Sherry A. Kempin

Molecular basis of the cauliflower phenotype in Arabidopsis

Genetic studies demonstrate that two Arabidopsis genes, CAULIFLOWER and APETALA1, encode partially redundant activities involved in the formation of floral meristems, the first step in the development of flowers. Isolation of the CAULIFLOWER gene from Arabidopsis reveals that it is closely related in sequence to APETALA1. Like APETALA1, CAULIFLOWER is expressed in young flower primordia and encodes a MADS-domain, indicating that it may function as a transcription factor. Analysis of the cultivated garden variety of cauliflower (Brassica oleracea var. botrytis) reveals that its CAULIFLOWER gene homolog is not functional, suggesting a molecular basis for one of the oldest recognized flower abnormalities.

Control of Fruit Patterning in Arabidopsis by INDEHISCENT

The Arabidopsis seedpod opens through a spring-loaded mechanism known as pod shatter, which is essential for dispersal of the seeds. Here, we identify INDEHISCENT (IND), an atypical bHLH protein, that is necessary for fruit opening and is involved in patterning each of the three fruit cell types required for seed dispersal. Previous studies suggested that FRUITFULL (FUL), a member of the MADS-domain transcription factor family, is required for fruit growth since ful mutant fruit fail to undergo the dramatic enlargement that normally occurs after fertilization. Here we show, however, that FUL is not directly required for fruit elongation and instead is required to prevent ectopic activity of IND. Our molecular and genetic studies suggest a model for the regulatory interactions among the genes that control fruit development and the mechanism that results in the expression of IND in a narrow stripe of cells.

Manipulation of flower structure in transgenic tobacco

Genetic studies suggest that three homeotic functions, designated A, B, and C, act alone and together to specify the fate of floral organ primordia in distantly related dicotyledonous plant species. To test the genetic model, we have generated transgenic tobacco plants that ectopically express the AGAMOUS gene from Brassica napus, which is necessary for the C function. Flowers on the resulting plants showed homeotic transformations of sepals into carpels and petals into stamens. These phenotypes are consistent with predictions from the genetic model, show that expression of AGAMOUS is sufficient to provide ectopic C function, and demonstrate that the structure of flowers can be manipulated in a predictable manner by altering the expression of a single regulatory gene. Furthermore, the generation of the predicted transformations by ectopic expression of the Brassica gene in transgenic tobacco indicates that gene functions are interchangeable between phylogenetically distant species.

Targeted disruption in Arabidopsis

Homologous recombination has been used for two decades to target insertions into cloned genes in bacteria and yeast, and more recently has become a routine method of gene inactivation in mammals. Arabidopsis is one of several multicellular model organisms (along with Drosophila, Caenorhabditis and zebrafish) in which mechanisms controlling development have been studied. Previously, traditional genetic methods have been used, as targeted disruption by homologous recombination has not been successful in any of these organisms. We have now successfully disrupted the AGL5 MADS-box gene in Arabidopsisby homologous recombination, providing a useful tool for future analyses.

Conversion of Perianth into Reproductive Organs by Ectopic Expression of the Tobacco Floral Homeotic Gene NAG1

Mutations in the AGAMOUS (AG) gene of Arabidopsis thaliana result in the conversion of reproductive organs, stamens and carpels, into perianth organs, sepals and petals. We have isolated and characterized the putative AG gene from Nicotiana tabacum, NAG1, whose deduced protein product shares 73% identical amino acid residues with the Arabidopsis AG gene product. RNA tissue in situ hybridizations show that NAG1 RNA accumulates early in tobacco flower development in the region of the floral meristem that will later give rise to stamens and carpels. Ectopic expression of NAG1 in transgenic tobacco plants results in a conversion of sepals and petals into carpels and stamens, respectively, indicating that NAG1 is sufficient to convert perianth into reproductive floral organs.

Brassicaceae INDEHISCENT genes specify valve margin cell fate and repress replum formation

Members of the Brassicaceae family, including Arabidopsis thaliana and oilseed rape (Brassica napus), produce dry fruits that open upon maturity along a specialised tissue called the valve margin. Proper development of the valve margin in Arabidopsis is dependent on the INDEHISCENT (IND) gene, the role of which in genetic and hormonal regulation has been thoroughly characterised. Here we perform phylogenetic comparison of IND genes in Arabidopsis and Brassica to identify conserved regulatory sequences that are responsible for specific expression at the valve margin. In addition we have taken a comparative development approach to demonstrate that the BraA.IND.a and BolC.IND.a genes from B. rapa and B. oleracea share identical function with Arabidopsis IND since ethyl methanesulphonate (EMS) mutant alleles and silenced transgenic lines have valve margin defects. Furthermore we show that the degree of these defects can be fine-tuned for crop improvement. Wild-type Arabidopsis produces an outer replum composed of about six cell files at the medial region of the fruits, whereas Brassica fruits lack this tissue. A strong loss-of-function braA.ind.a mutant gained outer replum tissue in addition to its defect in valve margin development. An enlargement of replum size was also observed in the Arabidopsis ind mutant suggesting a general role of Brassicaceae IND genes in preventing valve margin cells from adopting replum identity.

Genes Specifying Floral Meristem Identity in Arabidopsis

Genetic studies indicate that the Arabidopsis floral homeotic gene APETALA1 is one of several genes involved in the generation of floral meristems, the first step in flower development. Molecular analyses indicate that APETALA1 RNA is expressed in young floral primordia and in sepals and petals, and that APETALA1 encodes a transcription factor with a MADS-domain. These molecular studies, together with the apetalal mutant phenotype, suggest that APETALA1 acts locally to specify the identity of the floral meristem, and to determine sepal and petal development. Further studies demonstrate that the organ identity gene AGAMOUS negatively regulates APETALA1 RNA accumulation in the two inner whorls of wild-type flowers. Here we discuss the molecular characterization of the APETALA1 gene and its possible interactions with other genes involved in the specification of floral meristem identity in Arabidopsis.

The Role of MADS-Box Genes in the Control of Flower and Fruit Development in Arabidopsis

Arabidopsis flowers and fruit are typical of the more than three thousand species of Brassicaceae and have been the subject of intensive genetic and molecular studies. Among the many genes that have been identified that control various aspects of flower and fruit development, the MADS-box family has been shown to play central roles. MADS-box genes encode putative transcriptional regulators that play regulatory roles not only in diverse plant species, but also in fungal and animal development. The first Arabidopsis MADS-box gene to be cloned was AGAMOUS in 1990, and in the ensuing years, dozens of related genes have been cloned and functionally characterized (Yanofsky et al., 1990; Riechmann and Meyerowitz, 1997; Theissen, 2000). Three general lessons have been learned from these functional studies. The first, is that MADS-box genes play diverse roles in plant development, ranging from the control of flowering time, meristem identity, organ identity, fruit development, and they also appear to play roles during embryo, ovule, seed, root, stem and leaf development. A second lesson that we have learned is that MADS-box genes frequently play multiple roles during development. For example, the FRUITFULL gene is involved in leaf development as well as in fruit development. A third lesson is that functional redundancy is prevalent among MADS-box genes. In some cases, single mutants are indistinguishable from the wild type, whereas double mutants carrying mutations in two closely related MADS-box genes display striking phenotypic abnormalities. In this manuscript, we will focus on recent examples from our laboratory that highlight these three basic conclusions of MADS-box gene function.