Fibroblast Growth Factor Signalling: From Development to Cancer

Nicholas Turner; Richard Grose

In This Article

FGF Signalling

FGF signalling has evolved to become a highly complex growth factor signalling pathway, reflecting the multitude of physiological functions that are controlled by FGF signalling. The mammalian FGF family comprises 18 ligands ( Box 1 ), which exert their actions through 4 highly conserved transmembrane tyrosine kinase receptors (FGFR 1, FGFR2, FGFR3 and FGFR4) (Fig. 1). A fifth related receptor, FGFR5 (also known as FGFRL1), can bind FGFs, but has no tyrosine kinase domain, and might negatively regulate signalling[3] (Fig. 2).

Figure 1.

FGFR structure and control of ligand specificity.a | The basic structure of the fibroblast growth factor (FGF) –FGF receptor (FGFR) complex comprises two receptor molecules, two FGFs and one heparan sulphate proteoglycan (HSPG) chain. The FGF signalling pathway comprises 4 highly conserved transmembrane receptors and 18 FGF ligands ( Box 1 ). FGFs bind with low affinity to cell surface HSPGs (purple) and with high affinity to specific FGFRs. The FGFRs, which are phylogenetically closely related to the vascular endothelial growth factor receptors (VEGFRs) and platelet-derived growth factor receptors (PDGFRs), consist of three extracellular immunoglobulin (Ig) domains, a single transmembrane helix and an intracellular split tyrosine kinase (TK) domain. The second and third Ig domains form the ligand-binding pocket and have distinct domains that bind both FGFs and HSPGs. b | Ligand-binding specificity is generated by alternative splicing of the Ig III domain. The first half of Ig III is encoded by an invariant exon (IIIa), which is spliced to either exon IIIb or IIIc, both of which splice to the exon that encodes the transmembrane (TM) region. Epithelial tissues predominantly express the IIIb isoform and mesenchymal tissues express IIIc. FGFR4 is expressed as a single isoform that is paralogous to FGFR-IIIc. c | Examples of the extent to which ligand specificity can differ between FGFR-IIIb and FGFR-IIIc isoforms, illustrated with the differing ligand specificty of FGFR2 isoforms. The FGFR2-IIIb ligands are shown in blue and the FGFR2-IIIc ligands are shown in brown. For example, FGF7 and FGF10 bind specifically to FGFR2-IIIb and have essentially no binding to FGFR2-IIIc.[7] The mechanisms controlling splice isoform choice are becoming clearer and defined control elements have been identified in the introns surrounding alternatively spliced exons.[177–179]

Figure 2.

FGFR signalling network. The signal transduction network downstream of fibroblast growth factor (FGF) receptors (FGFRs), along with negative regulators. Following ligand binding and receptor dimerization, the kinase domains transphosphorylate each other, leading to the docking of adaptor proteins and the activation of four key downstream pathways: RAS–RAF–MAPK, PI3K–AKT, signal transducer and activator of transcription (STAT) and phospholipase Cγ (PLCγ) (green). FGFRs have also been shown to bind and directly phosphorylate ribosomal S6 kinase[17] (not shown). Signalling can be negatively regulated at several levels by receptor internalization or the induction of negative regulators, including FGFR-like 1 (FGFRL1), SEF, Sprouty (SPRY), CBL, MAPK phosphatase 1 (MKP1) and MKP3 (brown). These regulators may modulate ligand binding (FGFRL1 and SEF) or interfere with intracellular signalling, principally through modulation of the MAPK pathway. DAG, diacylglycerol; FRS2α, FGFR substrate 2α; GRB2, growth factor receptor-bound 2; IP3, inositol triphosphate; P, phosphorylation; PIP2, phosphatidylinositol-4,5-biphosphate; PKC, protein kinase C; Sos, son of sevenless.

FGFs are secreted glycoproteins that are generally readily sequestered to the extracellular matrix, as well as the cell surface, by heparan sulphate proteoglycans (HPSGs). To signal, FGFs are released from the extracellular matrix by heparinases, proteases or specific FGF-binding proteins, and the liberated FGFs subsequently bind to cell surface HPSGs (reviewed in Ref. [4]). Cell surface HPSGs also stabilize the FGF ligand–receptor interaction, forming a ternary complex with FGFR[5,6] (Fig. 1). The specificity of the FGF–FGFR interaction is established partly by the differing ligand-binding capacities of the receptor paralogues,[7,8] but also by alternative splicing of FGFR, which substantially alters ligand specificity (Fig. 1). Further control of FGF–FGFR specificity is mediated by the tissue-specific expression of particular ligands and receptors, coupled with several cell surface or secreted proteins that facilitate the FGF–FGFR interaction,[9] such as the Klotho family[10] for hormonal FGFs ( Box 1 ), which further increases ligand specificity.

