They might not beat insects in number of species but, compared with all other animals that roam the Earth, vertebrates have certainly hit a winning formula. More sophisticated brains had some say in this, although a lesser known morphological development — the invention of upper and lower jaws — almost certainly contributed to the success of this phylum. It might be some time before we're privy to the genetic basis of brain complexity; however, one paper now provides a developmental genetic model for the evolution of jaws. The explanation — as gleaned from knockout mice — lies in the peculiar expression pattern of a family of Distal-less homeobox (Dlx) factors along the proximal–distal axis of jaw precursor tissues.

Mice, like all mammals, have six Dlx genes (1–3 and 5–7), which are tandemly arranged in the genome. The genes are expressed in several head regions but of importance to jaw development is their expression in the branchial arches (BAs) — segmentally repeated structures with proximal–distal polarity. In the first arch, the more proximal (maxillary) portion develops into the upper jaw and the more distal (mandibular) portion into the lower jaw. The Dlx genes are expressed in nested pairs along this axis — Dlx1 / 2 are present along most of the axis, with Dlx5 / 6 and Dlx3 / 7 showing progressively more restricted distal expression. In previous work, mice in which either Dlx1 or Dlx2, or both, were knocked out had only proximal jaw defects, which was attributed to the rescue of their distal functions by other genes in the family, notably Dlx5. The authors therefore asked what would happen if both Dlx5 and Dlx6 were knocked out — the prediction being that distal BAs would acquire proximal properties. Although Dlx5/6−/− mice die at birth, their embryonic phenotype was even more striking than expected. Analysis of morphology and gene expression shows that the loss of these two genes leads to the homeotic, mirror image transformation of the lower jaw into the upper jaw.

Although vertebrates have had jaws for over 400 million years, jawless vertebrates, such as lampreys, are still around. Curiously, Dlx genes are expressed in the lamprey BAs, but the pattern is not nested. This gives weight to the theory that the transition from jawless to jawed vertebrates depends on the particular Dlx distribution pattern. As the authors point out, the cell fates imparted by Dlx genes probably depend on interpreting the one or more positional signals that act upstream of them. Candidates for these molecules exist (fibroblast growth factor 8, for example), but just how many convergent signals it takes to make a vertebrate jaw remains to be seen.