Inevolutionary developmental biology,homeosis is the transformation of oneorgan into another, arising frommutation in or misexpression of certain developmentally criticalgenes, specificallyhomeotic genes. In animals, these developmental genes specifically control the development of organs on their anteroposterior axis.[1] In plants, however, the developmental genes affected by homeosis may control anything from the development of astamen or petals to the development of chlorophyll.[2] Homeosis may be caused by mutations inHox genes, found in animals, or others such as theMADS-box family in plants. Homeosis is a characteristic that has helpedinsects become as successful and diverse as they are.[3]
Homeotic mutations work by changing segment identity during development. For example, theUltrabithoraxgenotype gives aphenotype wherein metathoracic and firstabdominal segments become mesothoracic segments.[4] Another well-known example isAntennapedia: again-of-functionallele causes legs to develop in the place ofantennae.[5]
Inbotany,Rolf Sattler has revised the concept of homeosis (replacement) by his emphasis on partial homeosis in addition to complete homeosis;[6] this revision is now widely accepted.
Homeotic mutants inangiosperms are thought to be rare in the wild: in the annual plantClarkia (Onagraceae), homeotic mutants are known where the petals are replaced by a second whorl of sepal-like organs, originating in a mutation of a single gene.[7] The absence of lethal or deleterious consequences in floral mutants resulting in distinct morphological expressions has been a factor in the evolution ofClarkia, and perhaps also in many other plant groups.[8]
Following the work on homeotic mutants byEd Lewis,[9] the phenomenology of homeosis inanimals was further elaborated by discovery of a conserved DNA binding sequence present in many homeotic proteins.[10] Thus, the 60 amino acid DNA binding protein domain was named thehomeodomain, while the 180bp nucleotide sequence encoding it was named thehomeobox. The homeobox gene clusters studied by Ed Lewis were named theHox genes, although many more homeobox genes are encoded by animalgenomes than those in the Hox gene clusters.
The homeotic-function of certain proteins was first postulated to be that of a "selector" as proposed byAntonio Garcia-Bellido.[11] By definition selectors were imagined to be (transcription factor) proteins that stably determined one of two possible cell fates for a cell and its cellular descendants in atissue. While most animal homeotic functions are associated with homeobox-containing factors, not all homeotic proteins in animals are encoded by homeobox genes, and further not all homeobox genes are necessarily associated with homeotic functions or (mutant) phenotypes.The concept of homeotic selectors was further elaborated or at least qualified byMichael Akam in a so-called "post-selector gene" model that incorporated additional findings and "walked back" the "orthodoxy" of selector-dependent stable binary switches.[12]
The concept oftissue compartments is deeply intertwined with the selector model of homeosis because the selector-mediated maintenance of cell fate can be restricted into different organizational units of an animal'sbody plan.[13]In this context, newer insights into homeotic mechanisms were found byAlbert Erives and colleagues by focusing onenhancer DNAs that are co-targeted by homeotic selectors and different combinations of developmental signals.[14]This work identifies a protein biochemical difference between thetranscription factors that function as homeotic selectors versus thetranscription factors that function as effectors of developmental signaling pathways, such as theNotch signaling pathway and theBMP signaling pathway.[14]This work proposes that homeotic selectors function to "license"enhancer DNAs in a restrictedtissue compartment so that the enhancers are enabled to read-out developmental signals, which are then integrated viapolyglutamine-mediated aggregation.[14]
Like the complexmulticellularity seen inanimals, the multicellularity ofland plants is developmentally organized intotissue andorgan units viatranscription factor genes with homeotic effects.[15]Although plants have homeobox-containing genes, plant homeotic factors tend to possessMADS-box DNA binding domains.Animal genomes also possess a small numberMADS-box factors.Thus, in the independent evolution ofmulticellularity in plants and animals, differenteukaryotic transcription factor families wereco-opted to serve homeotic functions. MADS-domain factors have been proposed to function as co-factors to more specialized factors and thereby help to determine organ identity.[15]This has been proposed to correspond more closely to the interpretation of animal homeotics outlined byMichael Akam.[16]