Genes Directing Flower Development in Arabidopsis

In 1989, our lab was analyzing the induction of transcription of the maize ALCOHOL DEHYDROGENASE gene under low-oxygen conditions (Walker et al., 1987). At a similar time, the article “Genes Directing Flower Development in Arabidopsis” by Bowman, Smyth, and Meyerowitz (Bowman et al., 1989) was published in the first issue of The Plant Cell. This article was important as it presented a new approach to understanding gene-based plant development and expanded our thinking with regard to genes and plant processes. It changed our lab focus from analyzing promoters and environmental responses to investigating genes controlling development. The article also showed the power of Arabidopsis (Arabidopsis thaliana) as a research system and was instrumental in our switching from maize (Zea mays). The Bowman et al. study of plant development provided a clear demonstration of how analysis of gene action through mutants can help to unravel the mechanisms of plant development and showed how the actions and interactions of a few genes can control organ development in flowering, a major developmental phase in the life of a plant. Flowers develop from groups of undifferentiated cells that form on the sides of the inflorescence meristem; these cells must differentiate appropriately into the various floral organs. Bowman et al. (1989) describe the roles of four genes (APETALA2, APETALA3, PISTILLATA, and AGAMOUS) in which recessive mutations remove the normal roles of each of the genes in floral organ development. In the mutants, floral organs are formed in whorls where they do not occur in wild-type flowers. In each case, two adjacent whorls are disrupted within the flower primordium, and during floral development, the specific function of each of the genes became apparent. The article is accompanied by beautiful scanning electron micrographs (SEMs) illustrating the morphological outcomes in the floral apex in the single and double mutants, which makes the article accessible to a broad array of biologists (see figure). The SEMs are particularly valuable for defining the variation of morphological outcomes in mutant lines where two gene products are disrupted. The SEMs also help in understanding the morphological effects of cellular outcomes induced by alterations of growth temperatures in temperature-sensitive mutants, showing that the timing of action of genes occurs very early as primordia arise. SEM Micrographs of the Inner Surface of Organs Observed in the Third Whorl of apetala3 Flowers. Wild-type-like stamens at 16°C (A), carpelloid stamens (B), stamenoid carpels (C), deformed filaments (D), and carpel-like organs at 29°C (E). Bars = 100 μm in (A), (B), (C), and (E) and 30 μm in (D). (Reproduced from Bowman et al. [1989] , Figure 7. ) This article also emphasizes that mutations in particular genes can give rise to homeotic conversions where the lack of a gene product results in transformation of an organ; for example, in mutants of agamous (ag), stamens are transformed to petals. These transformations imply that the product of one gene can limit the field of expression of the product of another gene, when a loss-of-function mutation in the first gene gives rise to an expanded field of action of the interacting gene. An example is seen in the interaction between APETALA2 and AGAMOUS. The work described by Bowman et al. (1989), together with subsequent articles, many by the same laboratory (reviewed in Bowman et al., 2012), led to the formulation of a predictive model, wherein the flower primordium can be divided into concentric whorls and the homeotic genes are expressed in overlapping fields in two adjacent whorls. As the whorls and regions were defined, “the ABC model of floral organ identity” came into being. The ABC model was popularized in an article in Nature by Coen and Meyerowitz (1991), pointing out the similarities in flower development in Arabidopsis and Antirrhinum majus. The fields of expression of these genes either singly or together specify the organs in a whorl. The early-acting AP2 gene may specify a developmental state for those cells that will become whorls 1 and 2, and the AG gene may specify a different state for cells that will become whorls 3 and 4, with these two genes acting antagonistically. Bowman et al. (1989) pointed out that “there are few if any restrictions on the ultimate fate of the cells in any whorl and the cells of a whorl can take on any fate—leaves, sepals, petals, stamens or carpels.” The homeotic genes AP3, PI, and AG encode MADS box proteins. By contrast, AP2 encodes a different class of transcription factor. A subsequent article in The Plant Cell, “Control of Arabidopsis Flower and Seed Development by the Homeotic Gene APETALA2,” by Jofuku, den Boer, Van Montagu, and Okamuro (Jofuku et al., 1994) describes the isolation of the AP2 gene, which has a widespread mode of action that extends beyond the floral primordium. AP2 plays an important role in flower development in three ways: the establishment of the floral meristem, the specification of floral organ identity, and the temporal and spatial regulation of floral homeotic gene expression. Unlike the other floral genes described in the Bowman article, AP2 is also expressed outside the floral organs, such as in developing ovules, indicating a role in seed development. The similarity, across many other plant species, of floral organ position and similar homeotic mutants, where one organ is replaced by another, suggests that the ABC model operates widely in angiosperms. Bowman, Smyth, and Meyerowitz (Bowman et al., 1989) showed how floral development can be explored in terms of single and double mutants affecting gene actions and interactions. The principles described in this article can be applied to development of structures other than flowers, such as embryos, roots, and leaves. In our study of the flowering repressor FLC in Arabidopsis (Sheldon et al., 1999), we followed the principles outlined by Bowman et al. (1989). Apart from demonstrating the power of double mutants, the article also showed the value of detailed analyses of the organs and their morphogenesis, for example by SEM. Even though this article appeared 30 years ago, the basic determination of floral organs described in the ABC model still applies.