PLS221                                        Instructor: Carlos F. Quiros

Cucurbitacea: Cucumber and muskmelonCucumis spp.

List of references (includes reading assignments) 


Genomic numbers in Cucumis

Interspecies relationships and evolution.

Cucumber: Origin and distribution, germplasm.

Interspecific hybridization

Genetic determination of most important traits

Sex expression


Mapping and markers

Disease resistance

The most important cultivated species within the family cucurbitacea are:

Cucumis sativus 2n=2x=14

Cucumis melo 2n=2x=24

Citrullus lanatus 2n=2x=22

Cucurbita spp 2n=2x= 40

The genus Cucumis can be separated in 2 subgenera on the basis of chromosome numbers and crossing ability: Cucumis x=7, (cucumbers), Asian species and Melo (muskmelons), African species, x=12.

In spite of sharing the same genus, these species are taxonomically quite distant and should be reclassified in different genera. This is supported by phylogenetic studies based on  isozymes and DNA markers.

Further, hybridization attempts between species of the two sub-genera involving the x=7 and x=12 species have always failed, even at different ploidy levels. The nuclear DNA content is  1.78 pg/1C and  2.48pg/1C, respectively.


Place of origin: Landraces are found in subtropical valleys bordering the Himalayas,India, Southern Asia, where the wild form, C. sativus ssp hardwickii, and possible progenitor of cultivated cucumber is also found. It has the same chromosome number and crosses readily producing F1 and F2 fully fertile. It has small fruits which are extremely bitter.

The wild species, C. callosus, with the same chromosome number but different karyotype, might be the ancestral to ssp. hardwickii. C. hystrix (2n=24) crosses with sativus with the aid of chromosome number manipulation (Chen et al 2003).

Cucumbers have been under cultivation in India for at least 3000 years. It was introduced very early into Southern Europe and China.

Many of the changes taking place during domestication relate to fruit morphology.

ssp. hardwickii have been exploited extensively for breeding. It is characterized by:

smaller seeds,

short day adaptation,

bitter fruits,

various disease resistances,

many fruits/plant.

In the cultivated types new fruit formation is inhibited by presence of mature fruits in the plant. This is an important limiting factor in the common cultivars which seldom they are able to produce only 2 fruits/plant.

Cucumber has certain advantages for genetical studies:

1) Low chromosome number,  facilitating linkage studies.

2) Short life cycle, being possible to grow several generations a year.

3) Easy to grow and propagated by sexual and asexual means.

4) Plants flower over an extended period of time, permitting crosses between plants of different maturities. Flowers are easily pollinated, emasculation is not needed since this is a monoeciuos species (unisexual flowers).

5) No problems with inbreeding depression or SI.

6) Many seeds/fruit. Seeds have long shelf life.

At least 70 genes have been reported for cucumber, many seedlings markers, mostly
chlorophyll mutants, leaf shape, morphology, hairs etc.

For example:

Determined habit, gene de,
dwarf plant , gene dw,
tendril inhibition, gene td.

Sex expression:

The wild type sex of cucumber is monoecious,  which displays 3 phases of growth:

1st phase: only staminated (male) flowers.

2nd phase: alternating flowers, male and female (pistilated flowers)

3rd phase: only female flowers

Several genes have been found to modify sex expression:

F: female tendency or gynoecy (also named  Acr), it probably corresponds to gene CsACS1G, encoding a key enzyme involved in  ethylene synthesis (Huang et al 2009)

a: male tendency in the presence of f

m: perfect flowers in presence of  f (hermaphroditic tendency). M recently cloned by Huang et al 2009, also involved in the ethylene pathway. It codes for  the 1 aminocyclopropane-1-carboxylic acid synthase (ACC gene CsACS2).

Additional loci might be involved as modifiers, and environmental factors such as temperature, light intensity and daylight might affect sex expression.



Sex type




Gynoecious (only female flowers)








Androecious (only 
male flowers)








Other sex variants includes:

Trimonoecious, male, female and perfect flowers, seems to require a fourth locus, Tr

Gynomonoecious,  female and perfect flowers

Andromonoecious: male and perfect flowers

Manipulation of sex expression:

Sex expression can be manipulated by growth regulators.

