For many years, developmental biologists, if challenged, have sought to justify the use of animal model systems by virtue of homology of form. This is likely to have been based loosely upon the a priori argument that, at least in early development, phenotypic similarities between human and non-human vertebrate species must reflect equivalence of the underlying generative mechanism. However, only slowly has evidence for such assumptions about homology of mechanism begun to accumulate. Perhaps one of the clearest early demonstrations of this relates to the zone of polarizing activity (ZPA) - the region of a limb bud which, by release of a diffusible morphogen, polarizes the distal part of the growing limb and controls the ante-roposterior pattern of digits (see Chapter 7). It was found that the ZPA taken from a human limb bud will, when assayed by grafting ectopically into a chick embryo wing bud, display the same activity as the equivalent region of a chick bud. Extra digits are formed in a predictable and organized fashion by the host, demonstrating that human and chicken ZPAs produce the same morphogen, but chicken host cells respond to it by forming additional chick digits (Fallon and Crosby, 1977). In other words, there is an equivalence of mechanism in the building of this particular bit of anatomy. Although this example deals with just a small part of the body plan (digit specification), it can be seen as exemplifying a widely held belief that similar equivalences exist at the mechanistic level in the building of much of the anatomy or, at least, that portion of it which is characteristically and uniquely 'vertebrate' in character.
This type of assumption has been cautiously held for a number of years and, in a rather piecemeal and limited fashion, evidence gradually accumulated to give it some justification. However, it has become clear in the last few years that the concept of homology is underpinned by an amazing degree of conservation of both gene sequence and function (reviewed by Scott, 2000). So fundamental is this to our understanding of the genotype-phenotype relationship and to our interpretation of data from model systems, that it is necessary to deal with the topic at some length.
The existence of Drosophila mutants in which body parts are transformed into recognizable structures but develop at an inappropriate site, the so-called homeotic mutants, has been known since the nineteenth century, when the phenomenon of homeosis was first discovered. Certain unidentified genes were thought to be involved in the specification of the segmented body plan of Drosophila, with mutation resulting in mis-specification of particular body parts. Cloning and sequencing revealed that the homeotic genes are in fact regulatory genes and contain a highly conserved motif, the homeobox (McGinnis et al., 1984), encoding a DNA-binding domain that subsequently became known as the homeodomain. Further analysis of homeobox-containing genes confirmed their role in morphogenetic specification and revealed a complex and hierarchical genetic control of the body plan in this arthropod (reviewed by Akam et al., 1994). The cloning of these genes provided probes with which to screen the genomes of other species, and screening revealed a surprising degree of conservation, with orthologous genes being found in a very wide and diverse range of species examined. The largest and best known of these homeobox-containing gene families are the Hox genes, of which there are 39, organized in four clusters on different chromosomes in all vertebrates including humans.
Sequence homology and position within each cluster is such that derivation of each gene can be traced from a single ancestral cluster similar to the HOM-C complex in Drosophila. Excellent reviews of the organization, evolution and functional roles of Hox genes have been published elsewhere (McGinnis and Krumlauf, 1992; Burke, 2000; Garcia-Fernandez, 2005) and accounts of their role in specification of major features of the vertebrate body plan are given here in Chapters 7, 11, 12 and 15.
Although these genes and others like them have only been identified in vertebrate genomes by virtue of their sequence homology with their Drosophila counterparts (remember that Drosophila probes were used in the screening), conservation of gene sequence is only one aspect of this remarkable evolutionary story. If there is truly homology of function, then we might expect conservation of expression domains of the gene(s) in question across a range of species, and this is indeed often found. The most rigorous test, however, has to be an operational one in which genes are moved into the genome of another and distant species, preferably into individuals in which the orthologue has been inactivated. Will the introduced 'foreign' gene be switched on in the correct spatiotemporal pattern and will it function to produce a normal embryo?
