Association of the Signaling Adaptor FRS2 with Fibroblast Growth Factor Receptor 1 (Fgfr1) Is Mediated by Alternative Splicing of the Juxtamembrane Domain*

Fibroblast growth factor receptors (FGFRs) are a family of transmembrane tyrosine kinases involved in signaling via interactions with the family of fibroblast growth factors. Alternative splicing of the juxtamembrane region of FGFR1–3 leads to the inclusion or exclusion of two amino acids, valine and threonine, the VT site. The presence or absence of VT (VT (cid:1) or VT (cid:2) , respec-tively) affects the signaling potential of the receptor. The VT (cid:1) receptor isoform is required for Erk2 phosphorylation, a component of the mitogen-activated protein kinase signaling pathway. FRS2 is an adaptor protein that links FGFRs to the mitogen-activated protein kinase signaling pathway. FRS2 interacts with a region of the juxtamembrane domain of FGFR1 that includes the alternatively spliced VT site. We investigated the interaction of FRS2 with murine Fgfr1 juxtamembrane domain. We showed the alternatively spliced VT motif, at the juxtamembrane domain of Fgfr1 is required for FRS2 interaction with Fgfr1. Activation of signaling pathways from FRS2 is likely to be regulated by control-ling the Fgfr1/FRS2 interaction through alternative splicing of the VT motif of Fgfr1. Fibroblast growth factor receptors (FGFRs) 1 are a

Fibroblast growth factor receptors (FGFRs) 1 are a family of four transmembrane receptor tyrosine kinases whose signaling functions are executed via interactions with the fibroblast growth factor family of ligands (FGFs) (1,2). FGFR activation induces signaling cascades that result in multiple biological responses, including cell proliferation, differentiation, migration, and survival (2)(3)(4).
Four unlinked FGFR genes have been cloned in humans (FGFR1-4), mice (Fgfr1-4), chickens, frogs, and newts (5)(6)(7)(8)(9)(10), and homologues have also been identified in Drosophila, Caenorhabditis elegans and sea urchins (11)(12)(13). A key feature of FGFR1 to -3 is that they are subject to post-transcriptional regulation by alternate splicing, resulting in the expression of multiple receptor isoforms (14). Alternative splicing of the third immunoglobulin-like domain (IgIII) in the extracellular region of the receptor determines ligand specificity (15). IgIII has a common IgIIIa exon that is spliced to either an IgIIIb or an IgIIIc variant exon (16 -20). The inclusion of the IIIb or IIIc exon determines receptor binding specificity.
A second splicing event involves the use of alternative donor splice sites located 6 bp apart at the 3Ј-end of exon 10, which leads to the inclusion or exclusion of two amino acids, valine 427 and threonine 428 (FGFR1 numbering) in the intracellular juxtamembrane region (21)(22)(23), producing two forms of the receptor, here termed VTϩ and VTϪ. Several studies have implicated the VT motif in the intracellular signaling functions of the receptor. In Xenopus, the Thr of the VT motif of the FGFR1 VTϩ isoform is phosphorylated by protein kinase C (PKC). The biological activity of VTϩ FGFR1 was reduced by phosphorylation by PKC, suggesting that PKC could negatively regulate FGFR1 VTϩ activity (22). Studies in transfected mammalian cells have shown that the VTϩ, but not the VTϪ form, could activate the Ras/MAPK signaling cascade, as measured by phosphorylation of Erk2 (23). This finding indicates that a major downstream channel of FGFR signaling is controlled by alternative splicing of the VT motif and that this splicing mechanism could therefore have important implications for FGFR function.
Two adaptor proteins, FGF receptor substrate 2 (FRS2, FRS2␣, or SNT1) and FGF receptor substrate 3 (FRS3, FRS2␤, or SNT2), have been implicated in downstream signaling from activated FGFRs by means of direct interaction with the receptor (24 -26). These proteins share 49% sequence identity and are composed of an N-terminal phosphotyrosine binding (PTB) motif containing an N-terminal myristoylation site and an effector domain, which contains multiple potential sites for tyrosine phosphorylation by activated FGFRs. FRS2 and FRS3 exhibit distinct spatial-temporal patterns of gene expression in vivo (27). Disruption of murine FRS2 is embryonic lethal at stage E7-E7.5 indicating that FRS2 plays an essential role in early development (28).