1) GA3, silver nitrate and silver thiosulfate promote development of male flowers in monoecious and gynoecious strains. These compounds interfere with endogenous ethylene production.

2) Ethylene (ethephon) induce production of female flowers. It induces transcription of MAD-box gene ERAF17 (Ando et al. 2001).

This ability is of great significance in breeding, since gynoecious lines can be induced to form male flowers for selfings and crossings and to produce gynoecious inbred lines.

Male sterility:

A few nuclear male sterile genes are available: ms-1, ms-2, cl and one for
female sterility, co.


This area of research is not well developed for this crop. Traditional techniques have been used for mitotic chromosome identification based on Feulgen banding and Giemsa C-banding. Recently Koo et al (2005) demonstrates that FISH is a powerful tool for molecular cytogenetics of cucumber.

Alien addition lines developed by Chen et al (2003) crossing C.  hystrix (2n=24) x C. sativus to transfer disease and abiotic resistances.  A sesquidiploid HSS was produced by crossing tetraploid sativus to the wild species.  Two monosomic addition lines were found after treating with colchinie shoots of the HSS hybrid, apparently by H chromosome elimination induced by the colchicine treatment.


Monoploids: This ploidy variant was obtained by Aalders (1958) with the original technique of suspending seeds in water. The embryos from floating seed, which were abnormally light,  yielded 5% monoploids after culture in vitro.

Doubled haploids have been successfully obtained by culturing unfertilized ovules and then doubling chromosome number of the resulting plantlets with colchicine. (Balkema-Boostra et al. 2003).

Tetraploids: These have been found spontaneously, 1 in 10,000. Also tetraploids have been obtained by colchicine treatment. Although they have low fertility, triploids have been obtained by crossing to diploids.

3n x 2n crosses after embryo rescue might produce trisomics. However, these stocks are not available in cucumber.

Breeding objectives:

F1 hybrids are becoming very popular.

There are two main types of cucumbers: Pickling and slicing cucumbers.
Practically all varieties of both types have now white spines in the skin.

1) Pickling types: have a smaller length/diameter ratio (2.8-3.2"), lighter skin, more pronounced warts when immature.

Before these types were black spined since this trait was associated to better processing qualities. These good qualities have been bred now into white spined cucumbers.

2) Slicing types: white spined and dark skin, 6-10" length, thicker skins, harvested later than processing varieties.

Lateral branching. Opens the possibility to increase yield and concentrate fruit production for mechanical harvest in combination with determinate types.

Inheritance of fruit traits:

Parthenocarpy: The incompletely dominant gene Pc is responsible for this trait which used extensively in Europe. This is a desirable trait for cucumber greenhouse production, since the fruits will develop without the need for pollinators. Pathenocarpic cucumbers do not have seeds and are sold as slicing cucumbers at prices often 10 times higher than the normal varieties.

Skin spines: white vs black.

Gene B, black is dominant over white.

B is linked to R, determining mature red fruit color, and to H, determining fruit netting,
no crossovers have been found.

Fruit warts: Most European cvs have gene Tu, which results in fruit wart development in the surface when immature. It is also a desirable treat preferred by some consumers.

Bitter fruits: Bt gene determining presence of C cucurbitacin, a terpenoid compound that confers resistance to mites, but serve as an attractant for cucumber beetles.

Yellow/orange flesh hi carotene cucumbers developed by Simo and Navazio (1997).

Disease resistance: Most cultivars today resistant to 8 diseases:







Bact Wilt






Cucumb. mosaic



Melon/Watermelon mosaic



Downy mildew



Powdery mildew


pm-1, pm-2, pm-3

Screening for disease is done at seedling stage. Some of the inoculations can be done simultaneously in a single cotyledon because of the large size of this organ.

RAPD markers flanking dm resistant gene are available, but at distances from 10 to 16 cM. Possible use in marker assisted selection schemes.

