Resolution by recombination: breaking up Solanum pennellii introgressions Saleh Alseekh 1, 2, 3 , Itai Ofner 2 , Tzili Pleban 2 , Pasquale Tripodi 4, 5 , Francesco Di Dato 4, 5 , Maria Cammareri 4 , Ayed Mohammad 3 , Silvana Grandillo 4 , Alisdair R. Fernie 1 , and Dani Zamir 2 1 Max Planck Institute of Molecular Plant Physiology, Am Mu ¨ hlenberg 1, 14476 Potsdam-Golm, Germany 2 The Institute of Plant Sciences and Genetics in Agriculture, Faculty of Agriculture, The Hebrew University of Jerusalem, P.O. Box 12, 76100 Rehovot, Israel 3 Hebron University, P.O. Box 40, Hebron, Palestine 4 Italian National Research Council – Institute of Plant Genetics (CNR – IGV), Research Division Portici, Via Universita ` 133, 80055 Portici, Italy 5 Consiglio per la Ricerca e la Sperimentazione in Agricoltura, Research Centre for Vegetable Crops (CRA-ORT), Pontecagnano (SA), Italy Quantitative trait locus (QTL) genetics retains an impor- tant role in the study of biological and agronomic process- es; however, its genetic resolution is often comparatively low. Community-based strategies are thus required to address this issue. Here we detail such a strategy wherein the widely used Solanum pennellii introgression lines (ILs) in the genetic background of the cultivated tomato (Sola- num lycopersicum) are broken up into molecular marker- defined sublines as a community resource for map-based cloning. Genome-wide association studies (GWAS) The emergence of GWAS and the adoption of genotyping by sequencing [1] has allowed the construction of ultrahigh- resolution haplotype maps of our major crops [2,3] and facilitated the identification of causative variation in im- portant agronomic characteristics to the resolution of a few genes [3]. Despite the resolving power of GWAS, pitfalls do exist [4] and these can be overcome by going back to the traditional experimental populations derived from bipa- rental crosses. A recurring theme in QTL analysis is the lack of sufficient mapping resolution. Although examples exist in which high-resolution mapping has been utilized to identify single-nucleotide changes [5] as causal for pheno- typic variation, most studies merely provide relatively coarse genome-wide surveys of the genetic architecture of the traits under study. One of the reasons behind the lack of resolution is the extreme time investment required to generate new subpopulations of specific regions to zoom in on the genetic factors underlying traits of interest. One solution to address this is to create genome-wide subpo- pulations harboring smaller, introgressed segments from the ‘donor parent’ (Box 1). Here we illustrate this approach by describing two publically available genetic resources derived from the extensively phenotyped S. pennellii (LA0716) ILs in the genetic background of S. lycopersicum cv M82: (i) a set of sublines with marker-defined introgres- sions smaller than the original set of ILs; and (ii) large quantities of F2 seeds derived from selfing of the hybrids of the ILs with the recurrent parent M82. S. pennellii ILs The S. pennellii ILs have been publically available since 1995 [6]. S. pennellii, a distant relative of the cultivated tomato S. lycopersicum, has evolved highly specific mor- phology, mating system, and phytochemical diversity and is particularly important given its desert habit responses to abiotic stress. Despite these specificities, it is sexually compatible with S. lycopersicum, producing fertile hybrids. For these reasons it was chosen as the founding donor parent of the first IL population used for interspecific QTL identification, cloning, and plant breeding [7]. The ILs represent whole-genome coverage of S. pennellii in over- lapping segments in the background of M82, presently comprising a core set of 76 genotypes (Figure S1 in the supplementary material online). A unique feature of the tomato IL phenotypes is that the raw data of the replicated measurements have been de- posited in the phenotype warehouse of Phenom Networks (http://phnserver.phenome-networks.com/). The data can be browsed and statistically analyzed online or down- loaded from the site for further analysis using other sta- tistical software. Phenom Networks harbors data from 45 IL experiments in which 355 traits were scored in replicat- ed measurements. These studies were conducted by mul- tiple laboratories and allowed the identification of diverse phenotypes underlying QTLs, such as whole-plant mor- phology and yield (including heterosis), metabolic compo- sition, fruit color, enzyme activities, leaf, fruit, and root morphology, cellular development, and biotic and abiotic stress tolerance (see, for example, [7–10]). Over 3069 QTLs have been identified in this population to date (Table S1 in the supplementary material online). Techniques & Applications Corresponding authors: Fernie, A.R. (fernie@mpimp-golm.mpg.de); Zamir, D. (dani.zamir@mail.huji.ac.il). Keywords: genome sequence; map-based cloning; interspecific cross; recombinant; sublines; tomato. 536