In addition, protein kinase C (PKC) and protein kinase C (PKC), members of the atypical PKCs, have been shown to interact with FRS2 following FGF stimulation (31). Activation of PKC is necessary and sufficient for its association with FRS2, although FRS2 was not a substrate for the PKCs. Activity of the atypical PKCs has been shown to be necessary for mitogenic signaling via the MAPK cascade (32,33). An active mutant of PKC activated MAPK and mitogen-activated pro-tein kinase/extracellular signal-regulated kinase kinase, and a dominant negative PKC impaired their activation. Therefore, FRS2 may activate MAPK by two pathways, through recruitment of Grb2 (28) and atypical PKCs (31).
FRS2 also mediates recruitment of phosphoinositol 3-kinase to FGFRs via interaction of an adaptor protein, Gab1, with FRS2-associated Grb2 (28,34). Activation of phosphoinositol 3-kinase is involved in signal transduction downstream of most tyrosine kinase receptors and is implicated in the regulation of mitogenesis, migration, and cell survival (35). Taken together, this evidence indicates that the interaction of FRS2 with FGFRs and the concomitant formation of a multiprotein signaling complex is a central element in multiple cellular responses to FGFR signaling.
The juxtamembrane region of FGFR1 and the PTB domain of FRS2 are required and sufficient for interaction and phosphorylation of FRS2 (26, 36 -38). The sequence within the juxtamembrane region required for specific interaction with FRS2 was identified as KSIPLRRQVTVS (amino acids 419 -430) (37), which includes the alternatively spliced VT motif. Alanine-scanned mutagenesis of residues Lys 419 , Ile 421 , Phe 422 , Leu 423 , Arg 425 , Val 427 , and Val 429 of FGFR1 significantly diminished the binding of FRS2 to FGFR1 and had an inhibitory effect on MAPK activation. The structure of FRS2 PTB domain in a 1:1 complex with a peptide from the juxtamembrane region of human FGFR1 (amino acids 409 -430) has been solved by nuclear magnetic resonance methods (38). In combination with site-directed mutagenesis, residues Leu 423 and Val 429 of FGFR1 were found to be most important for the FRS2/FGFR1 interaction followed by Val 414 , Leu 417 , Arg 425 , and Val 427 , supporting Ong et al. (37).
Taken together, this evidence suggests that the functional significance of the VTϩ and VTϪ alternatively spliced isoforms of FGFRs might be to control the formation of the FRS2dependant signaling complex, thereby regulating the signaling processes that are activated by ligand engagement. Here we test this hypothesis by analysis of the interaction of FRS2 with different forms of the juxtamembrane domain of murine Fgfr1. We show that inclusion of the VT motif is required for binding of FRS2 to Fgfr1 in a kinase-independent manner. We also show that VTϩ/VTϪ splicing is regulated independently of splicing in the extracellular domain and that both VT isoforms are co-expressed in the same organs in vivo.
Truncated FcFgfr1-Using the FcFgfr1 VTϩ as a template, the truncated receptor was amplified by PCR (primers: HB11 (CGCGGATC-CGAGCTCACCATGGTCAGCTGGGGTCGTTTC) and negative control (CGTAGATCGCGGCCGCTCAGCTCTTCTTGGTGCCGCT)). The PCR product was inserted into the FcFgfr1 construct using BamHI and NotI sites, replacing FcFgfr1 VTϩ with the truncated version.
Amplification of FRS2 cDNA-Rat PC12 RNA was used in RT-PCR. Qiagen Omniscript reverse transcriptase kit was used according to the manufacturer's instructions with a poly(dT) primer. Rat FRS2 cDNAs were amplified from the cDNA pool (primers: FRS2 5ЈKpnI (GCCGG-TACCATGGGTAGCTGTTGTAGCTGT) and FRS2 (3ЈXhoI CGGCCTC-GAGACATGGGCAGATCAGTACTATTGTG)). Using KpnI and XhoI sites, the cDNA was inserted into the vectors pSECTAG 2B (Invitrogen) and pcDNA3 (Invitrogen). The 3Ј-end of FRS2 linked to Myc and poly-(His) tags was amplified by PCR from the FRS2 pSECTAG 2B (primers: FRS2 5Јmiddle (TAGGGCCAACCCCTGTTC) and XbaI FRS2 3Ј (GGT-TCTAGATCAATGATGATGATGATGATG)). The PCR product was inserted into FRS2 pcDNA3 using HpaI and XbaI sites. The final construct had FRS2 tagged at the 3Ј-end with Myc and poly(His) in pcDNA3.
