Genetic mapping of the tomato Epinastic ( Epi ) locus Lee, S . , Yen, H - C . , and Giovannoni, J . * Department of Horticultural Sciences and Crop Biotechnology Center, Texas A&M University, College Station, TX   77843 - 2133 * corresponding author    jjg@unix . tamu . edu The tomato Epinastic ( Epi ) mutation was originally characterized as a semi - dominant, single locus mutation resulting in leaf epinasty, vertical growth, minimal branching, and highly branched root structure (1,2) .   These effects are consistent with ethylene over - production or constitutive ethylene signaling (3) .   Although elevated ethylene biosynthesis has been reported in some tissues of the Epi mutant, treatment with inhibitors of ethylene biosynthesis or action had little effect on mutant phenotype, suggesting that Epi represents a lesion in ethylene signal transduction (4) .   The Arabidopsis ctr1 mutant is also characterized by constitutive ethylene signal transduction, and the corresponding CTR1 gene has been isolated and shown to have homology to the Raf family of protein kinases (5) .   We report here genetic mapping of the Epi locus as a first step for testing linkage with tomato CTR1 - related sequences that represent candidates for the EPI gene, or alternately, for positional cloning of the Epi locus should none of the candidate genes co - segregate with Epi . While it is possible that Epi may represent a tomato homologue of the Arabidopsis CTR1 gene, Epi alternatively may represent an ethylene signal transduction component whose Arabidopsis counterpart remains to be identified or does not exist . Tomato cultivar VFN8 ( Epi/Epi ) was kindly provided by V . Ursin and crossed to the wild tomato relative L . cheesmanii (LA483; epi/epi ) to facilitate RFLP mapping .   Resulting F1 progeny were selfed and an F2 population of 962 plants was scored for the presence or absence of leaf epinasty .   A total of 123 mutant individuals were identified (or approximately half the number expected for a recessive mutation) .   In this regard it is noteworthy that poor transmission of the recessive Arabidopsis ctr1 mutant allele has also been reported (5) .   It is also important to note that we did not observe any effects of the mutant allele in the original F1 individuals, supporting the concept that the mutant phenotype results from a recessive allele, and in contrast to previous reports that Epi is semi - dominant (1,2) .   It is noteworthy that L . cheesmanii was used as the normal parent in the cross as opposed to L . esculentum parents in previous studies of Epi dominance, and thus that alleles derived from L . cheesmanii may have influenced