The making of a blossom: a flower's evolutionary past may be read in the genes that influence its development

Natural History, May, 2002 by Enrico Coen

Even without revealing just what that role may be, however, the cyc research to date has provided an important clue as to why bilateral asymmetry in flowers has evolved so many times. Since the asymmetric pattern of cyc activity is found in both snapdragons and Arabidopsis, it was presumably also present in their most recent common ancestor, a plant that would have lived about 100 million years ago. This ancestral plant probably had radially symmetrical flowers, and thus, as in Arabidopsis, the cyc gene must have had a different role to play. Whatever its role, the asymmetric pattern of cyc activity meant that, in terms of gene activity, the ancestor's flowers were already asymmetric from top to bottom. This may have made it relatively easy for differences between upper and lower petals to evolve numerous times in the descendants of the ancestral plant, through minor modifications in cyc or in the genes that respond to cyc.

The key point here is that much of what seems novel in the appearance of an organism stems from ancient patterns of gene activity manifesting themselves in new ways, rather than from the invention of something completely new. And we do not have to go back millions of years to find evidence of the importance of changes involving regulatory genes. A more recent example is the domestication of maize (corn) by the prehistoric peoples of Mexico. The maize we cultivate today has one main stem, from which grow large cobs with lots of accessible, nutritious seeds (the kernels). Teosinte--maize's nearest living wild relative--looks very different; it is a highly branched plant with relatively small cobs, each of which bears a few seeds that have a hard, inedible covering.

About ten years ago, John Doebley, then at the University of Minnesota, and colleagues, building on earlier work by George Wells Beadle, showed that changes in as few as five genes could convert teosinte into a useful food plant like maize. Recently, Doebley's group went on to isolate one of these genes. Called teosinte-branched, or tb1, this gene is largely responsible for the difference in branching patterns between maize and teosinte. As might have been expected from the appearance of the plant, tb1 was found to be most active in the developing side buds. Quite unexpectedly, however, the DNA sequence of tb1 turned out to be very similar to that of the cyc gene of Antirrhinum, and like cyc, tb1 seems to be a regulatory gene.

In the process of domesticating maize, the ancient peoples of Mexico seem to have chosen a plant with a mutant form of the tb1 gene that was particularly effective at preventing side buds from developing into long branches. They were unwittingly playing with regulatory genes, much as may have happened naturally in the evolution of bilateral symmetry. And the evolution of maize has another parallel with that of floral asymmetry: both enabled plants to establish new associations with animals--humans in one instance, insects in the other. By studying these genes, we are revealing not only the history of changes in plant development but also something of the habits and predilections of the animals that interacted with them.