each cross that plant breeders are interested in evaluating. Often, hybrids are produced before the parental lines have been thoroughly evaluated. In that way, the best hybrids are quickly available when the data from field and greenhouse trials have identified those that should be advanced to the next stage of breeding. In the same way, lines are used to start new crosses before they have been thoroughly evaluated. Seedling evaluations using molecular markers to select individuals with traits of interest are also decreasing generation times.
Numerous systems to detect variability in plant DNA have been developed and used to facilitate genetic mapping. Simple sequence repeats (SSRs) and single nucleotide polymorphisms (SNPs) are two of the most common molecular marker systems used for generating genetic maps currently. One use of genetic mapping is for the identification of molecular markers associated with traits of interest (high yield, fruit quality, disease resistance). Linkage between a molecular marker and a trait of interest allows one to identify the presence of the trait by checking the genotype at the marker locus. This testing can occur as early as the seed stage, allowing for rapid selection of plants that have the trait of interest. In cucumber, watermelon, squash and melon, markers are available to select for improved fruit quality and disease resistance.
In cases where simply inherited sources of disease resistance have not been found (for example, resistance to CMV in melon and squash), genetic engineering and transfer of the coat protein or replicase genes of the virus to cucurbits has given these cucurbits disease resistance. For example, Chee and Slightom (1991) developed the coat protein-mediated mechanism for CMV protection in cucumber. They used a disarmed strain of Agrobacterium tumefaciens to transfer the engineered viral coat protein to cucumber. Small pieces of young, growing cotyledon tissue were soaked in a solution containing the recombinant DNA-carrying bacterial plasmids. The transformed embryogenic calli were regenerated and the resulting plants were inoculated and shown to have resistance to CMV. Gonsalves et al. (1992) went on to prove that the transgenic plants had a high level of CMV resistance under natural field conditions, where viral exposure is usually via insect vectors.
Similar field performance experiments have been conducted for transgenic squash lines (C. pepo). The virus-susceptible hybrid ‘Pavo’ was compared with two transgenic selections from this cultivar that possessed CMV and WMV resistance (Arce-Ochoa et al., 1995). Under field conditions in Texas, the two transgenic selections had only 3% and 14% symptomatic plants, compared with 53% for ‘Pavo’.
In the USA, the first commercially available transgenic cucurbit was ‘Freedom II’ (previously called ‘ZW-20’), a yellow summer squash of the crookneck type (C. pepo). This cultivar was engineered with resistance to WMV and ZYMV. Again, A. tumefaciens was used, this time to transfer the coat protein genes of WMV and ZYMV to crookneck squash. Since the introduction of ‘Freedom II’ additional cultivars have been released, including ‘Liberator III’ and ‘Independence II’.
More recently, gene editing techniques have gained popularity. The clustered regularly interspersed short palindromic repeats (CRISPR)-associated protein system is universally recognized as a powerful genome editing tool. In brief, CRISPR is an RNA-guided editing tool that uses a short (21–72 nucleotides) seed sequence flanked by palindromic repeats to recognize and cleave a specific sequence of DNA. Unlike genetic engineering, where genes are inserted at random, genome editing allowed for specific changes with no to few off-target effects or foreign genes remaining in the new cultivar. CRISPR represents a highly efficient, low-cost genome editing tool and is now being developed for cucurbits. While the CAS9 system is the most common, multiple CRISPR systems are being developed to fully utilize this technology. In cucumber, CRISPR has been used to incorporate resistance to zucchini yellow mosaic virus, papaya ringspot mosaic virus-W and cucumber vein yellowing virus into a non-transgenic plant (Chandrasekaran et al., 2016). In watermelon, a CRISPR system has been tested using phytoene desaturase, but new quality traits have not yet been successfully incorporated (Tian et al., 2017).
CUCUMBER, MELON AND THEIR RELATIVES
Introduction
A biosystematic monograph of the genus Cucumis (Kirkbride, 1993) recognized 32 species, including two major crops, cucumber (C. sativus) and melon (C. melo), and two minor crops, West Indian gherkin (C. anguria) and African horned melon (C. metuliferus). Other species are sometimes cultivated (e.g. C. dipsaceus) or collected wild and used for food, water, or medicine (e.g. C. africanus L.). Molecular phylogenetic studies have placed additional genera into Cucumis based on chloroplast and nuclear DNA sequence similarities. A recent analysis of 100 Cucumis species from Africa to Australasia (Sebastian et al., 2010) provided a better picture of species diversity and relatedness and added 18 close relatives of cucumber and melon not previously included by Kirkbride. Of the 66 Cucumis species, 25 are Asian and Australian and 41 are African.
Cucumis is divided into subgenus Cucumis, composed of C. sativus and C. hystrix Chakravarty, and subgenus Melo (Mill.) C. Jeffrey, containing the remaining species. Phylogenetic investigations of Cucumis species have been based on crossing relationships, karyotypes, flavonoids, isozymes, chloroplast DNA and molecular markers. All of these studies agree that the subgenera are widely separated, to the point that it has been proposed they be in distinct genera. Puchalski and Robinson (1990) proposed seven groups of Cucumis species according to isozyme patterns. The groupings were generally similar to those above, except that the anguria group was separated into three isozyme groups and C. sagittatus was placed in a different isozyme group than C. melo.
Interspecific crossing relationships (see Chapter 3, Fig. 3.2) (Deakin et al., 1971) and molecular phylogenies (Garcia-Mas et al., 2004) suggest four groups of Cucumis species. Crossing relationship groups are: (i) the anguria group, composed of C. anguria and seven other intercompatible species (marked ‘2’ on Fig. 3.2) with softly spiny fruit; (ii) C. metuliferus; (iii) C. sativus, including the fully compatible C. sativus var. hardwickii (Royle) Gabaev; and (iv) the melo group, which includes C. melo, C. humifructus and C. sagittatus Peyritsch in Wawra & Peyritsch. Species of the melo group have hairy immature fruit, but, unlike other Cucumis species, their mature fruit lack spines. Interspecific hybrids have not been obtained in this group, although fruit with partly developed seeds were obtained in the cross C. melo × C. sagittatus.
Botany
Cucumber, Cucumis sativus, is the fourth most important vegetable crop in the world and the most important cucurbit. Cucumbers, like most cucurbits, are indeterminate, vining (1–3 m), frost-sensitive annual plants that produce cylindrical