FRS2 PTB-Cloned truncated FRS2 containing only the myristoylation and PTB domains using FRS2 as a template (primers: FRS2 5ЈKpnI (GCCGGTACCATGGGTAGCTGTTGTAGCTGT) and FRS2 PTB 3Ј (CCGCTCGAGAGTCTCCGAACGAGGGGTATC)) was inserted into the plasmid using KpnI and XhoI sites, replacing FRS2 with the truncated form.

Transient Transfections
293T cells were transiently transfected using CaPO 4 precipitation. Cells were grown to 70% confluency in 162-cm 2 flasks in 25 ml of media. To 155 l of 2 M CaCl 2 , 60 g of Qiagen prepared DNA were added. Water was added to a final volume of 1.5 ml, and the DNA/CaCl 2 mixture was added dropwise to 1.5 ml of 2ϫ HBS (1.6% NaCl (w/v), 1.2% Hepes (w/v), 0.04% Na 2 HPO 4 (w/v), pH 7.12). The DNA/CaPO 4 precipitate was incubated on the cells overnight and washed the following morning in media without serum. This was replaced with 25 ml of Ultracho media (Bio-Whittaker) supplemented with 0.1 mg/ml strepto-mycin and 0.2 units/ml of penicillin. Recombinant proteins were expressed for 48 h.
Immunoprecipitation-Transiently transfected 293T cells were washed twice in 25 ml of cold PBS and then lysed at 4°C using 2 ml of lysis buffer (10 mM Tris-HCl, pH 8.0, 2.5 mM MgCl 2 , 5 mM EGTA pH 8.0, 0.5% Triton X-100 (w/v), 1 mM Na 3 VO 4 , 50 mM NaF, and 1 tablet of protease inhibitor mixture (Roche Molecular Biochemicals) per 10 ml of buffer) per 162-cm 2 flask of cells. The supernatant was used for immunoprecipitation. To 15 l of protein A-Sepharose 4 fast flow (Amersham Biosciences, Inc.) 1 ml of 293T cell lysate was added and incubated for 1 h, at 4°C, with mixing. The protein A-Sepharose was washed three times with 1 ml of PBS. Washed protein A-Sepharose resuspended in 500 l of PBS was put down a sucrose column (400 l of 20% sucrose, 300 l of 10% sucrose) to remove precipitated proteins. After a final 1-ml wash in PBS, bound protein was eluted by boiling for 5 min in 15 l of 2ϫ SDS sample buffer (125 mM Tris/HCl, pH 6.8, 20% glycerol (w/v), 4% SDS (w/v), 0.1% bromphenol blue (w/v), 10% ␤-mercaptoethanol (w/v)) for analysis by SDS-PAGE.
Western Blotting-Each 4 -20% SDS-PAGE gel (Invitrogen) was calibrated with 3 l of high molecular weight rainbow markers (Amersham Pharmacia Biotech). Protein was transferred to polyvinylidene difluoride membrane (Millipore Corp.) at 100 V for 1 h. The membrane was blocked overnight at 4°C in 5% bovine serum albumin (w/v), 0.1% azide (w/v) in TBS-T (10 mM Tris/HCl, pH 7.4, 75 mM NaCl, 0.05% Tween 20 (v/v)). Primary antibodies were diluted in 1% bovine serum albumin (w/v) in TBS-T and incubated with the membrane for 1 h at room temperature. The blot was washed three times for 15 min in TBS-T and probed with the conjugated antibody for 30 min at room temperature, diluted in 1% bovine serum albumin (w/v) in TBS-T. The filter was washed five times in TBS-T, followed by a final wash in TBS (no Tween 20). Enzyme chemiluminescence (Pierce) was used for visualization and exposed to hyperfilm (Amersham Biosciences). Blots were stripped for reprobing with 0.1 M glycine-HCl, pH 2.5, for 5 min at room temperature.
