Characterization, Development And Application Of EST-SSR Markers In Cucumis Genus Based On Public Sequence Database

SSR markers are powerful tools for assaying genetic variation because of theirreproducibility, multiallelic nature, codominant inheritance, relative abundance and goodgenome coverage. SSR has become one of the most popular molecular markers, however,the development of SSR markers is low efficient with time and labor consuming whichlimits its application on more crops. With the development of functional genomics, ESTswere sequenced on large scale in many species for discovery of novel genes, analysis ofgene express regulation and application of substrate for gene microarray and deposited inpublic available sequence database, which offers an abundant resource for development ofSSR and other genetic markers. EST derived SSR markers have many intrinsicadvantages over genomic SSR markers for that they have a higher transferability amongrelated species and often have putative functions which can be deduced by homologysearches and can be developed easily at a lower cost. Because EST-SSRs derive fromtranscripts, they are more valuable for genetic diversity analysis, marker assisted selectionand comparative mapping.In Cucumis genus, there are two staple vegetable crops of cucumber (Cucumissativus .L) and melon (Cucumis melo .L), which world widely distribute and havesignificantly economic value. Development of SSR markers for cucumber and melon issignificant to the genetic and breeding study of the two crops. In this paper, employingthe data mining technique, we identified the presences of SSR on cucumber and melonESTs and analyzed their distribution characters. At the same time, we developed a total of42 polymorphic EST-SSR markers for cucumber and melon and evaluated theirapplication on the analysis of phylogeny for the two crops respectively. The main resultswe obtained are listed below.A total of 3627 Unigenes covering 2.2Mb was generated by assembling 5563 ESTand gene sequences from cucumber in GenBank. With the same way, 3189 Unigenes witha total length of 1.8Mb were obtained from 5748 melon EST and gene sequences. Thecriteria of minimum 5 repeat units for di-to hexa-nucleotide were adopted for screeningthe presence of SSR on cucumber and melon Unigenes respectively. A total of 479 SSRloci for cucumber with the frequency of 1/4.6kb and 385 SSR loci for melon with thefrequency of 1/4.7 kb were identified. The EST-SSR loci from cucumber and melo havemany similar characters. Dinucleotide repeat is the dominant motif and then istrinucleotide repeat among the EST-SSR loci. For dinucleotide repeat, AG is the mostabundant motif and then is AT. As for trinucleotide, AAG motif is dominant. Most of the SSR loci are located in CDS, and then are in 5'UTR and 3'UTR respectively. Test forgoodness-of-it with chi-square revealed that dinucleotide motif doesn't distributerandomly among CDS and UTR. The amount of SSR loci with dinucleotide motif in UTRis significant more than that in CDS. However, trinucleotide motif do distribute evenlyamong CDS and UTR indicating the existence of different selective forces in CDS andUTR for di- and trinucleotide motif.A total of 54 cucumber EST-SSR loci were selected for synthesis of flanking primersand polymorphism evaluation of those loci was carried out on 12 cucumber cultivars. Ofthe synthesized primers, two primer pairs had no products and three primer sets amplifiedaspecific products. For the rest primers with specific products, 23 primer pairs revealedpolymorphism on 20 EST-SSR loci. The maximum six alleles were detected on the 20loci with an average of 3.3 alleles per locus. The mean observed heterozygosity andexpected heterozygosity were 0.404 and 0.465 respectively. Thirteen EST-SSR locisignificant deviated from the Hardy-Weinberg equilibrium (P＜0.05). Ten SSR locishowed significant similarities to known genes or nucleotide sequence。Cross-speciesamplification showed that thirteen pairs of primer (65%) generated specific products inCucumis melo with seven of which (35%) detecting polymorphism.56 pairs of primer were synthesized according to the flanking sequences of melonEST-SSR loci. Of which, 4 pairs of primer amplified no products, 5 primer sets generatedaspecific products. A total of 47 pairs of primer successfully amplified the expectedproducts with 22 pairs of primer exhibiting polymorphism. The maximum number of 5alleles was detected on the 22 polymorphic EST-SSR loci with an average number of 2.9alleles per locus. The mean observed heterozygosity and expected heterozygosity were0.340 and 0.442 respectively. Seventeen EST-SSR loci significant deviated from theHardy-Weinberg equilibrium (P＜0.05). In addition, significant linkage disequilibria (P＜0.05) were observed between the loci of CM33 and CM01, CM16 and CM33. FifteenSSR loci showed significant similarities to known genes or nucleotide sequences.Cross-species amplification showed that 15 pairs of primer (68.2%) generated specificproducts in Cucumis sativus.The performance of EST-SSR in the application of phylogenetic analysis wasevaluated on cucumber and melon separately. The PIC value, scored on 12 cucumbercultivars by 20 cucumber EST-SSR markers, ranged from 0.153 to 0.733 with a mean of0.446. The pair wise similarity coefficient of Jaccard among the 12 cultivars ranged from0.142 to 0.935. The UPGMA dendrogram clearly separated the 12 cultivars into Europeoriginated cucumber germplasm group and domestic cucumber group with high Bootstrap value (＞99%). The Principle Coordinate Analysis further classified the domesticcucumber germplasm into types of cucumber with green skin and cucumber with whiteskin according to the fruit color, demonstrating the powerful discriminating capacity ofEST-SSR.The genetic relationship of 27 melon cultivars was also assessed with 22 melonEST-SSR markers. The PIC value scored by 22 EST-SSR markers on those cultivarsranged from 0.168 to 0.679 with an average of 0.433. The pair wise similarity coefficientof Jaccard revealed by the markers among the 27 cultivars ranged from 0.393 to 0.946.Both the UPGMA dendrogram and the Principle Coordinate Analysis discriminated thecultivars into the groups of melon with thin rind and melon with thick rind.