61 NBS-LRR genes identified by Huang et al (2009) after analyzing the cucumber genome sequence. This number is v small compared with other genomes (up to 600 of these genes in rice). 20 of the genes are on chromosome 2 , ¾ of the genes are arranged in 11 clusters.  Lypogenase (LOX) gene family involved in plant defense expanded to 23 genes in cucumber, compared to 6 in Arabidopsis.

Cucumber Genome

It has been sequenced by Huang et al (2009) using inbred line  ‘Chinese Long” . It has 243.5 Mbp and ~27,000 predicted genes of average size ~1kb and 4 exons/gene. 82% of the genes had known homologs or known function. No evidence for recent whole genome duplication found.

73% of the sequences anchored the seven chromosomes of a genetic map of C. sativus x C hardikii containing 1885 markers including 995 SSR.

54.4 Mb (`24% oft the genome) is repeated. LTR retransposons copia and gypsie made the majority of the TE elements, occupying 10.4% of the genome

66.7% of melon markers and 58.6% of watermelon aligned to the cucumber chromosomes.

Mapping and markers:

A number of cucumber linkage map has been produced in the past. The first one consisting of six groups and based on morphological traits, disease resistant genes  and other genes of interest (Pierce and Wehner, 1990).

A linkage map has also been constructed for isozyme loci consisting of 4 groups (Knerr and Staub 1992). Later  DNA-based markers have been added to this map and has also aligned in part with the morphological map using a narrow and a wide cross (Kennard et al. 1994).

More recently maps were developed based on RAPD, SSR, AFLP, SCAR, SRAP and ISSR markers. Yeboah et al (2007) reported construction of SRAP and ISSR map. The latest map is the one based on the cucumis genome sequence with close to 2000 markers mentioned above (Huang et al 2009).

 RFLP and isozymes have been used to infer phylogenetic relationships between cultivated types of cucumber and ssp hardwickii. Additionally RAPD markers allow the separation of cucumber cultivars and landraces by geographical origin (Horejsi and Staub, 1999).

QTL Also RFLPs have been applied to detect QTL associated to quantitative fruit traits, such as fruit diameter, length and weight (Kennard and Havey 1995) and other genes (Serquen et al. 2000). For example, the F gene (sex determination) was mapped. Three other regions in the genome were found to be involved in sex determination.  QTL analysis for lateral branching detected as well as fruit traits such as weight , length, diameter, flesh thickness,  seed cavity etc. (Yuan et al 2008).

BAC libraries are also available.


Cucumbers are amenable to organogenesis and regeneration from protoplasts, but this technique has not had any major impact in variety development so far.

Agrobacterium mediated transformation with coat protein of cucumber mosaic virus has resulted in acceptable levels of resistance to this virus, comparable to the resistant standard (Gonsalves et al. 1992). Also transformation with rice chitinase has resulted in enhanced resistance to gray mold (Botrytis cinerea). 

Cucumber has been used as a bioreactor for increased activity (3X) of superoxide dismutase (SOD) from cassava to cucumber as for higher antioxidant activity. It is proposed to use it as a functional cosmetic to combat ageing by placing cucumber slices over the skin, which is already done in many countries with conventional cucumber.  Lee et al. (2003). Mol Breed. 11:213-220.



Muskmelons and cantaloupes:

Origin and distribution

Related species and interspecific hybridization

Genetic determination of most important traits

Male sterility

Sex determination and schemes for F1 hybrid production

Linkage maps and markers


Biotechnology and breeding

Cucumis melo: x=12

This species probably originates in Africa. The primary center of diversity is southwest and central Asia. Secondary centers are found in Spain, Portugal and China. A number of wild relatives are found in East and West Africa. The cultivated species widespread in arid and semi arid regions of Africa and Asia. Relatively small genome, 4.5 x 108 , roughly 3x the A. thaliana genome. Approximately 30% of the genome is duplicated.

These Cucumis species have been grouped on the basis of crossing ability and chromosome pairing by Singh and Yadava (1984). Among them, annuals and perennials, monoecious and dioecious, diploid and polyploids are found.

Crossing groups:

1) Spiny fruit species: All these are monoecious.

