The Arabidopsis floral homeotic genes APETALA3 (AP3) and PISTILLATA (PI) are necessary and sufficient for petal and stamen identity within the flower. They encode MADS box transcription factors thought to act as obligate heterodimers. Reflecting this, loss of function mutations in either gene give similar phenotypes consisting of homeotic conversions of organs in the second and third whorls of their flowers: petals are converted to sepals and stamens are converted to carpels.

AP3 expression is restricted to flowers, largely to the second and third whorls of the flower where petals and stamens arise. Both loss of expression and ectopic expression of AP3 interferes with organ identity in the flower. Therefore, the proper transcriptional regulation of this gene is essential for flower development. Our work to date has concentrated on the initiation of AP3 expression within the petal and stamen domains. LEAFY (LFY), which encodes a plant specific transcription factor, directly regulates the expression of several homeotic genes within the flower. Our work has determined that LFY regulates the AP3 gene in a complex manner acting through a number of pathways (Lamb et al., 2002 ). Currently, we are investigating the mechanisms by which LFY specifically turns on AP3 only in the second and third whorls of the flower, and the potential role of post-translational modifications in this process. In addition, we are identifying LFY interacting proteins.

APETALA3/PISTILLATA function

C hanges in homeotic gene expression patterns or in the functions of the encoded proteins are thought to play a prominant role in the evolution of new morphologies. Previous work has suggested that differeneces in the expression patterns of AP3 and PI homologs are correlated with changes in floral structure in the angiosperms. We investiged the possiblity that changes in the functions of the encoded gene products may also have played a role in the evolution of different floral morphologies. The AP3 and PI proteins, like many other plant MADS box proteins, consist of a N-terminal MADS DNA binding domain, a short I (intervening) region, a K box, and a more divergent C-terminus. The AP3 and PI lineage genes are thought to represent paralogous gene lineages that arose from a duplication event prior to the orgin of the angiosperms. In addition, the AP3 lineage underwent another major duplication event at the base of the core eudicots, the group of plants that includes Arabidopsis, giving rise to two AP3 sub-lineages: the euAP3 and the TM6 gene lineages. Each of these gene lineages is characterized by short, lineage specific motifs found in the C-terminus of the predicted proteins. A series of deletion and domain swap constructs of the C-terminal motifs between Arabidopsis PI and AP3 and a paleoAP3 gene from a basal eudicot, Dicentra eximia. Analysis of both ectopic expression and complementation experiments revealed that the lineage specic motifs are necessary for function and have diverged in function. They are also necessary and sufficient for protein function specificity. These results suggest that the divergenece in these sequence motifs has contributed to the evolution of distinct functions for these floral homeotic gene products (Lamb and Irish, 2003, ).

Current work to follow up these results are concentrated on determining the functions of these C-terminal motifs of AP3 and PI. Currently, we have two possible functions for the motifs: recruitment of transcriptional co-activators or stabilization of larger ternary complexes (see figure). In order to identify possible co-activators, we have identified six putative APETALA3 BINDING PROTEINS (ABPs) by yeast two-hybrid screen. We are currently characterizing these proteins genetically and biochemically. In addition, we plan to look at the larger AP3/PI complex in planta and the possible role of the C-terminal motifs in this complex.

Transcriptional networks during flower development:
Arabidopsis 2010: Establishing Regulatory Networks in Arabidopsis: Integrating AGRIS with the Identification of Direct Targets for Transcription

In order to understand flower development, we will need to understand the gene networks underlying this process. To start to understand the gene network underlying flower development, we will identify direct targets of a selected group of transcription factors involved in flower development. This information will be integrated into the existing AGRIS (Arabidopsis Gene Regulatory Information Server).

More information on this project can be found at: http://arabidopsis.med.ohio-state.edu/NSF2010Project/