Expression and Purification of GST Fusion Proteins-A 1-liter culture of Escherichia coli XL-1 Blue cells transformed with the appropriate pGEX construct was grown in Lennox Broth (10 g/liter tryptone, 5 g/liter yeast extract, 10 g/liter NaCl, pH 7.4) supplemented with ampicillin (100 g/ml) at 37°C to an A 600 of 0.6 -0.7. Protein was expressed at 25°C for 4 h with 0.1 mM isopropyl-1-thio-␤-D-galactopyranoside. The bacteria were pelleted and resuspended in 5 ml of MT-PBS (150 mM NaCl, 16 mM Na 2 HPO 4 , 4 mM NaH 2 PO 4 ). Phenylmethylsulfonyl fluoride and EDTA were added to a 1 mM final concentration. Cells were sonicated four times for 15 s on ice and incubated on ice for 5 min in 1% Triton X-100. Cell debris was pelleted at 15,000 rpm for 20 min at 4°C. The proteins were purified on 0.8 ml of glutathione-Sepharose 4B (Amersham Pharmacia Biotech). The column was prewashed with 5 ml of MT-PBS and MT-PBS, 1% Triton. The cell supernatant was applied, and the column was washed with 2.5 ml of MT-PBS, 0.1% Triton, 2.5 ml of MT-PBS and twice with 3 ml of wash buffer, pH 8.0 (50 mM Tris-HCl pH 8.0, 150 mM NaCl). The GST fusion was eluted with 0.5-ml aliquots of 20 mM glutathione in wash buffer, pH 8.0. To cleave the protein, the column was washed twice in 3 ml of TNED (50 mM Tris HCl, pH 8.0, 150 mM NaCl, 10 mM EDTA, pH 8.0, 1 mM dithiothreitol) and incubated with 10 g of 3C protease overnight at room temperature with rotation. The cleaved protein was collected and washed from the column with 0.5-ml aliquots of TNED. Purity of the GST fusion and cleaved protein was assessed by Coomassie staining of SDS-PAGE gels and was assessed to be ϳ50%.
FRS2 Enzyme-linked Immunosorbent Assays-Purified, cleaved VTϩ or VTϪ Fgfr1 juxtamembrane region was coated onto Nunc Im-munoTM plates at 5 g/well in a 100-l volume. Wells were washed three times each with PBS, 0.1% Tween and blocked with 150 l of 1% bovine serum albumin in PBS for 4 h at room temperature. A doubling dilution series of purified GST-FRS2 (0.25 g/l to 2.44 ϫ 10 Ϫ4 g/l diluted in PBS) was incubated overnight at room temperature (100 l/well). A blank (PBS only) was also included. The following morning the wash step was repeated, and 100 l of anti-GST diluted 1:1000 in PBS was incubated for 1 h at room temperature. The wash step was repeated, and 100 l of anti-mouse IgG HRP diluted 1:5000 in PBS was incubated for 1 h at room temperature. The wells were washed again before applying the OPD substrate (Dako) used according to the manufacturer's instructions. The reaction was stopped after 5 min with 100 l of 0.5 M H 2 SO 4 . Absorbance at A 490 was measured. Triplicate samples were used.
RNA Isolation and RT-PCR-Total RNA was extracted from mouse tissues using Trizol® Reagent (Life Technologies, Inc.). One-step RT-PCR was performed on 2 g of total RNA (primers: fgfr.22L, GGC TCA CTG CAC CAG ACC TG; fgfr.27R, GAG CTC ATA TTC GGA GAC TCC AGC) using the Qiagen OneStep RT-PCR kit. The reverse transcription was performed according to the manufacturer's specifications. The PCR was performed in 30 cycles: 92°C, 20 s; touch-down 60 to 50°C, 15 s; 72°C, 20 s. The PCR product was digested with PshAI (New England Biolabs) and run on an agarose gel with a 1-kb ladder (Invitrogen).

