Histological study of organogenesis in Cucumis melo L. after genetic transformation: why is it difficult to obtain transgenic plants?

Melon (Cucumis melo L.) is widely considered as a recalcitrant species for genetic transformation. In this study, we developed different regeneration and transformation protocols and we examined the regeneration process at different steps by histological studies. The highest regeneration rate (1.13 ± 0.02 plants per explant) was obtained using cotyledon explants of the 'Védrantais' genotype on Murashige and Skoog (MS) medium supplemented with 0.2 mg/l 6-benzylaminopurine (BAP) and 0.2 mg/l dimethylallylaminopurine (2-iP). Agrobacterium tumefaciens-mediated transformations with the uidA reporter gene were realized on cotyledon explants cultivated in these conditions: 70-90% of explants expressed a transient GUS activity during the early stages of regeneration, however, only few transgenic plants were obtained (1.8-4.5% of stable transformation with the GV2260pBI101 strain). These results revealed a low capacity of melon GUS-positive cells to regenerate transgenic plants. To evaluate the influence of the Agrobacterium infection on plant regeneration, histological analyses were conducted on explants 2, 7, 15, and 28 days after co-culture with the GV2260pBI101 strain. Genetic transformation occurred in epidermal and sub-epidermal cells and reached the meristematic structures expressing a high level of GUS activity during 14 days of culture; but after this period, most of the meristematic structures showed premature cell vacuolization and disorganization. This disruption of the GUS-positive meristematic areas could be responsible of the difficulties encountered to regenerate melon plants after genetic transformation.

Strain-specific and recessive QTLs involved in the control of partial resistance to Fusarium oxysporum f. sp. melonis race 1.2 in a recombinant inbred line population of melon.

Fusarium oxysporum f. sp. melonis (FOM) causes serious economic losses in melon (Cucumis melo L.). Two dominant resistance genes have been identified, Fom-1 and Fom-2, which provide resistance to races 0 and 2 and races 0 and 1, respectively, however FOM race 1.2 overcomes these resistance genes. A partial resistance to FOM race 1.2 that has been found in some Far East accessions is under polygenic control. A genetic map of melon was constructed to tag FOM race 1.2 resistance with DNA markers on a recombinant inbred line population derived from a cross between resistant (Isabelle) and susceptible (cv. Védrantais) lines. Artificial root inoculations on plantlets of this population using two strains, one that causes wilting (FOM 1.2w) and one that causes yellowing (FOM 1.2y), resulted in phenotypic and genotypic data that enabled the identification of nine quantitative trait loci (QTLs). These QTLs were detected on five linkage groups by composite interval mapping and explained between 41.9% and 66.4% of the total variation. Four digenic epistatic interactions involving seven loci were detected and increased the total phenotypic variation that was explained. Co-localizations between QTLs and resistance gene homologs or resistance genes, such as Fom-2 and Vat, were observed. A strain-specific QTL was detected, and some QTLs appeared to be recessive.

Relationship between Loci conferring downy mildew and powdery mildew resistance in melon assessed by quantitative trait Loci mapping.

ABSTRACT Partial resistance to downy mildew (Pseudoperonospora cubensis) and complete resistance to powdery mildew (Podosphaera xanthii races 1, 2, 3, and 5 and Golovinomyces cichoracearum race 1) were studied using a recombinant inbred line population between 'PI 124112' (resistant to both diseases) and 'Védrantais' (susceptible line). A genetic map of melon was constructed to tag these resistances with DNA markers. Natural and artificial inoculations of Pseudoperonospora cubensis were performed and replicated in several locations. One major quantitative trait loci (QTL), pcXII.1, was consistently detected among the locations and explained between 12 to 38% of the phenotypic variation for Pseudoperonospora cubensis resistance. Eight other Pseudoperonospora cubensis resistance QTL were identified. Artificial inoculations were performed with several strains of four races of Podosphaera xanthii and one race of G. cichoracearum. Two independent major genes, PmV.1 and PmXII.1, were identified and shown to be involved in the simple resistance to powdery mildew. Three digenic epistatic interactions involving four loci were detected for two races of Podosphaera xanthii and one race of G. cichoracearum. Co-localization between PmV.1, resistance genes, and resistance genes homologues was observed. Linkage between the major resistance QTL to Pseudoperonospora cubensis, pcXII.1, and one of the two resistance genes to powdery mildew, PmXII.1, was demonstrated.

