Keywords

Introduction

Tomato has been a model species for genetic analyses for years. The diversity of its fruit colour, shape and size has interested geneticists since the early work of genetic mapping. Butler (1952) proposed one of the first genetic maps including more than 50 loci corresponding to phenotypic mutations. Nevertheless, until the discovery of molecular markers in the late 1980s, the location of mutations on genetic maps was not really precise as it was impossible to simultaneously map many loci. Molecular markers have enabled biologists to construct saturated linkage maps of the genome and to systematically localize mutations of interest on these maps. Over years, more and more markers were discovered and the genotyping cost decreased. Following isozymes, the first DNA markers, based on the detection of Restriction Fragment Length Polymorphisms (RFLP), allowed the construction of a reference map of the tomato genome (Tanksley et al. 1992). With more than 1000 loci, spread on the 12 chromosomes, this map allowed the precise localization of several mutations and of a few genes of interest. New mutations or genes of interest were subsequently mapped using either F2 populations or pairs of near isogenic lines differing only in the region of the interesting gene (Laterrot 1996). Bulks of individuals were later used (following the Bulk Segregant Analysis method), together with markers based on PCR amplification of the DNA (RAPD or AFLP markers). Following the identification of PCR markers linked to the gene of interest, specific PCR markers were set up, simplifying the genotyping step for breeders. Nevertheless, PCR markers such as RAPD or AFLP are dominant and map for the most part close to the centromeres, reducing their potential efficiency for gene mapping in tomato (Grandillo and Tanksley 1996; Haanstra et al. 1999; Saliba-Colombani et al. 2000). Markers based on the variation in the number of small sequence repeats (microsatellites or SSR) were then discovered and mapped on the reference map or used for the construction of new maps (He et al. 2002; Liu et al. 2005). To increase the number of markers available and to use the microsynteny observed with the Arabidopsis thaliana genome, Fulton et al. (2002) proposed the use of Conserved Ortholog Sequences (COS) as markers.

The polymorphism revealed by RFLP markers among cultivated accessions was very low and only a few markers were polymorphic and thus useful for mapping genes in such genetic background (Saliba-Colombani et al. 2000). Interspecific progenies were much more polymorphic and maps based on progenies derived from crosses with every wild species related to tomato were constructed (Labate et al. 2007). A population of introgression lines derived from a cross with a Solanum pennellii accession (Eshed and Zamir 1995) was particularly useful to discover new genes and quantitative trait loci (QTL) involved in fruit colour, size and plant traits (Zamir 2001).

More recently, several tomato accessions were used to sequence fragments of expressed sequences and identify Expressed Sequence Tags (ESTs), allowing the first Single Nucleotide Polymorphism (SNP) markers to be discovered and mapped (Labate and Baldo 2005; Sim et al. 2009). With the access to the tomato genome sequence (Tomato Genome Consortium 2012), the increased throughput of sequencing and the advances in Next Generation Sequencing technologies, it has been possible to discover thousands of SNPs through RNA sequencing (RNAseq). The SolCAP consortium developed a SNP array carrying more than 8000 SNPs chosen to reveal polymorphisms among cultivated accessions (Sim et al. 2012). Another SNP array was developed by Víquez-Zamora et al. (2013). Today, thanks to the tomato genome sequence availability, several projects of resequencing whole genomes of tomato accessions allowed the discovery of several millions of SNP (Causse et al. 2013; Aflitos et al. 2014; Lin et al. 2014) and the construction of genetic maps at the intraspecific level is now possible (Shirasawa et al. 2010). Large SNP arrays permit the rapid mapping of new loci of interest (Viquez-Zamora et al. 2014).

Genes and Loci Involved in Morphological and Fruit Characteristics

Among the major mutations used in tomato, the self-pruning (sp) mutation was discovered about 100 years ago and confers the determinate growth behaviour. It was largely used in processing tomato for field grown production. The tomato SELF-PRUNING (SP) gene is the homolog of the Antirrhinum majus CENTRORADIALIS (CEN) and Arabidopsis thaliana TERMINAL FLOWER1 (TFL1) genes (Pnueli et al. 1998).

