Promotion and Attenuation of FGF Signaling Through the Ras-MAPK Pathway

See allHide authors and affiliations

Science's STKE  13 Apr 2004:
Vol. 2004, Issue 228, pp. pe17
DOI: 10.1126/stke.2282004pe17


The fibroblast growth factors (FGFs) represent a large family of ligands that activate signal transduction pathways leading to diverse biological responses, including many involved in various processes during development. Here, we discuss the discovery of a subset of conserved FGF target genes that encode feedback regulators of FGF signaling itself. Members of the Sprouty, Sef, and mitogen-activated protein kinase phosphatase families are negative modulators of FGF signaling, whereas positive factors that promote FGF signaling include the ETS transcription factors ERM and PEA3 and the transmembrane protein XFLRT3. These molecules affect the FGF signaling cascade at different levels to regulate the final output of the pathway. This multilayered regulation suggests that precise adjustment of FGF signaling is critical in development.

The process of embryogenesis requires cell-to-cell communication mediated by secreted factors and involves the complex interactions of multiple signaling pathways. The focus of this Perspective is to highlight some recent findings on the role of feedback modulators pertaining to the fibroblast growth factor (FGF) family of ligands. FGFs represent a large family of secreted molecules [22 genes in humans, 18 of which are true secreted FGF ligands (1, 2)]. Upon binding to their cognate receptors (FGFRs), FGFs activate signal transduction pathways required for multiple developmental processes, including the specification of cell fates, determination of axial polarity, and promotion of cell survival (3, 4). The FGFRs are members of the receptor tyrosine kinase (RTK) class of transmembrane proteins and activate several signaling cascades, including the phospholipase C gamma (PLC-γ), phosphatidylinositol-3 kinase (PI3K), and Ras to extracellular signal–regulated protein kinase (ERK) pathways; ERKs are a subclass of mitogen-activated protein kinases (MAPKs) (3) (Fig. 1). The wide-ranging biological roles of FGFs and the multitude of signaling pathways activated by this family of ligands suggest that FGF signaling must be tightly regulated, and both positive and negative feedback loops have been discovered in this pathway.

Fig. 1.

FGF signaling pathways. Ligand binding initiates receptor dimerization and activation of kinase activity, promoting binding of the adaptor molecule FRS2 and the subsequent recruitment of Grb2 and SOS to the FGFR complex. SOS facilitates guanine nucleotide exchange to activate Ras, which stimulates the Raf-to-MEK-to-ERK pathway. The PLC-γ and PI3K pathways are not discussed here. Phosphorylated forms of ERKs activate transcription factors, including members of the ETS family. These helix-turn-helix factors bind DNA as monomers, but require the formation of ternary complexes to initiate transcription. Certain transcriptional targets of FGF activity are feedback modulators of the FGF signaling cascade, as illustrated in Figs. 2 and 3.

The first bona fide feedback regulator of the FGF pathway, Sprouty (Spry), was discovered through a genetic screen in Drosophila (5). Airway branch formation is governed by FGF activity, and spry mutants exhibit ectopic airway branches due to excessive FGF signaling. Spry proteins, which contain an invariant tyrosine phosphorylation site and cysteine-rich domain at the C terminus, are widely conserved, with four members having been identified in mammalian species (6, 7). Overexpression and gene depletion studies in zebrafish and mouse confirm that vertebrate Sprys are functionally conserved and antagonize FGF signaling (810). The biochemical mechanism of Spry function has been under intense investigation, and whereas Spry is thought to be a general inhibitor of Ras-MAPK signaling, it behaves differently depending on the RTK that is activated. In fact, under certain conditions, Spry has been shown to prolong epidermal growth factor (EGF) to Ras-MAPK signaling, and this is dependent on the presence of Cbl, an E3 ubiquitin ligase and adaptor protein that interacts with Sprys (11, 12). In this instance, Spry prevents Cbl-mediated EGFR ubiquitination and degradation (11, 12). It is clear that in regard to FGF signaling, Spry acts as an antagonist; however, the requirement for Cbl in this context is uncertain (12, 13). A proposed mechanism is that after FGF signaling, Spry is tyrosine-phosphorylated, creating a decoy site that binds the docking molecule Grb2 and prevents SOS from activating Ras (Fig. 2) (14). Another mechanism involves the cysteine-rich domain, which contains a Raf-binding motif; interaction of Spry and Raf interferes with the activation of the MAPK pathway downstream of Raf (Fig. 2) (15, 16). It is likely that both mechanisms act during development, where the former is posttranslationally controlled to allow fine-tuning and the latter depends on transcriptional regulation of Spry expression by FGF.

