In 1991, researchers at the California Institute of Technology described the basic genetic system behind how flowers are made.  The “ABC model” that they proposed was so popular and successful that it was quickly taken up by the community and was even included in textbooks when I did my undergraduate studies just six years later. I remember being fascinated the first time I heard about it. This was the sort of thing that got me really excited — disparate facts were brought together in an elegant mechanism that could robustly set up a flower but was still flexible enough to create different forms. Now, over a decade later, I’d like to share this inspirational work with you.

What goes into building a flower, anyway? Some familiar flowers, like sunflower, daisy and gerbera, aren’t actually a single flower but rather a flower head (called pseudanthium) made up of many small individual flowers; the “petals” of these flowers are actually each an individual flower (called a “ray flower”) modified to look like a petal. More typical flowers, like a tulip or petunia,  are made up of four basic parts arranged in a series of rings, or whorls. The prominent, colourful petals which make up one whorl are probably the first thing that comes to most people’s mind when thinking about flowers. Just outside the petals are the sepals, small green coverings that protect the flower as a bud and support it in bloom; some flowers, like tulips, have modified sepals that look like petals. The male and female organs, the stamens and carpels, make up the inner two whorls encircled by the petals.  The challenge facing scientists was not only to explain this stereotypical arrangement, but to do so in a way which could account for the kinds of mutations they had seen.  One mutant, for example, lacks the innermost whorls containing the sexual organs; in their place the mutant plants form a second set of petals around an internal flower.  This kind of mutation is a favourite in ornamental flowers, since it produces blossoms with more petals, such as the double lily.  In another mutant, sepals and petals are replaced by carpels and stamens, respectively, while a third mutant replaces petals and stamens with sepals and carpels.  Despite their differences, these mutants suggest some kind of underlying order.  Each mutant shows defects in two adjacent sets of organs, transforming them into other organs in a systematic manner.

By studying these various mutants, the researchers came up with what has become known as the “ABC model” of floral development.  They suggested that three classes of genes interact to set up the pattern of a flower.  The class A genes are expressed in cells of the outer two whorls, the sepals and petals, while the class C genes are active in the cells which become stamens and carpels.  Genes from these two classes suppress one another; the negative interaction between them leads to their mutually exclusive domains in the outer and inner whorls.  The class B genes straddle the intersection between the two domains; these genes are expressed in cells which will become petals and stamens.  These overlapping domains make it possible for the three classes of genes to generate four different patterns of gene expression and thus four different cell types.  Cells with only class A genes become sepals, while those with genes from both class A & B turn into petals; likewise, those which have class B & C genes active will become stamens and cells with only class C genes turn into carpels.

The model also beautifully explains the order apparent in the mutants. Each mutant has a defect in a gene of one class and each class controls two adjacent whorls; when one of the classes is missing, the combination of genes needed for those whorls will be missing, transforming them into something different. Since the class A & C genes mutually suppress one another, absence of either one will lead to the spread of the other, resulting in doubled inner or outer whorls. For example, the double lily has a mutation in a class C gene, allowing class A genes to spread to the inner whorls.  The presence of A & B genes in the third whorl (rather than B & C genes) causes it to form petals rather than stamens; the fourth whorl, containing only class A genes, becomes the sepals of a new flower. The researchers were also able to use their model to predict what would happen when the mutants were combined to produce plants with two different mutations.  If both the class A and class B genes were defective, they reasoned, then all of the cells would have only class C genes and would turn into carpels.  The same reasoning could be used to predict the outcome of different combinations of mutants, up to plants with defective members of all three classes.  These triple mutants have flowers made from whorls of leaf-like organs, confirming a long-held suspicion that flowers are, essentially, modified leaves.

The ABC model is simple, beautiful and powerful.  It’s also basically correct, even in the extremely simplified form I’ve presented here.  The model has been elaborated in the 20 years since it was first proposed, with further research identifying the molecular nature of the three classes and clarifying how the class A & C genes repress each other, but the core idea has remained unchanged.  Looking back on their work today, the team pointed out that it “could have been carried out a century earlier”.  All of the technology, techniques and mutants they used were already widely available in the early and mid-20th century.  The strength of their model didn’t come from novel observations or a technological breakthrough, but rather from a different approach, a new conceptual framework in which to look at the data.  The kind of thinking that informed their model is commonplace in developmental biology today but it was still relatively new in the early 1990s. By interpreting existing data in a different way, these scientists were able to come up with a model that could explain their observations and make accurate new predictions. The ABC model isn’t just an elegant and appealing description of an developmental system, but also an example of how sometimes what’s needed most is a change of perspective.

Refs
Bowman JL, Smyth DR, & Meyerowitz EM (1991). Genetic interactions among floral homeotic genes of Arabidopsis. Development (Cambridge, England), 112 (1), 1-20 PMID: 1685111

Bowman, J., Smyth, D., & Meyerowitz, E. (2012). The ABC model of flower development: then and now Development, 139 (22), 4095-4098 DOI: 10.1242/dev.083972