C. ficifolius, perennial, diploid and tetraploid forms

C. africanus,  annual, only diploid

C. myriocarpus ssp. leptodermis, annual, diploid

C. anguria, annual, diploid

In addition there are 5 other species, including diploid, tetraploid and hexaploid forms.

Genomic differentiation within this group is due to chromosomal translocations and inversions.

These species cross to each other, producing hybrids with different levels of fertility.

There are several reports on isozymes (29 loci) and cpDNA in these species, agreeing with their classification based on cytogenetic and hybridization studies.

2) Non-spiny fruit species:

a) C. sagittatus

b) C. metuliferus and C. melo

Species of groups 2a and 2b are partially cross-compatible, C. melo has been crossed to C. sagittatus.

Species of group 1 and 2 are cross-incompatible.

Although C. melo and C. metuliferus are in the same sub-group, there is difficulty hybridizing them. Furthermore, isozyme phylogeny separates these two species even further than the spiny-fruit species are from C. melo.

Wild varieties of melo classified as C. melo var. agrestis

Cultivars of C. melo have been divided into several botanical varieties based on various fruit traits, and more recently by isozymes and DNA markers. The main varieties are:

1) cantalupensis: netted muskmelon, aromatic melons with orange or green flesh, typical American cantaloup.

2) inodorus: fruits lack aroma,  smooth rind. Includes Honeydews, casabas and others. White or green flesh.

3) chito: Mango melon, used for pickles or cooked.

4) conomom: used for pickling in Asian countries

5) dudaim: Pomegranate melon. Popular in Asia and Africa, aromatic fruits, used sometimes as an ornamental.

6) momordica: Snap melon, popular in India, cracked fruit surface, light orange or white flesh

7) flexuosus: Snake melon,  consumed immature, resembles cucumber in fruit shape.

Other classifications: Two subspecies of C. melo are sometimes recognized. Subsp melo, including cantelupensis and inodorus, and subsp agrestis, including conomon and flexuosus.

Important fruit traits:

Flesh color: Determined by two loci, gf for salmon or green color, and wf, color inhibitor in presence of gf allele.

Genotypes: green (gfgfWf_) green is recessive to salmon

salmon (Gf_Wf)

white (gfgfwfwf) double recessive

Skin color:  Also determined by single genes.

Suture: Determined by single recessive s

Netting: At least 5 genes involved.

Fruit shape: O gene determines oval fruit

Fruit ridges: Determined by ri


Five nuclear male-sterile genes have been reported: ms-3 is the most promising, used already for some F1 hybrids.

Sex expression

Most American cultivars are monoecious and andromonoecious.

Two genes involved in sex expression:

Sex type















A gene seems to be homologous to cucumber M (Huang et al 2009)

Schemes for F1 hybrid seed production:

Similar to cucumber, sex expression in melons can be manipulated by use of chemicals for F1 seed production.

1) monoecious (+ethephon) x andromonoecious
    Seed parent (female)         pollen parent (male)

Problems: monoecious trait often linked to elongated fruit. Change to
to femaleness is not always complete, causing possible seed contamination.

2) gynoecious x andromonoeciuos

3) Gynoecious x gynoecious (+AgNO3), not used commercially yet.

Gynoecious lines can be maintained by spraying with them with AgNO3 to induce male flowers, relying in bees for pollination.

Plant bushy type, is determine by gene b and 2 modifiers. This trait has potential for developing melons for mechanical harvest.


Fusarium, at least two dominant gene for resistance, multiple races Fom-1, Fom-2. Co-dominant markers for Fom-1 reported in 2011.

Powdery mildew: single dominant gene for resistance, several races exist. Chinese melons are resistant.
This disease can be effectively controlled by sulfur dusting, however many cultivars are susceptible to to sulfur. Resistance to sulfur is dominant, 2 to 3 genes involved.

Verticillium, multiple genes for resistance.


This area of research has not been developed for melons, due to small chromosome size.

Haploids: Anther culture has been unsuccessful for this crop, however haploids have been obtained by gynogenesis.

