Dimerization of Floral Organ Identity Proteins in Arabidopsis

Dartmouth College, Hanover NH

Investigators

Abstract

Considerable evidence supports the hypothesis that the gene products and mechanisms that control flower development in different flowering plant species are largely conserved. Many plant developmental control proteins contain a conserved DNA binding domain called the MADS domain. The MADS family in Arabidopsis consists of more than 30 genes. MADS proteins are involved in diverse aspects of plant development including flowering time control, meristem specification, floral organ identity, and fruit development. Despite the importance of these MADS proteins in plant development, little is known about how these proteins function and interact. This proposal focuses on the floral organ identity MADS proteins APETALA3 (AP3) and PISTILLATA (PI) in Arabidopsis thaliana. AP3 and PI function to direct the development of petal and stamens in the Arabidopsis flower. All MADS proteins bind to DNA either as heterodimers or homodimers. Most MADS proteins in Arabidopsis are capable of forming both homodimers and heterodimers with a variety of partner proteins. AP3 and PI, by contrast, are unable to homodimerize and form an obligate heterodimer. The AP3/PI heterodimer is highly specific and is evolutionarily conserved as evidenced by a similar interaction between the AP3 and PI homologs in the distantly related plant species Antirrhinum majus. This project is focused on identifying the amino acids and subdomains in the AP3 and PI proteins that mediate the highly specific heterodimer interaction between AP3 and PI; this will be done both by site-specific mutagenesis and by utilizing "reverse" yeast two-hybrid approaches. Through this research, we hope to be able to define the rules for dimerization of this importance class of plant developmental regulators. A second focus of this research centers on an unusual allele of PI called pi-5. Unlike all other ap3 and pi mutants, pi-5- exhibits phenotypic defects only in a single whorl of the flower. The PI-5 protein exhibits defects in protein-protein interaction with a wild-type AP3 partner protein. Genetic experiments with pi-5 suggest that whorl-specific factors modulate the activity of PI-5. Using both genetic and molecular approaches, whorl-specific factors that affect floral organ identity will be isolated and characterized.