Developmental biology

Hox gene variation and evolution

Article metrics

Both developmental and evolutionary biologists try to explain patterns in the diversity among organisms, and the Hox genes encode a class of transcription factors that have provided ample material for such discussions. Because they may be pivotal in specifying regional identity in body plans, differences in their expression could (at least partly) explain the evolution of animal phyla. The Hox genes are arranged in genomic clusters and, importantly, they are expressed in a spatially colinear fashion — anterior genes are expressed early in development and towards the front of the body, posterior genes later in development and in more distal portions of the body.

In invertebrates, only a single Hox gene cluster has been found (although it is split in Drosophila). The common ancestor of all chordates is surmised to have had a single cluster as well. This cluster is thought to have duplicated to four clusters (A-D) on different chromosomes, accompanying the increasing complexity of body plans during the evolution of vertebrates (Fig. 1). But a report by Prince et al.1, shortly to appear in Development, is likely to cause some questioning of this commonly held hypothesis — that genomic and morphological complexity are causally linked2,3,4,5,6,7.

Figure 1: Evolutionary conservation of the Hox gene clusters in zebrafish, pufferfish, mouse and Amphioxus.

The zebrafish Hox gene clusters are shown with the 34 genes confirmed by Prince et al.1, and an additional eight, for the total of 42 that the zebrafish is likely to have. Ten gene losses differentiate the mouse from the pufferfish, but there are only three losses between the mouse and zebrafish, even though fish and tetrapods shared their last common ancestor about 400 million years ago. The common ancestor of human and mouse probably lived around 60 million years ago, but the architectures of their Hox gene clusters are identical. The losses in the AbdA (abdominal A)-related genes (light green) in the D-cluster probably occurred early in the history of vertebrates. (Figure drawn by Heike Haunstetter.)

Using an experimental approach based on the polymerase chain reaction, Prince et al. unambiguously identified 34 Hox genes and determined their linkage with somatic-cell hybrids. Surprisingly, they found that the zebrafish has three Hox genes (HoxC3, HoxA8 and HoxB10), with no direct mouse equivalents (Fig. 1). Moreover, the expression domains of the anterior Hox genes are partly overlapping, and restricted to a shorter anterior region. Possibly the most important finding is that the zebrafish has at least two additional Hox gene clusters for a total of six and not, as previously thought8, the typical vertebrate number of four. All of the genes on these additional clusters have probably not yet been discovered, and three are reported so far1.

These two additional clusters lead us to question a simple, ‘more clusters, more complexity’ model of evolutionary diversification — in terms of phenotypic complexity, however measured, a zebrafish is probably not 50 per cent more complex than a mouse or a human. The extra clusters cannot be explained by entire-genome duplications because, although polyploidy is known from other carp-like fish9 and is common in salmonids, zebrafish are diploid. It also seems unlikely that the additional clusters are remnants of a polyploid ancestral condition. Instead, the Hox-cluster duplications in zebrafish might be a unique evolutionary event. But such events may turn out to be common, at least in fish.

Prince and colleagues' work on zebrafish1, combined with studies on the pufferfish10, now enables us to reconstruct the evolutionary history of the Hox gene clusters in vertebrates (Fig. 1). The initial chordate ancestral cluster of 13 Hox genes (the architecture that is still present in the cephalochordate Amphioxus7) probably duplicated in a three-step process, to form four complete clusters with a total of 52 genes. One phylogenetic study11 indicates that the D-cluster is the most ancestral, and that the B- and C-clusters are the youngest. Hagfish and lamprey are phylogenetic intermediates between Amphioxus and more derived vertebrates such as zebrafish. So, if these fishes have only two or three clusters (which is not precisely known), they would be more likely to contain a D-like Hox gene cluster than a B- or C-cluster. The suggestion that the D- and A-clusters are the oldest also seems to fit the observation that the D-cluster is the most ‘deteriorated’ of all (Fig. 1).