Homeobox-containing genes provide a number of examples in which these three criteria of sequence homology, equivalence of expression domain and functional homology are satisfied. Thus, a regulatory sequence of the Drosophila homeotic gene, Deformed, which supports expression in subregions of posterior head segments, can be replaced by the equivalent mouse sequence and still result in normal embryonic development (Awgulewitsch and Jacobs, 1992). The mouse gene, Hoxb-6, can be moved into the Drosophila embryo and specify normal thoracic segments (Malicki et al., 1990) and even the regulatory element of a human Hox gene, HOXB4, is expressed rostrally and supports head development when introduced into Drosophila (Malicki et al., 1992). Finally, we should not assume that such exchanges only operate between species with segmented body plans, no matter how divergent they may be, since it has also been shown that equivalent functional homology exists between the Hox genes of Drosophila and those of the unsegmented nematode worm, Caenor-habditis elegans (Hunter and Kenyon, 1995).
The existence of such amazing functional homology might suggest that there has been some conservation of downstream target genes for the homeoproteins. But how would this degree of conservation of homeobox gene function across a wide range of species correlate with the diverse range of phenotypic form displayed by these species? In other words, how do we reconcile functional homology, and all that that entails, with the evolution of the disparate body plans displayed by mammals and insects, for example? Such questions are currently unresolved but various possibilities, such as homeoproteins acquiring new targets, homeobox genes changing expression domains, changes in the function of downstream target genes and the emergence of new modes of regulation, are all under consideration (Kenyon, 1994; Manak and Scott, 1994; Hughes and Kaufman, 2002). Meanwhile, similar levels of conservation for genes involved in major morphogenetic events are being discovered, with functional homology apparently being retained by other key regulatory genes and pathways, such as goosecoid, Brachyury and non-canonical Wnt signalling controlling the very different modes of gastrulation across species as diverse as zebrafish, Xenopus, chick and mouse (Beddington and Smith, 1993; De Robertis et al., 1994; Tada et al., 2002).
However, it is not just regulatory genes that display such conservation of sequence, expression domain and function. It is rapidly emerging that genes encoding a number of secreted molecules involved in signalling between cells have been similarly conserved. Genes homologous to the Drosophila hedgehog gene family (so named because of the 'spiny' appearance of the mutant larvae) encode secreted proteins that appear to have a pivotal role in patterning a number of structures in vertebrates (reviewed by Hammerschmidt et al., 1997; Nybakken and Perrimon, 2002). The product of sonic hedgehog (shh) has a major role in notochord induction of the ventral floor plate of the neural tube (e.g. Roelink et al., 1994; and see Chapter 8). A parallel signalling role for this secreted protein is seen in limb development. Thus, shh is expressed in the posterior region of both fin (zebrafish) and limb buds (chick and mouse) where it is thought to be active in establishing pattern across the anteroposterior axis of the bud and is a component of the ZPA (see earlier). Ectopic expression of this gene in the anterior part of the chick limb bud produces duplication of anterior structures, paralleling the mirror-image duplication of the anterior wing compartment in Drosophila resulting from ectopic hedgehog expression (Fietz et al., 1994). Functional homology is even maintained amongst some of the other signalling molecules thought to be downstream from the hedgehog proteins, such as decapentaplegic (dpp) in Drosophila, and the related transforming growth factor-p (TGF-ft) gene family in vertebrates (reviewed by Hogan et al., 1994), and the proteins with which they interact during specification of dorsoventral pattern in the neural primordium (Holley et al., 1995).
