We analyzed recessive mutants of two homeotic genes in rice, SUPERWOMAN1 ( SPW1 ) and DROOPING LEAF ( DL ). The homeotic mutation spw1 transforms stamens and lodicules into carpels and palea-like organs, respectively. Two spw1 alleles, spw1-1 and spw1-2 , show the same floral phenotype and did not affect vegetative development. We show that SPW1 is a rice APETALA3 homolog, OsMADS16 . In contrast, two strong alleles of the dl locus, drooping leaf-superman1 ( dl-sup1 )and drooping leaf-superman2 ( dl-sup2 ), cause the complete transformation of the gynoecium into stamens. In these strong mutants, many ectopic stamens are formed in the region where the gynoecium is produced in the wild-type flower and they are arranged in a non-whorled, alternate pattern. The intermediate allele dl-1 ( T65 ), results in an increase in the number of stamens and stigmas, and carpels occasionally show staminoid characteristics. In the weakest mutant, dl-2 , most of the flowers are normal. All four dl alleles cause midrib-less drooping leaves. The flower of the double mutant, spw1 dl-sup , produces incompletely differentiated organs indefinitely after palea-like organs are produced in the position where lodicules are formed in the wild-type flower. These incompletely differentiated organs are neither stamens nor carpels, but have partial floral identity. Based on genetic and molecular results, we postulate a model of stamen and carpel specification in rice, with DL as a novel gene controlling carpel identity and acting mutually and antagonistically to the class B gene, SPW1 .

Cells in developing organisms do not only differentiate, they differentiate in defined patterns. A striking example is the differentiation of flowers, which in most plant families consist of four types of organs: sepals, petals, stamens and carpels, each composed of characteristic cell types. In the families of flowering plants in which these organs occur, they are patterned with the sepals in the outermost whorl or whorls of the flower, with the petals next closest to the center, the stamens even closer to the center, and the carpels central. In each species of flowering plant the disposition and number (or range of numbers) of these organs is also specified, and the floral ‘formula’ is repeated in each of the flowers on each individual plant of the species. We do not know how cells in developing plants determine their position, and in response to this determination differentiate to the cell types appropriate for that position. While there have been a number of speculative proposals for the mechanism of organ specification in flowers (Goethe, 1790; Goebel, 1900; Heslop-Harrison, 1964; Green, 1988), recent genetic evidence is inconsistent with all of them, at least in the forms in which they were originally presented (Bowman et al. 1989; Meyerowitz et al. 1989). We describe here a preliminary model, based on experiments with Arabidopsis thaliana . The model is by and large consistent with existing evidence, and has predicted the results of a number of genetic and molecular experiments that have been recently performed.