Hox gene clusters notate relative positions along the A/P axis of the endoderm and mesoderm, and are usually expressed from a particular anterior position all the way to the posterior.
Hox gene clusters specify relative position, not specific structures. Hox gene expression oft coincides with visibly distinct boundaries between different parts of the intestinal tract. It is unknown how Hox gene expression is restricted to specific regions along the A/P axis of the endodermal tube. Two known inducing signals are Shh and ParaHox.
Only three human birth defects have been linked to Hox mutations.
HOXD13-/- defects the digits; HOXA13-/- defects the digits and genitals; and HOXA9-/- defects the forearm. Generally, mice with specific HOX genes knocked out have intestinal malformations (not transformations): Hox3-/- defects the pharyngeal pouch, leading to a retarded thyroid, parathyroid and thymus; Hox5-/- retards the trachea and lungs, it is expressed in mesoderm proximal to the tracheal bud; and Hox13-/- defects the anal sphincter and cloaca.
Retinoids (like Accutane) are well known to have teratogenic effects.
Retinoic acid can induce Hox gene expression such that anterior genes are expressed earlier. Furthermore, cells are exposed to varying concentrations of retinoic acid (produced by somites) over time. Perhaps an endogenous retinoic acid gradient controls sequential expression of Hox genes. However, it is unknown what controls retinoic acid expression in somites.
Hox Gene Control
It remains unclear how Hox genes are expressed in precise patterns that are related to their location within the complex. A clue has come from the discovery of an enhancer element located outside of the HoxD cluster. This enhancer acts as a global control region that controls the timing and pattern of expression of many genes in this region of the chromosome.
It is likely that similar global control regions will be found for the other Hox complexes.
HOX and Homeotic genes are homologs.
A screen of vertebrate genomic DNA libraries with Drosophila homeobox DNA identified vertebrate gene clusters with homology to Drosophila Hom-C. Verebrate Hox genes and Drosophila AntpC and BxC are both temporally collinear along the A/P axis and share a similar homeodomain sequence. Perhaps a primordial Hox complex existed before arthropods and chordates diverged, and that Hox and Hom-C have similar roles in development. However, insects have only one Hox cluster while vertebrates have four.
Amazingly, Hox genes are conserved enough that the human HOX 4.2 gene can partially substitute for the Drosophila Deformed gene (giving weak antenna- to-leg transformation when expressed ectopically). Similarly, ectopic expression of either the Drosophila Antennapedia gene or the mouse Hoxb-6 gene in Drosophila causes transformation of antenna into a leg. This gives credence to the notion of a primordial Hox gene cluster.
In Drosophila, naturally occurring mutations in Hox genes have been noted and studied.
Mouse Models: First, the gene must be knocked out of an embryonic stem (ES) cell line, followed by introduction of the mutated cel into a blastocyst and breeding of the resultant chimeric mice. Defects are usually observed in the anterior part of the Hox gene's expression domain. This is a phenomenon knows as posterior prevalence, whereby when a gene loses function then the most anterior segment of the domain where the gene is usually expressed takes on the identify of the next most anterior segment. Sometimes, these defects are homeotic (ie, rib-less lumbar vertebrae become ribbed thoracic vertebrae).
ParaHox is a complex of the three homeodomain genes that, like the Hox genes, are ordered along the chromosome in the same pattern as their sequential expression along the antero-posterior axis. From posterior to anterior, the three ParaHox genes are: Caudal in Drosophila and Cdx in vertebrates, important for posterior gut development and expressed in a gradient opposite to bicoid; Pdx in vertebrates, important for determining pancreatic fate; and Nkx in vertebrates, important for lung and thyroid morphogenesis and expressed in ventral foregut endoderm.
Abnormalities in vertebrate (including human) development have been observed that can be interpreted as possible segmental homeotic transformations. These include: i.) the occurrence of ribs on cervical (neck) vertebrae. ii.) a mutation in mice (called rachiterata) that results in replacement of the 7th cervical vertebra with a 1st thoracic vertebra with a rib. iii.) the occurrence of rare ground sloths with thoracic-like vertebrae in the lumbar region (Raff and Kaufman, 1983).
5. MAKING MUTATIONS IN VERTEBRATE HOX GENES Knockout mutations in various mouse Hox genes have been generated in order to understand the requirement for these genes in embryonic development. Mutants have been obtained in all of the Hox genes. Defects are observed in embryos in the anteriormost domain of the expression of the mutated Hox gene. In some cases, these defects are homeotic transformations (i.e., knockout of the Hoxc-8 gene causes several lumbar vertebrae (which do not have ribs) to become transformed into thoracic vertebrae (which do have ribs)). These types of results have been summarized by the term "posterior prevalence", which means that the defects are seen in the domain where the mutated gene is the most posteriorly expressed Hox gene. These homeotic transformations are analogous to the homeotic transformation described above for Ubx: in the absence of a particular Hox gene product, the most anterior segment where that gene is expressed becomes transformed into the identity of the next more anterior segment. The mechanisms underlying posterior prevalence are not understood. Posterior prevalence is seen in flies and mice, so it appears to be a conserved feature of Hox gene regulation. It has been known that while Hox gene mRNAs are found in overlapping patterns along the A/P axis, the protein products show a more restricted distribution. For example, the Ubx mRNA is found in a broad domain along the A/P axis of the Drosophila larva, but the protein is much more restricted to the A end of the axis. A recent clue to one possible mechanism underlying posterior prevalence is the discovery of microRNA (miRNA) genes within Hox clusters. Recall that miRNAs lead to inhibition of translation or degradation of the mRNAs of their target genes. Recent estimates suggest that up to 4% of the vertebrate genome encodes miRNAs, and these target up to 1/3 of all genes. We are only beginning to understand miRNA function. Within the Hox clusters, multiple miRNAs are found in conserved locations, and remarkably, their primary targets appear to be the Hox genes. Even more remarkably, they target genes that are expressed more anteriorly along the Hox cluster.