Many mutations in genes related to the carotenoid pathway were identified and correspond to specific fruit colours (Hirschberg 2001). Among them the B/ogc locus has been shown to correspond to two mutations in the same gene responsible for either yellow or dark red colour of the fruit (Ronen et al. 2000). Recently the gene conferring the uniform ripening (u) phenotype was cloned and shown to correspond to a Golden 2-like (GLK) transcription factor, which determines the chlorophyll accumulation and distribution in developing fruit (Powell et al. 2012). The y locus, responsible for the pink fruit colour (due to a colourless peel which lacks the yellow flavonoid pigment naringenin chalcone), was also cloned. It corresponds to a MYB transcription factor (Adato et al. 2009; Ballester et al. 2010). Several alleles and their polymorphisms were identified at the y locus, thanks to the recent resequencing of more than 300 tomato accessions (Lin et al. 2014). Several mutations confer a long shelf life to the fruit. The most widely used, rin (for ripening inhibitor) corresponds to a deletion in a MADS BOX transcription factor (Vrebalov et al. 2002). Another important discovery was the mutation at the Cnr locus (Colourless non-ripening), which was one of the first epiallele discovered in tomato (Manning et al. 2006). Table 3.1 lists the genes involved in morphological and fruit mutations.

Table 3.1 Cloned genes with a phenotyped mutant mapped on the tomato genome assembly
Full size table

Disease Resistance Genes

Tomato is susceptible to many pathogens and all the resistance genes (R) were discovered in wild relatives. Many tomato disease resistance genes were mapped and characterized (Table 3.2). Since the first positionally cloned R gene (Pto, by Martin et al. 1993), more than 20 genes were cloned and characterized. Their structure and evolution was analyzed and the great conservation among genes conferring resistance to different types of pathogens revealed. The majority of R genes cloned so far encode proteins with a nucleotide-binding site (NBS) and a leucine-rich repeat (LRR) region (Ellis et al. 2000).

Table 3.2 Disease resistance genes cloned
Full size table

Mutant Collections

Many natural mutations were discovered in tomato. The Tomato Genetic Resources Center (TGRC, Davis, California, USA) collection encompasses more than 1000 monogenic mutants at over 600 loci, including spontaneous and induced mutations affecting many aspects of plant development and morphology, disease resistance genes, protein marker stocks, and other traits of economic importance (Chetelat 2005). Genetic data on individual stocks, including phenotypes, images, chromosome locations, etc. are available at the TGRC website (http://tgrc.ucdavis.edu/).

An additional series of provisional (i.e. less well-characterized) mutants is also available. The Hebrew University of Jerusalem developed an isogenic mutant library in the genetic background of cv. M82 (http://zamir.sgn.cornell.edu/mutants/index.html). A total of 13,000 M2 families, generated by ethylmethane sulfonate (EMS) and fast-neutron mutagenesis, were phenotypically analyzed and catalogued into at least 3417 mutations (Menda et al. 2004). This series of mutations includes many previously described mutant phenotypes as well as many novel mutants, and multiple alleles per locus. Screening this collection allowed the discovery of interesting alleles which interact with the SP gene and whose mutation modify its expression and may allow optimization of crop productivity (Park et al. 2014). Other collections of mutants are available (Okabe et al. 2011). Together these mutant collections provide important tools for analyses of gene function either through forward or reverse genetic approaches (Chap. 5).

Recombination Heterogeneity

Many genes/mutations were mapped on a genetic map but not yet cloned (Table 3.3). The recent availability of the tomato genome sequence confirmed earlier observations that recombination is unevenly distributed along chromosomes and that large pieces of the chromosomes around the centromeres do not recombine at all (Sim et al. 2012; Tomato Genome Consortium 2012). If the recombination frequencies may vary from one progeny to the other (Fig. 3.1), these regions do not recombine more in any. The ratio of kb per cM thus greatly varies hampering the characterization of some mutations due to the lack of recombination. Hopefully, these regions of low recombination also correspond to regions with lower gene density.

Table 3.3 Genes mapped on a genetic map with a phenotype but not yet cloned
Full size table
Fig. 3.1
figure 1

Relationships between physical and genetic distances: example of chromosome 3

Full size image

Many genes involved in morphological traits or disease resistances remain to be characterized. The high-quality genome sequence and millions of SNPs available today constitute unique resources to rapidly identify new genes of interest. High throughput genotyping technologies combined to the information on gene annotation and expression in various tissues should make the task much easier.