valuable tool for analyzing the genetic basis of flower 

 development in many other species, as well. 



But as powerful and useful as the ABC model is, 

 flower forms are more complicated than ABC. 

 In recent years several discoveries have forced plant 

 biologists to reevaluate the model. First, new genes 

 were discovered that are necessary for flower organs 

 to form, but that have a different function than A, 

 B, or C genes. In 2000 Soraya Pelaz and Martin F. 

 Yanofsky, both plant biologists at the University of 

 California, San Diego, and several colleagues dis- 

 covered three closely related genes (called SEPAL- 

 LATA) in arabidopsis that were needed for petals, 

 stamens, and carpels to form. Without the SEPAL- 

 LATA genes, sepals grow in all four flower whorls. 

 The investigators designated the new function E (D 

 had already been applied to an aspect of ovule for- 

 mation), and concluded that E genes join A, B, and 

 C genes as the parties responsible for forming the 

 inner three flower whorls. 



Then, in 2001, Takashi Honma and Koji Goto, 

 both plant biologists at Kyoto Uni- 

 versity in Japan, discovered that when 

 A, B, and E genes are all activated during 

 leaf development, a petal forms in- 

 stead of a leaf. This discov- 

 , ery, in addition to con- 



firming the E function in 

 flower formation, provided the 

 first experimental evidence for one 

 of the oldest and best-known 

 theories in botany. In 1790 the 

 German philosopher, poet, and 

 polymath Johann Wolfgang von 

 Goethe outlined his theory that 

 all plant parts — and flower 

 parts in particular — are 

 actually modified 

 leaves. Thus the dis- 

 covery of E genes 

 showed that, as 



Goethe aphorized two centuries ago, "Alles ist Blatt" 

 (all is leaf). 



Another challenge to the ABC model is its failure to 

 identify genes that perform function A in species other 

 than arabidopsis. B, C, and E genes, however, have 

 been confirmed in other species. Oddly, the species 

 that serves as an experimental model for the entire 

 plant kingdom appears to be unique in having A genes 

 (a condition it most likely shares with other members 

 of the mustard family). Why should that be so? 



The evidence points to a phenomenon known 

 as gene duplication. Throughout evolution, it has 

 been extremely common for groups of genes, or 

 even an organism's entire genome, to double, giv- 

 ing rise to two copies of every duplicated gene. 

 The copies are usually unnecessary, and 

 often they are simply lost. Sometimes, 

 however, the "new" copy, since it 

 serves no critical function, can ac- 

 cumulate mutations, or random 

 changes, in its DNA sequence. 



Most of the mutations render a 

 gene copy useless, or even harm- 

 ful, but in some cases they give a 

 gene new capabilities. Occasion- 

 ally, they may even bring about 

 profound changes — perhaps a new 

 flower form, or even a brand-new spe- 

 cies. Over time, in fact, gene duplication has pro- 

 vided the genetic raw material for the evolution of 

 diversity and complexity. The /I, B, C, and E genes, 

 for instance, have all undergone multiple duplica- 

 tions during their evolutionary history. Since these 

 genes are transcription factors that turn molecular 

 cascades on and off, their duplications may have 

 been particularly important in the evolution of new 

 flower forms. 



Duplications in the evolutionary history of the 

 A genes may explain why arabidopsis (and 

 probably other mustards) is alone among flowering 

 plants in having A genes. When an A gene called 

 APETALA1 (API for short) is experimentally inacti- 

 vated, stamens and carpels form normally But in- 

 stead of sepals, modif ied leaves grow, and instead of 

 petals, there are branches bearing additional unusual 

 flowers. The pattern of branching flowers growing 

 from other branching flowers repeats several times. 



The role of API in forming sepals and petals in 

 arabidopsis is largely responsible for introducing the 

 concept of A genes. In other species, genes homol- 

 ogous to API — genes, that is, that share a common 

 ancestor with API — do not perform the same role. 

 Even in the snapdragon, the other species on which 

 the ABC model was based, the .-I/'/ homolosue, 



Morning 

 glory 



June 2006 NATUH.AI HISTORY 37 



