USE OF SHORT MICROSATELLITES FROM DATABASE SEQUENCES TO GENERATE POLYMORPHISMS AMONG LYCOPERSICON-ESCULENTUM CULTIVARS AND ACCESSIONS OF OTHER LYCOPERSICON SPECIES

A search of nearly 2000 sequences from Solanacene species in the EMBL and Genbank databases yielded 220 microsatellites. Among these were 80 microsatellites from 675 Lycopersicon entries. Dinucleotide repeats, as well as (CAA)(n) and (TAA)(n) repeats, were over-represented in non -coding DNA. The other trinucleotide repeats were predominantly found in exonic DNA. PCR analysis of 44 of the microsatellite-containing Lyc opersicon loci identified 36 primer pairs that yielded well-scorable f ragments, or groups of fragments, in L. esculentum cultivars and acces sions of Lycopersicon species. Twenty-nine of these amplified bands th at were polymorphic among the four Lycopersicon species. Ten primer pa irs generated polymorphic bands among seven tomato cultivars. Upon exa mining the number of microsatellites and the degree of polymorphisms i n relation to the repeat type and motif, the type of DNA the microsate llite resided in, the length of the microsatellite, and the presence o f imperfections in the microsatellite, only two significant correlatio ns were found. (i) Imperfect repeats were less polymorphic among speci es than perfect repeats. (ii) The percentage of loci polymorphic among cultivars increased from 6% for the shortest loci (with eight or less repeat units) to 60% for the group with the longest repeats (12 repea t units or longer). Among the species, however, all length classes con tained about 83% polymorphic loci. In general, 2-4, alleles were found for each locus among the samples of the test set. In a few cases, up to eight alleles were found. A combination of these microsatellite loc i can therefore be useful in distinguishing cultivars of tomato, which are genetically very closely related to each other.