Downstream Signalling

FGF receptors signal as dimers, and ligand-dependent dimerization leads to a conformational shift in receptor structure that activates the intracellular kinase domain, resulting in intermolecular transphosphorylation of the tyrosine kinase domains and intracellular tail. Phosphorylated tyrosine residues on the receptor function as docking sites for adaptor proteins, which themselves may also be directly phosphorylated by FGFR,[11] leading to the activation of multiple signal transduction pathways (Fig. 2). FGFR substrate 2 (FRS2) is a key adaptor protein that is largely specific to FGFRs, although it can also bind other tyrosine kinase receptors, such as neurotrophic tyrosine kinase receptor type 1 (NTRK1), RET and anaplastic lymphoma kinase (ALK).[12] FRS2 binds to the juxtamembrane region of FGFRs through its phosphotyrosine-binding (PTB) domains. The activated FGFR phosphorylates FRS2 on several sites, allowing the recruitment of the adaptor proteins son of sevenless (SOS) and growth factor receptor-bound 2 (GRB2) to activate RAS and the downstream RAF and MAPK pathways.[11] A separate complex involving GRB2-associated binding protein 1 (GAB1) recruits a complex, which includes PI3K, and this activates an AKT-dependent anti-apoptotic pathway.[13]

Elsewhere on the intracellular portion of the activated receptor, and independently of FRS2 binding, the Src homology 2 (SH2) domain of phospholipase Cγ (PLCγ) binds to a phosphotyrosine residue towards the carboxyl terminus[14] (Fig. 2). After PLCγ is activated, it hydrolyses phosphatidylinositol-4,5-biphosphate (PIP2) to phosphatidylinositol-3,4,5-triphosphate (PIP3) and diacylglycerol (DAG),[15] activating protein kinase C (PKC), which partly reinforces the activation of the MAPK pathway by phosphorylating RAF. Several other pathways are also activated by FGFRs, depending on the cellular context, including the p38 MAPK and Jun N-terminal kinase pathways, signal transducer and activator of transcription (STAT) signalling[16] and ribosomal protein S6 kinase 2 (RSK2).[17]

Negative Regulation of Signalling

The mechanisms of attenuation and negative feedback control of FGFR signalling are only partly understood. Following activation receptors are internalized, resulting in receptor degradation or recycling, in a process that is partly controlled by CBL-mediated monoubiquitylation[18] (Fig. 2). MAPK signalling, particularly ERK1 and ERK2 signalling, has been shown to phosphorylate FRS2 on many serine and/or threonine residues, inhibiting the recruitment of GRB2 (Ref.[12]). Downstream signalling can be attenuated through the induction of MAPK phosphatases such as MAPK phosphatase 3 (MKP3),[19] Sprouty (SPRY) proteins[20,21] and SEF[22,23] family members that modulate receptor signalling at several points in the signal transduction cascade[24] (Fig. 2). MKP3 dephosphorylates ERK1 and ERK2 to attenuate MAPK signalling,[19] and SPRY proteins are thought to function in either a dominant-negative fashion, by competing for GRB2 binding and so preventing SOS-mediated RAS activation, or by directly binding to RAF and blocking subsequent MAPK signalling.[20,21] Similarly, SEF may function at multiple levels and has a transmembrane form that can directly interact with FGFRs; both the transmembrane form and a splice variant that is confined to the cytoplasm seem to be capable of inhibiting ERK phosphorylation.[25]

Context-dependent Signalling

Studies of FGF signalling during development reinforce the crucial importance of cellular context in determining the functional outcome of FGFR activation ( Box 2 ). Although in most cellular contexts FGFs induce proliferation and migration,[2] in certain cell types physiological FGF signalling induces differentiation and/or cell cycle arrest. This has been most clearly demonstrated in the development of endochondral and membranous bone, in which mouse models have shown that FGFR3 and FGFR2 can negatively regulate proliferation and positively drive differentiation.[26,27] In humans, several gain-of-function germline mutations in the FGFRgenes result in skeletal dysplasias, with FGFR2 mutations a common cause of craniosynostosis syndromes and FGFR3 mutations common in chondrodysplasia syndromes.[28] Similarly, activating germline mutations in FGFR1 are a cause of Pfeiffer syndrome, a rare craniosynostosis syndrome.[29] Through several different mechanisms, these germline mutations result in ligand-independent activation of the receptor, which induces the premature fusion of cranial sutures[28] and the differentiation of chondrocytes in the endplates of long bones.[28] There is a lack of evidence regarding whether germline FGFR mutations predispose to cancer.

The predominant signalling pathway activated downstream of FGFRs in development seems to be MAPK signalling,[30] and although it is incompletely understood current evidence points to the importance of differential effects of MAPK signalling in determining the cellular response of FGFR activation. The downstream effect of MAPK signalling is mostly cell proliferation; for example, through the induction of cyclin D1 expression, thereby promoting entry into the S phase of the cell cycle.[31,32] However, MAPK signalling induces cell cycle arrest and differentiation in other contexts, such as the p21-dependent cell cycle arrest in chondrocytes[33] and the differentiation of neuronal PC-12 cells.[34] Similarly, divergent responses can occur in cancer development (discussed below).

There are many factors that underlie context-dependent signalling, including the cell type-specific expression of different adaptor molecules, signal transduction enhancers, transcription factors and co-activators.[32] Other factors that are important in context specificity include crosstalk with other signalling networks, such as Wnt signalling,[35] and importantly the kinetics of signalling, with sustained strong signalling inducing differentiation, senescence and apoptosis, and less pronounced signalling inducing proliferation and survival.[36]

A further factor that affects the context specificity of FGF signalling might be differences in signalling between the FGFRs. Although all four FGFRs signal through a similar network of pathways in general, the kinase domain of FGFR1 drives stronger downstream pathway activation than FGFR4 (Ref.[37]). There is also some evidence that differential responses to signalling are initiated by the FGFR1 and FGFR2 kinase domains, with more rapid attenuation of FGFR2 signalling mediated by receptor internalization and degradation.[38–40] Although splicing of the extracellular domain controls ligand specificity, there is no evidence to suggest that this affects intracellular signalling per se.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: