Gibberellins (GAs) are endogenous plant growth regulators that are involved in the regulation of many aspects of plant growth and development, including seed germination, extension growth and flowering. An approach to understand GA action is to isolate genes which are regulated by GAs and use these genes both as molecular markers for GA response and also to isolate the molecules which are involved in this regulation. GA-regulated genes are found in the aleurone layer of germinating cereal seeds, in vegetative shoot tissue and in flowers. We have initiated a project to investigate the role of GA in the promotion of anther development in the gib-1 mutant of tomato, Lycopersicon esculentum. This mutant is deficient in GAs because its ability to convert geranylgeranyl pyrophosphate to copalyl pyrophosphate is reduced. The phenotype of this mutant, which includes dwarfism, failure to germinate and failure to flower normally, is reversed by exogenously applied GAs, The anthers become developmentally arrested when the flower bud is 2-5 mm in length and remains responsive to a single treatment with 50 ng gibberellic acid (GA3) per bud until a length of 3-7 mm. Developmentally arrested anthers contain pollen mother cells which are in the G1 phase of the premeiotic interphase, and outer and inner tapetum cells which are at the uninucleate and binucleate stages, respectively (Jacobsen & Olszewski, 1991, Plant Physiol. 97: 409-414). The GA3 treatment of developmentally arrested flowers is known to cause specific changes in the gene expression of the gib-1 anthers (Jacobsen & Olszewski, 1994, Planta 192: 372-378). A differential screening of a gib-1 anther cDNA library made 48 hours after GA treatment resulted in the isolation of several cDNAs. The mRNAs of the corresponding cDNAs could be placed in two classes with respect to the kinetics of their accumulation patterns in the gib-1 flowers. The first class consists of genes that increased in expression 8 hours after GA3 treatment and they were maximally abundant either 24 or 48 hours post-treatment, while the second class was not detected until 48 hours post-treatment. The genes of class I already showed a response when arrested buds were treated with 0-5 ng GA3/bud, while at least a single treatment of 5 ng GA3/bud was necessary for rescuing normal flower development. In situ hybridization experiments showed that 24 hours after a single treatment with 50 ng GA3/bud the localization of the gene expression was restored as in wild-type. Also the timing of the formation of interlocking hairs on the mutant anthers after treatment was comparable to that in wild-type, while these hairs were never found on anthers of untreated buds.