Fig. 2.

Feedback inhibitors of FGF signaling. Sef interacts with FGFRs and prevents phosphorylation of FRS2. A second less-defined mechanism suggests that Sef blocks FGF signaling downstream of MEK, and this is also the case for the intracellular splice variant Sef-b. Spry inhibits FGF signaling by sequestering Grb2, preventing its binding to FRS2; another inhibitory mechanism involves direct interaction of Spry with Raf. MKPs are negative regulators that work by dephosphorylating activated ERKs. MKP3 functions within the cytoplasm, whereas MKP1 is localized in the nucleus.

Screens to identify genes with restricted expression patterns during development have revealed further modulators of the FGF pathway (17, 18). The seemingly naïve premise that genes with similar complex expression patterns (syn-expression groups) belong to common pathways led to the discovery of several FGF target genes in Xenopus and zebrafish, including the Xenopus homolog of the fibronectin-leucine-rich transmembrane protein 3 gene (XFLRT3), Sef, Spry4, and map kinase phosphatase 3 (Mkp3) (10, 1721). Although Sprys and MKP3 were previously known as feedback inhibitors of Ras-MAPK signaling, their identification validated the approach (5, 22). The two FGF target genes Sef and XFLRT3 appear to be restricted to vertebrates, and both encode single-pass transmembrane proteins (1921). Homology searches with Sef revealed similarities to the interleukin-17 receptor, a conserved and functionally important putative tyrosine phosphorylation site within the intracellular region, and two fibronectin type-III (FNIII) domains facing the extracellular space (20, 21, 23). XFLRT3 represents a class of proteins that contains at least three members in humans. These proteins are characterized by a cluster of leucine-rich repeats and one FNIII domain within the extracellular region, and no protein signature motifs in the intracellular region (19).

In zebrafish, Sef functions as an antagonist of FGF signaling, and gene depletion studies have shown that it is required to limit the extent of FGF signaling during development (20, 21). The mechanism of Sef action is not fully understood (Fig. 2). The intracellular domain of Sef is required for its function as an inhibitor and for its interaction with FGF receptors FGFR1 and FGFR2 (20). Ectopic Sef expression blocks the phosphorylation of FGF receptor substrate 2 (FRS2), a mediator of FGF signaling that bridges FGFRs to the Ras-MAPK and PI3K pathways (24). This implies that Sef acts at the level of FGFRs and can block all signaling pathways downstream of the receptor, which was confirmed by showing that phosphorylation of ERK and MEK (the kinase downstream of Raf in the Ras-MAPK cascade), as well as phosphorylation of Akt, an effector of the PI3K pathway, was suppressed (2426). Overexpression assays in zebrafish, however, point to a role for Sef at the level of or downstream of MEK, because the activity of constitutively active MEK was blocked by ectopic Sef expression (21). To further complicate matters, at least two alternatively spliced isoforms of Sef exist in humans. Sef-b, an isoform that lacks the signal peptide and therefore is localized within the cytoplasm, can suppress FGF signaling downstream of MEK (27). Given that FGFs can deliver multiple biological signals, it is conceivable that like Spry, Sef can attenuate the FGF pathway at multiple points depending on context, possibly allowing fine-tuning of signal regulation (Fig. 2).

In contrast to Sef and Spry, XFLRT3 acts as a positive regulator of FGF signaling during Xenopus development (Fig. 3) (19). The embryonic expression of XFLRT3 mirrors that of XFgf8, and its expression varies with experimental changes in FGF activity. XFLRT3 overexpression in animal explants can activate the Ras-MAPK pathway as measured by the phosphorylation of ERK and the activation of the FGF target gene Xbrachyury (Xbra), a T-box transcription factor. The activation of Xbra by XFLRT3 was inhibited by dominant negative FGFR1 or by overexpression of MKP1, an ERK-specific phosphatase. In contrast, inhibitors of the PI3K pathway did not influence the ability of XFLRT3 to stimulate Xbra expression, pointing to specificity of this factor for the Ras-MAPK pathway. From gene knockdown studies, it seems that XFLRT3 is only required to relay Ras-MAPK signaling through the FGF8 ligand and not through the related FGF2 (bFGF) or FGF4 (eFGF) ligands. Because XFLRT3 can interact with FGFR1 and the FNIII domain is required for this interaction, specificity for certain ligands or FGFRs is indicated. The role of XFLRT3 may be to enhance FGF8-mediated signaling through the Ras-MAPK pathway, a hypothesis supported by the exact correspondence of initial XFLRT3 and XFgf8 expression patterns in the embryo (19).

