Signaling by FGFR3 in disease

Activating mutations in FGFR3 are associated with the development of a range of skeletal dysplasias that result in dwarfism (reviewed in Webster and Donoghue, 1997; Burke et al, 1998; Harada et al, 2009). The most common form of human dwarfism is achondroplasia (ACH), which is caused by mutations G380R and G375C in the transmembrane domain of FGFR3 that are thought to promote ligand-independent dimerization Hypochondroplasia (HCH) is a milder form dwarfism that is the result of mutations in the tyrosine kinase domain of FGFR3. Two neonatal lethal conditions, thanatophoric dysplasia type I and II (TDI and TDII) are also the result of mutations in FGFR3; TDI arises from a range of mutations that either result in the formation of unpaired cysteine residues in the extracellular region that promote aberrant ligand-independent dimerization or by mutations that introduce stop codons (Rousseau et al, 1995; Rousseau et al, 1996, D'Avis et al,1998). A single mutation, K650E in the second tyrosine kinase domain of FGFR3 is responsible for all identified cases of TDII (Tavormina et al, 1995a, b). Other missense mutations at the same K650 residue give rise to Severe Achondroplasia with Developmental Disorders and Acanthosis Nigricans (SADDAN) syndrome. The severity of the phenotype arising from many of the activating FGFR3 mutations has recently been shown to correlate with the extent to which the mutations activate the receptor

In addition to mutations that cause dwarfism syndromes, a Pro250Arg mutation in the conserved dipeptide between the IgII and IgIII domains has been identified in an atypical craniosynostosis condition. This mutation, which is paralogous to mutations seen in FGFR1 and 2 in Pfeiffer and Apert Syndrome, respectively, results in an increase in ligand-binding affinity for the receptor.

Of all the FGF receptors, FGFR3 has perhaps the best established link to the development in cancer. 50% of bladder cancers have somatic mutations in the coding sequence of FGFR3; of these, more than half occur in the extracellular region at a single position (S249C) (Cappellen et al, 1999; Naski et al, 1996; di Martino et al, 2009, Sibley et al, 2001). Activating mutations are also seen in the juxta- and trans-membrane domains, as well as in the kinase domain (reviewed in Weshe et al, 2011). As is the case for the other receptors, many of the activating mutations that are seen in FGFR3-related cancers mimic the germline FGFR3 mutations that give rise to autosomal skeletal disorders and include both ligand-dependent and independent mechanisms (reviewed in Webster and Donoghue, 1997; Burke et al, 1998). In addition to activating mutations, the FGFR3 gene is subject to a translocation event in 15% of multiple myelomas. This chromosomal rearrangement puts the FGFR3 gene under the control of the highly active IGH promoter and promotes overexpression and constitutive activation of FGFR3. In a small proportion of multiple myelomas, the translocation event is accompanied by activating mutations in the FGFR3 coding sequence.

More recently, a number of fusion proteins of FGFR3 have been identified in various cancers (Singh et al, 2012; Williams et al, 2013; Parker et al, 2013; Wu et al, 2013; Wang et al, 2014; Yuan et al, 2014; reviewed in Parker et al, 2014). The most common fusion protein is TACC3, a coiled coil protein involved in mitotic spindle assembly. FGFR3 fusion proteins are constitutively active and appear to contribute to proliferation and tumorigenesis through activation of the ERK and AKT signaling pathways (reviewed in Parker et al, 2014).