Genetic control of fruit shape acts prior to anthesis in melon ( Cucumis melo L.).

Genetic control of fruit shape in Cucumis melo was studied using QTL analysis in two Recombinant Inbred (RI) populations consisting of 163 and 63 individuals, respectively, obtained by crossing the same round-fruited parent with two different elongated-fruit lines. Fruit shape is mainly explained by fruit length in these two populations. Most QTLs for fruit shape and ovary shape detected were found to co-segregate, thus demonstrating early control of fruit shape during ovary development. A high level of correlation between fruit shape and ovary shape was also found in 14 unrelated genetic lines, a finding which suggests that control of fruit shape by gene(s) active early in the ovary is a general feature in C. melo. Two major flower genes, a ( monoecious) and p ( pentamerous), were shown to have major effects on fruit shape. Major tightly linked QTLs for fruit and ovary shape were found close to the a and p genes, probably reflecting their pleiotropic effect on fruit shape. Moreover, one of the two QTLs detected in the Védrantais x PI 414723 population was also found in the Védrantais x PI 161375 population. Variation of fruit shape in melon could be due to variations having quantitative effects on a large set of genes that are probably involved in ovary development.

A reference map of Cucumis melo based on two recombinant inbred line populations.

A composite genetic melon map was generated based on two recombinant inbred line (RI) populations. By analyzing the segregation of 346 AFLPs, 113 IMAs and phenotypic characters on a RI population of 163 individuals derived from the cross Védrantais x PI 161375, a first map was constructed. About 20% of the molecular markers were skewed, and the residual heterozygosity was estimated at 4.43% which was not significantly different from the theoretical value of 4.2%. The genome distribution of molecular markers among the 12 linkage groups was not different from a random distribution with the exception of linkage group XII which was found significantly less populated. The genome distributions of IMAs and AFLPs were complementary. AFLPs were found mainly in the middle of each linkage group and sometimes clustered, whereas IMAs were found mainly at the end. A total of 318 molecular markers, mainly AFLP and IMA markers, were mapped on 63 RIs of the second population, Védrantais x PI 414723. Comparison of the maps enables one to conclude that AFLPs and IMAs of like molecular size, amplified with the same primer combination, correspond to the same genetic locus. Both maps were joined through 116 common markers comprising 106 comigrating AFLPs/IMAs, plus five SSRs and five phenotypic markers. The integrated melon map contained 668 loci issuing from the segregation of 1,093 molecular markers in the two RI populations. The composite map spanned 1,654 cM on 12 linkage groups which is the haploid number of chromosomes in melon. Thirty two known-function probes, i.e. known-function genes (9) and morphological traits (23), were included in this map. In addition, the composite map was anchored to previously published maps through SSRs, RFLPs and phenotypic characters.

Resistance gene homologues in melon are linked to genetic loci conferring disease and pest resistance.

Genomic and cDNA fragments with homology to known disease resistance genes (RGH fragments) were cloned from Cucumis melo using degenerate-primer PCR. Fifteen homologues of the NBS-LRR gene family have been isolated. The NBS-LRR homologues show high divergence and, based on the partial NBS-fragment sequences, appear to include members of the two major subfamilies that have been described in dicot plants, one that possesses a TIR-protein element and one that lacks such a domain. Genomic organization of these sequences was explored by DNA gel-blot analysis, and conservation among other Cucurbitaceae was assessed. Two mapping populations that segregate for several disease and pest resistance loci were used to map the RGH probes onto the melon genetic map. Several NBS-LRR related sequences mapped to the vicinity of genetic loci that control resistance to papaya ringspot virus, Fusarium oxysporum race 1, F. oxysporum race 2 and to the insect pest Aphis gossypii. The utility of such markers for breeding resistant melon cultivars and for cloning the respective R-genes is discussed.