Construction of a Constitutively Active Form of Murine Fgfr1
(FcFgfr1)-A constitutively active form of Fgfr1 (FcFgfr1) was generated by replacing the extracellular region of murine Fgfr1 with the Fc region of a human antibody molecule, IgG1. Part of the extracellular domain (last 24 bp) and the whole of the transmembrane and intracellular regions of Fgfr1 were fused to the Fc (Fig. 1). In this configuration, Fgfr1 kinase is activated constitutively via disulfidemediated dimerization of the Fc extracellular domain. Functional studies were used to test for constitutive activation by assessing receptor dimerization and intracellular tyrosine phosphorylation. The receptor was transiently expressed in 293T cells, the cells were lysed, and the receptor was immunoprecipitated by direct interaction of the Fc of the receptor with protein A-Sepharose. Expression of the receptor was observed using an anti-Fc HRP antibody by Western blot (Fig. 2a). Reprobing the same blot with an anti-phosphotyrosine antibody revealed that the receptor had tyrosine kinase activity (Fig. 2b). Under reducing conditions, the receptor ran as a monomer of ϳ100 kDa. Under nonreducing conditions, disulfide bonds remained intact, and oligomeric complexes of the receptor were detected (Fig. 2c). Therefore, the Fc fusion receptor was constitutively active in the absence of ligand. Interaction of the receptor with known signaling proteins (PLC␥ and Grb2) was confirmed, and the receptor was shown to correctly localize to the plasma membrane of transfected cells (results not shown).
Using site-directed mutagenesis, mutant forms of the constitutively active FcFgfr1 receptor were constructed that block docking sites for intracellular signaling proteins. A "kinase- dead" receptor was made that has the catalytic base in the phosphotransfer reaction Asp 623 mutated to Ala (40), rendering the receptor enzymatically inactive despite receptor dimerization. A Y766F mutant blocks the association and activation of PLC␥, as previously described (41)(42)(43). Mutants were made of the alternatively spliced VT motif, which has been implicated in Ras/MAPK regulation (23). Point mutations of the VT site were constructed, a VT to VA mutation and VT to TT mutation. Finally, a truncated form of the constitutively active receptor was generated in which the entire intracellular region of the receptor was deleted, leaving only the Fc for dimerization and the transmembrane domain for localization at the plasma membrane.
The VT Motif of Fgfr1 Is Required for FRS2 Interaction-A full-length FRS2 cDNA was isolated from PC12 cell RNA by RT-PCR and tagged at the N terminus with Myc and poly-His epitopes for immunodetection. The interaction of FRS2 with FcFgfr1 was studied by comparing FRS2 association with the constitutively active Fgfr and the mutant forms described above by transient co-expression in 293T cells and immunoprecipitation (Fig. 3a).
Under these experimental conditions, FRS2 bound to both the kinase-active and kinase-inactive VTϩ Fgfr1 isoforms, confirming constitutive interaction of FRS2 with Fgfr1 as previously described (26,37). This association was abolished by removal of the VT site (Fig. 3a). Thus, in co-transfected cells, FRS2 associates with the VTϩ but not the VTϪ form of the Fgfr1 intracellular domain. The requirement for specific residues in the VT motif was explored by analysis of point mutants. A VT to VA mutation had a minimal effect on the Fgfr1/FRS2 interaction. However, mutation of VT to TT almost abolished complex formation. This finding indicates that the identity of the valine residue in the VT motif has a more significant contribution to the FRS2 interaction than the neighboring threonine. In this experimental system, the Thr has little impact on FRS2 recognition. Mutation of Tyr 766 in the C-terminal tail had no effect on FRS2 association, and removal of the entire intracellular region of the receptor abolished FRS2 interaction with the receptor.
The ability of FRS2 to associate with the endogenous effector protein Grb2 was examined by reprobing the immunoprecipitates with an anti-Grb2 antibody (Fig. 3a). Grb2 association was found to require both association of FRS2 with FcFgfr1 and an active Fgfr1 kinase; no Grb2 was isolated with the kinasedead form of Fgfr1. In addition, no Grb2 was found to be associated with the VTϪ form of the receptor, which correlates with the absence of FRS2 association with this form of the receptor. This indicates that, in 293T cells, association of Grb2 with tyrosine-phosphorylated FRS2 is the principle potential connection with the Ras/MAPK pathway. Grb2 has been reported to associate with Fgfr1 via binding the adaptor protein Shc (44) that requires a functional kinase domain (45). Since loss of FRS2 binding to the receptor leads to concomitant abrogation of Grb2 binding, it seems probable that Shc-dependent activation of the MAPK signaling cascade may not be a major pathway in 293T cells. been previously shown that FRS2 interacts with the Fgfr1 juxtamembrane region through the FRS2 PTB domain (26,37,38). The coexpression/immunoprecipitation experiments described above were repeated with a truncated form of FRS2, containing the myristoylation site and PTB domain of FRS2 (Fig. 3b). The same interaction of FRS2 PTB domain with Fgfr1 as wild type FRS2 was observed, confirming that FRS2 interacts with the Fgfr1 VT site through its PTB domain. In these experiments, no association of Grb2 was detected, since the Grb2 recognition epitopes are located in the C-terminal effector domain of FRS2 (24).
