Original URL: http://www.mpiz-koeln.mpg.de/~theissen/grouphome/evolution.html
It has been demonstrated that homeotic organ identity genes belonging to the MADS-box gene family sculpt the structure of dicotyledonous flowers. The establishment of floral homeotic genes and changes in their regulation or function, therefore, may have contributed to the establishment and structural evolution of flowers. Understanding the phylogeny of MADS-box genes might thus help us to gain a better understanding of flower origin and diversity.
We started to characterize the MADS-box gene family in phylogenetic informative taxa. In addition to our interest in the monocotyledonous plant maize, we also study MADS-box genes in the monocots lily and tulip. In that case we are especially interested to understand the molecular and genetic basis of the difference in perianth structures between lily/tulip and maize. While grasses such as maize have a highly reduced perianth, lilies and tulips have two whorls of well developed petaloid tepals to attract pollinators.
Moreover, we are currently analyzing MADS-box genes the gymnosperm Gnetum gnemon, in the ferns Ceratopteris and Ophioglossum and in the moss Physcomitrella patens. Our analyses enabled us already to get a rough reconstruction of the history of MADS-box genes, which is described below. Our current knowledge about MADS-box gene evolution in vascular plants is summarized.
We think it is quite safe to assume that the last common ancestor of extant plants, animals and fungi, which existed about one billion years ago, had already at least one MADS-box gene. Within the lineages leading to animals and fungi, MADS-box genes evolved into two different subfamilies (ARG80- and MEF2-like genes), whose members are involved in a variety of functions, including signal transduction, muscle development etc. Somewhere in the lineage leading to green plants genes appeared in which the MADS-box is followed by three other defined sequence elements, called I-, K- and C-region. These genes - we have termed them MIKC-type MADS-box genes, where "M" stands for "MADS-box" and "IKC" for the other regions mentioned above - were clearly homologs, but very likely not orthologs of the floral homeotic genes.
The last common ancestor of ferns and seed plants, which existed about 4oo
million years ago, had already at least two different genes of that type.
Gene duplications independent of those that gave rise to the variety of seed plant
MADS-box genes established large numbers of MADS-box genes in ferns, since we could
detect already more than 15 different gene family members in just one Ceratopteris
species. 11 of them have already been cloned, and turned out to be members of
subfamilies different from those of seed plants.
The last common ancestor of extant seed plants, which existed about 300 million years ago, possessed already members of at least six different gene clades also present in angiosperms, namely AGAMOUS-, AGL2-, AGL6-, DEF/GLO-, GGM13- and TM3-like genes. Representatives of three additional clades (AGL15-, AGL17- and SQUA-like genes) as well as separate clades of DEF- and GLO-like genes had already been established in the last common ancestor of monocotyledonous and eudicotyledonous flowering plants. Thus the last common ancestor of all higher angiosperms had already at least 10 different MIKC-type MADS-box genes, among them representatives of all the gene clades from which the floral homeotic genes have been recruited. This evidence, therefore, that the establishment of the MADS-type floral homeotic genes has been a prerequisite for the "invention" of flowers. On the other hand, it seems that the radiation of angiosperms and the diversification of the flower structure was mainly based on more subtle changes, e.g., in expression patterns and regulatory interactions between already existing floral homeotic genes, or on changes in the targets of these genes.
In future studies including gene cloning, expression analysis and functional tests in additional phylogenetic informative taxa, we want to see wether our speculations turn out to be correct.
for questions and remarks please send