Normal development depends crucially on precise communication between embryonic tissues, which in turn depends on finely tuned gene expression. But how do cells discriminate between the signals of a large family of ligands that are often coexpressed? In such cases, signalling specificity can be imposed by tightly regulating the expression of receptors with different affinities. Tissue-specific alternative splicing can achieve this — and is the focus of a recent paper in which the inactivation of one isoform of mouse Fgfr2 (fibroblast growth factor receptor 2) had disastrous developmental consequences, producing defects similar to the human Apert and Pfeiffer syndromes. These disorders are characterized by craniosynostosis (the premature closure of cranial sutures) and short stature.

Four Fgf receptors transduce signals from ∼22 Fgf ligands during mammalian development, requiring that receptor affinity and specificity be tightly regulated. This is brought about by the alternative splicing of exons that encode the extracellular domains (loops I–III) of each Fgfr. This splicing generates different receptor isoforms (called a, b or c), each with distinct ligand-binding affinities. For Fgfr2, this splicing is highly tissue specific — in epithelial tissue, exons 7, 8 and 10 are used to generate the Fgfr2IIIb isoform, and in mesenchymal tissue, exons 7, 9 and 10 create the Fgfr2IIIc isoform. Reciprocal signalling occurs because epithelially expressed Fgfs activate only the mesenchymally spliced Fgfr2IIIc, and mesenchymally expressed Fgfs activate only epithelial Fgfr2IIIb.

Patients with Apert syndrome commonly have FGFR2 mutations in these extracellular regions generated by alternative splicing — mutations that might, according to previous work, create a mutant form of FGFR2IIIc that aberrantly responds to mesenchymally expressed FGFs. However, Hajihosseini et al. have now found that abrogating the expression of the Fgfr2IIIc isoform in mice, by deleting exon 9 from mouse Fgfr2, results in a dominant gain of function that causes premature ossification of the cranial sutures and intersternebral cartilage, and defects in organs that undergo branching morphogenesis. This mutant phenotype recapitulates many features of the Apert and Pfeiffer syndromes.

How can loss of Fgfr2IIIc result in the same defects that are caused by mutations that alter the activity of FGFR2IIIc? Hajihosseini et al. found that deletion of exon 9 leads to the upregulated expression of the Fgfr2IIIb isoform in cells that normally express Fgfr2IIIc, which causes those cells to respond to the wrong Fgfs. The mesenchymal expression of Fgfr2IIIb disrupts the epithelial–mesenchymal reciprocal signalling required for branching morphogen-esis. And because Fgfr2IIIb signalling probably also mediates normal endochondrial bone formation, its ectopic activation might underlie the abnormal and excessive bone ossification in these mutant mice. Future work should pinpoint whether other Fgf-related developmental disorders are caused by similar ligand-independent, receptor-activation mechanisms.