Variation in the branching of plant inflorescences determines flower number and, consequently, reproductive success and crop yield. Nightshade (Solanaceae) species are models for a widespread, yet poorly understood, program of eudicot growth, where short side branches are initiated upon floral termination. This “sympodial” program produces the few-flowered tomato inflorescence, but the classical mutants compound inflorescence ( s) and anantha ( an) are highly branched, and s bears hundreds of flowers. Here we show that S and AN, which encode a homeobox transcription factor and an F-box protein, respectively, control inflorescence architecture by promoting successive stages in the progression of an inflorescence meristem to floral specification. S and AN are sequentially expressed during this gradual phase transition, and the loss of either gene delays flower formation, resulting in additional branching. Independently arisen alleles of s account for inflorescence variation among domesticated tomatoes, and an stimulates branching in pepper plants that normally have solitary flowers. Our results suggest that variation of Solanaceae inflorescences is modulated through temporal changes in the acquisition of floral fate, providing a flexible evolutionary mechanism to elaborate sympodial inflorescence shoots.
Among the most distinguishing features of plants are the flower-bearing shoots, called inflorescences. Despite a solid understanding of flower development, the molecular mechanisms that control inflorescence architecture remain obscure. We have explored this question in tomato, where mutations in two genes, ANANTHA ( AN) and COMPOUND INFLORESCENCE ( S), transform the well-known tomato “vine” into a highly branched structure with hundreds of flowers. We find that AN encodes an F-box protein ortholog of a gene called UNUSUAL FLORAL ORGANS that controls the identity of floral organs (petals, sepals, and so on), whereas S encodes a transcription factor related to a gene called WUSCHEL HOMEOBOX 9 that is involved in patterning the embryo within the plant seed. (F-box proteins are known for marking other proteins for degradation, but they can also function in hormone regulation and transcriptional activation) Interestingly, these genes have little or no effect on branching in inflorescences that grow continuously (so-called “indeterminate” shoots), as in Arabidopsis. However, we find that transient sequential expression of S followed by AN promotes branch termination and flower formation in plants where meristem growth ends with inflorescence and flower production (“determinate” shoots). We show that mutant alleles of s dramatically increase branch and flower number and have probably been selected for by breeders during modern cultivation. Moreover, the single-flower inflorescence of pepper (a species related to tomato, within the same Solanaceae family) can be converted to a compound inflorescence upon mutating its AN ortholog. Our results suggest a new developmental mechanism whereby inflorescence elaboration can be controlled through temporal regulation of floral fate.
Plant flower production is largely determined by the number of inflorescences, the branches produced on flower stems. Two genes identified in tomato reveal a new phase transition that may explain the mechanism of evolution of compound inflorescences in the Solanaceae family.