BIOL 606 Home

Hox Genes and the Evolution of Development

Lecture by Corey Davis

Based on the focal paper by Holland and Garcia-Fernandez 1996, concerning Hox genes and the Evolution of Development

Rapporteur: Stefanie Zaklan

Viewing the fields of Evolution and Development as interdisciplinary studies or separate lines of thought, has vacillated throughout the last century. In the 1930's the once highly acclaimed Law "Ontogeny recapitulates phylogeny" slowly eroded into two areas of interest, development and evolution, each possessing their own vocabulary, journals and ardent defenders. This speciation event, due to loss of intellectual hybridization, resulted in several potentially harmful ramifications. First, the study of macroevolution was no longer considered a discipline on its own, but instead an accumulation of microevolutionary events. Second, understanding homology was no longer deemed necessary to understand evolutionary patterns, as selection acted on differences in allele frequencies (variation). Finally, the basal level of selection was thought to occur at the genomic level (see Dawkins, The Selfish Gene 1976) and thus excluded the possibility that cellular interaction could produce large morphological changes (i.e. cellular level of selection was not relevant). Recently, with the onset of comparitive embryology in evolution, these aforementioned problems were rectified as the two disciplines were again united.

As the fields of developmental and evolutionary biology continue to converge, an underlying pattern is beginning to emerge: namely that the astonishing diversity of morphological variation in animals is built on a scaffolding of a seemingly quite limited set of developmental programs, such as the homeobox gene cluster (defined by a 180 bp DNA segment shared by all homeobox genes; the four groups of Hox genes are: 1) Dorsal-ventral patterning 2) paired structures, e.g. Pax-6 3) Segment polarity eg. Distaless and 4) Anteroposterior patterning e.g. Antennapedia). The homeobox gene expression data and organizational patterns have been found to be similar throughout the Metazoa, but are these homologous (resemblance caused by a continuity of information)? If Hox genes mark fixed axial positions in different species and control the development of appropriate phenotypic characters, they will not be good indicators of homology. Recently however, experimental evidence has revealed a tight correlation between vertebral identity and Hox gene expression, suggesting that these gene clusters may be good indicators of homology. However, not all are convinced that homology and not homoplasy (resemblance by a discontinuity of information) explain these patterns.

The recent reconciliation between the two disciplines has also allowed interchange of methodology. First, phylogenetic inferences using sequence information have been used to study the evolution of Homeobox gene clusters (e.g. collagen). These cladograms suggest that there is very little sequence variation amongst high level taxa, and that duplication and divergence has played a large role in Homeobox evolution. Plotting Hox gene numbers and morphological complexity on phylogenies suggested that there is a positive correlation between these two characters. Second, using gene expression data, organization patterns have been found to be similar within the Metazoa suggesting homology exists within clusters (Paralogue) and between taxa (Orthologue). Pax-6 is a pair rule Homeobox gene involved in the development of visual structures. This gene has been isolated from three protostomes and two deuterostome lineages and homology has been advocated in explaining this wide distribution.

Overall, homeobox genes have a very conserved role in the development of the metazoans (o.k..... so only 3 of the 35 phyla have been studied in any detail). Further investigations into the functioning and the possibility of homology of gene clusters will be the focus of many papers for years to come.

Discussion summary

Discussants: Jan Jekielek and Mark Steinhilber

Q: Are Hox genes useful in inferring phylogenetic relationships? Is timing of genomic expression different in different organisms? If this timing is modified does that mean homology does not exist and thus this information cannot be useful for phylogenies? What information can you obtain from Hox genes, is this just a band wagon of information with no direction? Perhaps we should be trying to understand the function of hox genes then we can infer phylogenies. We need to understand Hox functioning across the metazoa and not only within several well-analyzed taxa. We also need to understand why, within one phyla, we have different numbers of Hox clusters (6 clusters in fish versus 4 in mammals). Thus, duplication is not a mere precursor to large morphological changes suggesting that this information is not phylogenetically useful. Basically, we can all agree that if one was to use Hox genes to defend an unorthodox phylogeny the authors of the paper would be standing on shaky grounds.

Q: What is an orthologue? When a gene in one taxa (e.g. Hox 1A in Drosophila) is homologous to a gene in another taxa (e.g. Hox 1A in mouse).

Q: What is a Paralogue? A homologous gene in the same gene cluster. Thus, within a Hox cluster, two similar genes may be identical by descent (i.e. Homologous) and thus would be considered paralogues of each other.

Q: What does the word homology mean to a molecular biologist? They have terms (Paralogue and Orthologue) to distinguish origins of the gene. They will use the word homology to imply similarity by descent (i.e. evolutionary origins of the molecule in question are taken into account) or similarity due to structural identity. Morphologists would suggest that similarity by decent is homology and similarity due only to structural identity should be termed analogy. In cladistics the words apomorphy (derived character) and pleisiomorphy (ancestral character) are used and may be better terminology for molecular works, as these terms are new and possess no ambiguity in their meaning (understanding if a character is derived or ancestral is, of course, another can of worms).

Q: At what level of selection can you truly say something is homologous? We must be careful with this term. For example, bats wings and birds wings can be considered analogous in their dermal coverings, but homologous in their underlying bone structure. What about the discovery that distaless is expressed in such diverse taxa as echinoderms, chordates and arthropods? In chordates and arthropods it codes for the distal end of the organism (i.e. fingertips). In echinoderms, antibody staining shows that distaless shows in their tube feet. Does this mean that the cellular components that make up fingertips and tube feet are homologous structures, or just that their placement is homologous? Homology is not an absolute property. There are levels.

Q: Do Arthropods (insects), Vertebrates (mice) and Molluscans have homologous eyes? If homology exits in the Pax-6 gene cluster (important in the early development of vision) between these phyla, then at some basal level these structures are homologous (at this point no mulluscan Hox genes have been isolated, thus this is purely speculative). As the basic functioning of the eyes are similar, they all function in capturing photons and transmitting the information to ganglia, this may suggest homology. However, it was pointed out that vertebrate eyes are mesodermally derived whereas in molluscans it is ectodermally derived. Thus the question of the level at which a structure is homologous rears its ugly head once again (i.e. genomic level-homologous; cellular-analagous; organismal-homologous).

Dawkins, R. The Selfish Gene. 1976. Oxford, Oxford University Press.

Holland, W. H. and J. Garcia-Fernandez. 1996. Hox Genes and Chordate Evolution. Developmental Biology. 173:382-395.

BIOL 606 Home