Another example of evolutionary conservation of function has recently been demonstrated with the signalling pathway for planar (epithelial) cell polarity. Epithelial cell orientation and cross-talk to the surrounding cells is critical for correct assembly of the Drosophila compound eye and uniform positioning of hairs on the wing and thorax (Strutt, 2003). This is dictated by a secreted Wnt ligand that binds to a frizzled receptor protein complex, which then signals to the nucleus via an intracellular protein called dishevelled. A similar signalling pathway with essentially the same protein components has now been found to organize cellular convergence on the dorsal midline of the mammalian embryo in order to regulate formation of the neural tube. Disruption to any of the core protein components leads to a failure of neural tube closure, as demonstrated in mice, Xenopus and zebrafish (reviewed in Copp et al., 2003; and see Chapter 8). This pathway is also required for correct orientation of the stereociliary bundles found in the mammalian inner ear (Mon-tcouquiol et al., 2003; Curtin et al., 2003). In contrast to neural tube development, this represents a vertebrate phenotype more closely resembling the invertebrate wing hairs.
Similarly conserved function through evolution is elegantly illustrated by the study of mutations in different orthologues of the PAX6 gene (see Chapter 9). Mutations in
PAX6 give rise to the eye defect aniridia, while the mouse orthologue turned out to be a gene formerly known as Small eye, since a loss-of-function mutation produced a microphthalmic phenotype. Both of these genes are the functional orthologues of the Drosophila Eyeless (ey) gene (Quiring et al., 1994); ey is also involved in eye development, and a loss-of-function mutation eliminates the compound eye. As a result, Aniridia, Small eye and Eyeless are collectively regarded as Pax-6 homologues with pivotal roles in eye development, whether it be the compound eye of an arthropod or the vertebrate eye (Quiring et al., 1994). This has been assessed by ectopic expression of the ey gene, which results in ectopic compound eyes with relatively normal facet organization and arrays of photoreceptor cells (Figure 1.4a). More relevant to this discussion is the finding that ectopic expression of the mouse Pax-6/Small eye gene introduced into Drosophila will also generate ectopic compound eyes that are morphologically equivalent to the normal compound eye (Figure 1.4b; Halder et al., 1995). In other words, the generative programme for assembling an arthropod compound eye can be activated and controlled by a mouse Pax-6 gene. It is concluded that these various Pax-6 homologues constitute master genes, arising from a common ancestral gene and with conserved function in controlling eye morphogenesis.
With the advent of more efficient positional cloning strategies, mouse knockout technology and the development of large scale ENU mutagenesis programmes, more and more genes are being assigned to function and phenotype. As a consequence, the extent of regulatory gene involvement in birth defects is becoming better defined. There are now numerous examples of regulatory gene families that are grouped
Figure 1.4 (a) Ectopic compound eye (white arrowhead) formed adjacent to the normally Located compound eye (on the right, black arrowhead) in the head of a Drosophila fly; this is the result of the ectopic expression of the ey gene. (b) Ectopic compound eye formed, in this case, on the leg of a fly, under the control of an ectopically expressed mouse Pax-6 gene introduced experimentally. In both (a) and (b), note the similarity of the ommatidial organization and interommatidial bristles, in the ectopic eyes and in their normal counterpart in (a). Photographs supplied by Professor Walter Gehring
Figure 1.4 (a) Ectopic compound eye (white arrowhead) formed adjacent to the normally Located compound eye (on the right, black arrowhead) in the head of a Drosophila fly; this is the result of the ectopic expression of the ey gene. (b) Ectopic compound eye formed, in this case, on the leg of a fly, under the control of an ectopically expressed mouse Pax-6 gene introduced experimentally. In both (a) and (b), note the similarity of the ommatidial organization and interommatidial bristles, in the ectopic eyes and in their normal counterpart in (a). Photographs supplied by Professor Walter Gehring through both sequence and functional homology that are also directly implicated in dysmorphogenesis and neoplasia. These include the PAX, HOX, ZIC, ZNF, SOX, FOX and TBX families. When the first T-box gene, T (Brachyury), was identified (Herrmann et al., 1990) it was thought to be unique. However, most species studied have multiple family members and mammals contain a total of 17 different functional T-like genes (see Table 1.1). Family members are based on their
Was this article helpful?