var. Moneymaker. All the lines were genotyped with our SNP array and 1974 SNPs made it possible to construct a linkage map based on 715 genetic loci. In this way we could compare genetic linkage and physical positions. Additionally, a subset of the lines was genotyped by sequencing (GBS). We identified two QTLs for resistance to Tomato Yellow Mosaic Virus (TYCV) and the sequence information was used to saturate the Quantitative Trait Loci with more markers. We found that the resistance to TYLCV was associated to a region on chromosome 11 close to the region of qTy-p11 (~51.3 Mb) and to another region on chromosome 3 near qTy-p3 (~46.5 Mb) (Chapter 3). We also used this genotyping approach to target mQTL hotspots for fruit related metabolites.
Three different metabolomics platforms were used to phenotype the metabolome of the RIL population. Liquid chromatography coupled with mass spectrometry (LS) was used to detect semi-polar compounds such as flavonoids, alkaloids, phenylpropanoids, saponins, phenolic acids, polyamines and products thereof. Gas chromatography (GC) coupled with electron impact time of flight (TOF) was used for detection of primary metabolites and solid phase microextraction (SPME)-GC for the analysis of volatiles. We performed QTL analysis on leaf and fruit samples of the RIL population. The TYLCV resistance mechanism is likely associated with sucrose and flavonoid glycosides related regions on chromosomes 11 and 3, respectively. With the combination of different ~omics platforms we provided a valuable insight into the genetics behind S. pimpinellifolium-derived TYLCV resistance.
For fruits, we found clear metabolite QTL-hotspots on chromosomes 1 and 10. Our results show that to increase the antioxidant properties of tomato, the region between 71-87 Mb on chromosome 1 has to originate from Moneymaker while other regions on chromosome 6 (35-44 Mb), chromosome 10 (~44.3 Mb) and chromosome 12 (~48 Mb) have to be of S. pimpinellifolium origin. The above-mentioned region on chromosome 6 also affects the concentration of malic acid in the fruits. Sugars can be increased by combining the wild alleles on chromosome 2 (~41.7 Mb) for sucrose and chromosome 10 (1.7 Mb) for fructose with the Moneymaker alleles on the hotspot region of chromosome 1 and chromosome 4 (~55 Mb) for fructose and glucose respectively. Off flavour regions that should be avoided in crosses with S. pimpinellifolium are the ones at the top of chromosome 1 and on chromosome 9 around 65 Mb where we found loci associated with the concentration of the compounds putrescine and dimethyl disulfide. An aromatic boost to the fruits can be given by the introgression of parts of the wild chromosome 8; this will increase the concentration of phenolic VOCs (Chapter 4).
S. pimpinellifolium certainly harbours characteristics that could be (re-) introduced in tomato. Therefore, we explored the metabolome of several accessions of S. pimpinellifolium during ripening. Clear metabolic profile differences were identified between species, especially related to the phenylpropanoid pathway. S. pimpinellifolium is a potential source to improve the flavonoid content of tomatoes and several other fruit aromas. Certain accessions looked even more promising than the RIL parent as a source for quality traits. All this helped us to better understand particular differences with wild relatives or even between genotypes (Chapter 5).
In general, our results give an insight in the physical positions of metabolite related QTLs that could be used by breeders that would like to exploit S. pimpinellifolium to improve tomato quality.