Fig. 3.

Positive regulation of FGF signaling by XFLRT3. The positive regulator FLRT3 is a transmembrane protein that interacts with XFGFRs to facilitate FGF signaling and may provide selectivity for specific FGF ligands.

MKPs represent a family of dual-specificity phosphatases consisting of an N-terminal MAPK (ERK)-binding domain and a C-terminal phosphatase site that specifically inactivates phosphorylated forms of ERK (28, 29) (Fig. 2). The fact that MKPs are themselves under the regulation of RTK signaling implies that they represent feedback modulators (3032). Although both MKP1 and MKP3 function in a similar fashion in that they dephosphorylate the p42/p44 class of ERKs (ERK1 and ERK2), MKP1 is localized in the nucleus, whereas MKP3 is cytoplasmic, indicating that the Ras-MAPK pathway must be regulated in both compartments (33). MKP3 feedback modulation of the FGF pathway has been illustrated in chicken and zebrafish (3032). In both species, Mkp3 expression is regulated by FGF signaling, and ectopic expression of Mkp3 in the chick limb bud results in the disruption of limb outgrowth, a phenotype characteristic of blocking FGF signaling (3032). Likewise, Mkp3 limits the extent of FGF activity in the early zebrafish embryo, which is critical to maintaining dorsal-ventral polarity (32). Mkp3 may be a feedback inhibitor of pathways other than those stimulated by FGF, because genetic experiments in Drosophila indicate that Mkp3 acts to regulate EGF to Ras-MAPK signaling during ommatidial patterning (34). In Xenopus embryos, Mkp3 expression can be induced directly by retinoic acid and, in zebrafish embryos, the maternal β-catenin pathway appears to be responsible for the initiation of mkp3 expression; thus, multiple signaling pathways use Mkp3 to regulate Ras-MAPK signaling (32, 35, 36).

FGF activity eventually results in the regulation of gene expression through the modification of transcription factors by activated ERKs (Fig. 1). The ETS (E26 transformation-specific) transcription factors, specifically Ets-related molecule, PEA3-like (ERM), and polyomavirus enhancer activator 3 (PEA3), represent examples of proteins that are phosphorylated by activated ERKs, facilitating their interaction with DNA and with cofactors to induce gene expression (3739). Further, the erm and pea3 genes are transcriptionally activated by FGF signaling during zebrafish development, although it is not clear whether this is a direct activation or proceeds through a positive feed-forward network (4042). Such networks have been established in Xenopus, where XFgf4 regulates Xbra expression and in turn Xbra induces XFgf4 expression (43). At least in Xenopus, it is possible that Ets2 fulfills the role of the ETS transcription factor that is involved in mediating FGF target gene expression, providing a feed-forward link involving FGF, Ets2, and Xbra (44). This system of regulation allows the establishment of a biological signaling loop that is self-regulated, providing a basis for the formation of local organizing centers that are capable of providing patterning information during development.

It is clear that multiple feedback regulators of the FGF pathway act at different levels of the signal transduction cascade to promote or attenuate FGF signaling. To determine how these components are integrated within the pathway and between different pathways is the next challenge. In a simple model that includes feedback regulation, transcriptional output depends on signal strength, with feedback inhibition ensuring containment of activity within desired bounds; feedback regulation may also enhance threshold-type responses. This is undoubtedly too simplistic a model, because posttranslational modifications governed by FGF influence the activity of modulators such as Spry and MKP3 (14, 45). Cross-pathway modulation adds further complexity. FGF itself signals through three pathways, only one of which was considered here, and signaling pathways initiated by different classes of ligands also interact. Further, the modulators discussed here have different ranges of activity, so that Sef and XFLRT3 appear to be specific for FGF-induced signals, whereas Spry and Mkp3 affect RTK signals initiated by different ligands. Thus, the remarkable complexity of signaling networks is coming more clearly into focus with every new insight in the field.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
View Abstract

Stay Connected to Science Signaling

Navigate This Article