Following the principle of Dollo parsimony — which assumes that losses of genes are much more common and likely than independent evolutionary origins12 — we can speculate which Hox genes might have been present in the common ancestors of vertebrates, tetrapods and fish (Fig. 1). Based on the genomic organizations available so far, the rates of evolution of Hox clusters do not seem to be constant. For example, whereas the zebrafish is likely to have lost only one Hox gene since it shared a common ancestor with the pufferfish (probably more than 200 million years ago), the losses along the pufferfish lineage were possibly several times faster (12 Hox genes were lost, if the zebrafish really has 42 Hox genes) (Fig. 1). Ignoring the additional clusters of the zebrafish, and estimating that it has 42 Hox genes, we find that 13 differences separate the zebrafish from the pufferfish. In the pufferfish, the HoxC1 and C3 genes are still recognizable, but they are only pseudogenes (genes that are not transcribed). They might have lost their function at different points in the evolution of fish, because HoxC3, at least, is still present in the zebrafish.

The apparent acceleration of genomic evolution along the pufferfish lineage might be correlated with accelerated morphological evolution. Pufferfish belong to one of the most morphologically derived groups of fish, and they lack ribs, pelvic fins and the pelvic girdle. Are the missing genes those that are no longer necessary because these structures have been lost during evolution? If so, the missing Hox genes in the pufferfish might also be absent in the other groups of fish which have secondary loss of pelvic fins (such as eels) or even tail fins (for example, the ocean sunfish Mola mola). Moreover, the loss of Hox genes might also be accompanied by the secondary loss (or simplification) of appendages in land vertebrates, such as in limbless amphibians, reptiles and whales.

What selective forces maintain or modify genomic organizations? The observation that Hox genes are clustered, and that the architecture of these clusters is highly conserved in evolution, has led to the suggestion that the regulatory elements that control expression of the Hox genes cannot be separated from these genes without jeopardizing their proper functioning and, possibly, determination of morphology along the antero-posterior axis. For vertebrates13,14 these ideas have been partially confirmed experimentally15, and this tight functional linkage might be particularly strong along the lineage that leads to reptiles and mammals (Fig. 1).

The long-standing question of whether the evolution of genes or networks of interactions through regulatory elements drives most morphological diversification might, then, have different answers in different evolutionary lineages. In the most species-rich group of vertebrates — fish — organization of the Hox genes might not be completely constrained by interwoven regulatory networks, and differentiation might be driven by gene evolution. However, in the lineage that leads to reptiles and mammals, the driving force behind morphological diversification might have been newly evolving interactions in networks of regulatory elements16.


  1. 1

    Prince, V. E., Jolly, J., Ekker, M. & Ho, R. Development 125, 407–420 (1998).

  2. 2

    Akam, M. Phil. Trans. R. Soc. 349, 313–319 (1995).

  3. 3

    Holland, P. W. H. & Garcia-Fernández, J. Dev. Biol. 173, 382–395 (1996).

  4. 4

    Lewis, E. B. Nature 276, 565–570 (1978).

  5. 5

    Krumlauf, R. Cell 78, 1991–202 (1994).

  6. 6

    Duboule, D. & Morata, G. Trends Genet. 10, 358–364 (1994).

  7. 7

    Garcia-Fernández, J. & Holland, P. W. H. Nature 370, 563–566 (1994).

  8. 8

    Misof, B. Y. & Wagner, G. P. Mol. Phylogen. Evol. 5, 309–322 (1995).

  9. 9

    Buth, D. G., Dowling, T. E. & Gold, J. R. in Cyprinid Fish, Systematics, Biology and Exploitation (eds Winfield, I. J. & Nelson, J. S.) 83-126 (Chapman & Hall, London, 1991).

  10. 10

    Aparicio, al. Nature Genet. 16, 79–83 (1997).

  11. 11

    Bailey, W. J., Kim, J., Wagner, G. P. & Ruddle, F. H. Mol. Biol. Evol. 14, 843–853 (1997).

  12. 12

    Meyer, A. in New Uses for New Phylogenies (eds Harvey, P. H., Leigh Brown, A. J., Maynard Smith, J. & Nee, S.) 322-340 (Oxford Univ. Press, 1996).

  13. 13

    Gérard, al. Genes Dev. 10, 2326–2334 (1996).

  14. 14

    Gould, A. et al. Genes Dev.11, 900-913 (1997).

  15. 15

    Mann, R. S. BioEssays 19, 661–664 (1997).

  16. 16

    Davidson, E. H., Peterson, K. J. & Cameron, R. A. Science 270, 1319–1325 (1995).

Download references

Author information

Rights and permissions

Reprints and Permissions

About this article

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.