Gynogenesis: Female parthenogenesis induced by pollen irradiation, yielding  1.5 to 3% haploids.

Pollen is irradiated between 30 to 100 Krads, then used for pollination. It results in seeds lacking endosperm, apparently empty. The embryos are then extracted and cultured in vitro. The amount of pollen used for pollination increases the yield of haploids. Haploid frequency varies for varieties used as female.

The effect of irradiation dose is also important, below 30Krads results in low haploid yield, since pollen will be able to fertilize the egg, resulting in aneuploids and other chromosome aberrant types.

The chromosomes of the resulting haploids are doubled in vitro by dipping stem pieces in 0.5% colchicine for 2 hrs. Autofecundation of these yield a high number of doubled haploids.

Doubled haploid lines widely developed for construction of mapping populations.

Linkage groups and markers:

A linkage map based on morphological markers and traits of interest has been constructed for melons consisting of 8 groups (Pitrat 1991). Later this map was expanded to 14 groups using isozyme, RFLP and RAPD markers (Baudracco-Arnas and Pitrat, 1996).  Katzir et al. (1997) reported microsatellite markers for this crop and Wang et al. (1997) constructed a linkage map based on AFLP markers consisting of 14 major and six minor groups. Danin-Poleg (2002) developed a partial linkage group of molecular markers were various genes of importance where located, such as andromonecious (a), striped epicarp (st), fruit flesh (pH), resistance to zucchine yellow mosaic virus ZYMV (zym-1). Yuste-Lisbona 2011 produced a linkage map and identified a QTL for resistance to powdery mildew.

Marker developed for Fusarium resistance gene Fom-2. Candidate NBS-LRR sequences identified (Wang et al. 2002, Genome 45:473). BAC clones available for this work.  Microlinarity with Arabidopsis for segments harboring disease resistance genes reported by van Leuween et al. (2002).

Markers used to study genetic variation and divergence of the various botanical groups. These include isozymes, RFLPs (using cucumber DNA probes), RAPDs, AFLPs and SSRs. Sweet groups (and flexuosus) separate from the rest of the groups (non-sweet) [(Garcia-Mas et al. 2000)]. Momordica types seems more distant than the rest, whereas inodorus and cantelupensis, are close to each other (Silberstein et al. 1999). Greek melon landraces found also to be fairly unique based on molecular marker surveys (Staub et al (2004).


(Tree adapted from Silberstein et al. 1999)

Good level of hybridization of cucumber DNA probes to melon DNA opens possibility for comparative mapping.

Gonzalo et al developed and mapped 118 SSR markers merging with the existing RFLP maps. QTL for fruit quality traits have been studied. Doubled haploid and NILs available for these studies by crossing Spanish with Asian melon cultivars as a source of important traits, including disease resistance (Eduardo et al 2005). Zapala et al (2007) detected QRT for yield related traits using RI lines. Traits included branch number, fruit number per plant, fruit weight, and % mature fruit per plot.

Hi microsynteny conservation found between cucumber and melon. Cucumber and melondivergence estimated between 4-7 mya. (Huang et al 2009).

Biotechnology and Breeding:

Micropropagation: Protocols for regeneration by organogenesis in cotyledon and leaf explants have been reported.  Success on somatic embryogenesis has been limited.

Transformation: Cotyledon explants were used for transformation with Agrobacterium tumefaciens and microprojectiles. Agrobacterium transformation often results in a high proportion of tetraploid plants. Leaf explant transformation method of high efficiency and high diploid yield has been reported by Guis et al. (2000).

Transformation has also been accomplished by biolistic techniques in this crop.

Gonsalves et al report attempts to obtain melons resistant to cucumber mosaic virus (CMV) by introduction of coat protein gene. This approach is called CP-mediated protection. Also transformation with antisense ACC oxigenease gene for longer shelf life has been accomplished.

Cucumber and melon evolution

Acording to Huang et al (2009), x=12 is the ancestral chromosome number, explain the Cucumber genome as result of fusion of melon ancestral chromosomes.

 Move to watermelon

 Back to Syllabus

 Last modified, May11, 2011

© Carlos F Quiros, 1998