Birth, life and death of developmental control genes: new challenges for the homology concept.

Understanding the interrelationship between the phylogeny of developmental control genes and the evolution of morphological features is a central goal of evolutionary developmental biology (evo-devo). It requires that one distinguishes properly between gene genealogy and function. Gene duplication, gene loss and speciation in combination with differential changes in gene function can generate complex evolutionary scenarios that require additional terms beyond homology for a proper description. Use and possible misuse of these terms, including "orthology", "paralogy" and "subfunctionalization", is exemplified with AGAMOUS-like genes encoding transcription factors involved in flower and fruit development. This MADS-box gene subfamily demonstrates that homologous genes in different species with (almost) identical functions can be paralogues rather than orthologues, corroborating that functional similarity of genes is not a valid criterion for orthology. Homeosis fails some tests of homology, but might be of greater evolutionary importance than previously assumed, justifying yet another term, "homocracy". It describes organs that share the expression of the same patterning genes, irrespective of the homology of these organs. All in all this article opts for a careful use of a limited and well-chosen set of terms describing gene relationships and function, rather than the inflationary production of novel terms that may seem to be precise, but whose obscurity hampers communication.