Direct Interaction of FRS2 with Fgfr1 VTϩ-The interaction of FRS2 with Fgfr1 VTϩ could either be direct or indirect mediated by additional proteins present in 293T cells. To test this, we examined the interaction of purified Fgfr1 juxtamembrane domain with purified FRS2 in solution. The juxtamembrane regions of Fgfr1 with and without the VT motif and FRS2 fused to GST were expressed and purified from bacteria. Direct interaction of the VTϩ/VTϪ juxtamembrane proteins with GST-FRS2 was investigated by a modified enzyme-linked immunosorbent assay (Fig. 4). GST-FRS2 bound preferentially to the VTϩ form of the Fgfr1 juxtamembrane region but not to the VTϪ form of the receptor. This results shows that FRS2 can interact directly with the Fgfr1 VTϩ form of the juxtamembrane domain, although the need for further co-factors to form a stable high affinity complex between FRS2 and Fgfr1 cannot be eliminated, since the estimated EC 50 of the interaction in solution is ϳ240 nM.
Expression of Fgfr1 VTϩ and VTϪ RNAs in Mouse Tissues-The experiments described above show that alternative splicing of the VT motif determines binding of FRS2 to Fgfr1 (and presumably Fgfr2 and -3, which also exhibit VT splicing (23)). As documented above, the association of FRS2 links Fgfr activation with multiple signaling pathways. Therefore, activation of these pathways could be regulated in part by the expression levels of the Fgfr1 VTϩ and VTϪ isoforms. RT-PCR was used in order to determine relative expression levels of Fgfr1 VTϩ and VTϪ in different mouse tissues. The VT motif is produced by alternative splicing of six nucleotides at the end of exon 10. Exons 8 and 9 of the Fgfr1 are also alternatively spliced to exon 10, giving rise to the IIIb and IIIc splice variants, respectively. The forward primer was selected to distinguish the IIIb (exon 8) and IIIc (exon 9) splice variants, and enzyme digestion was used to detect the VTϩ and VTϪ forms of Fgfr1. In all tissues studied, the major splice variants IIIb and IIIc existed in combination with both the VTϩ and VTϪ forms of the receptor (Fig. 5 and data not shown). It is interesting to note that differential tissue-specific expression of the VTϩ and VTϪ isoforms was not observed in the tissues examined. DISCUSSION Here we have shown that alternative splicing of the juxtamembrane region of Fgfr1, leading to the inclusion or omis-sion of the two-amino acid VT motif, controls the ability of the receptor to associate with the signaling adaptor protein FRS2. As a result of this splicing event, signaling functions associated with adaptors, such as Grb2, that associate with activated Fgfrs via FRS2 is abrogated. This finding explains our previous report (23), which showed that the VTϪ splice variant was unable to activate the Ras/MAPK pathway in transfected cells.
These findings also confirm that the FRS2 PTB domain constitutively interacts with the juxtamembrane region of Fgfr1 (26, 36 -38). From single amino acid substitutions of the VT site, it emerges that FRS2 interaction with valine of the VT site is most significant. By contrast, mutation of the threonine residue has little effect on FRS2 binding. This indicates that phosphorylation of the threonine by PKC or other kinases may not play a significant role in FRS2 docking. This is supported by mutagenesis studies and NMR structural data by others (37,38). The structural data also provide an explanation for the loss of affinity for FRS2 exhibited by the VTϪ variant; it essentially creates a "frameshift" between residues that are required for binding to the PTB domain of FRS2.
Here we have shown that the loss of FRS2 adaptor binding to Fgfr1 results in concomitant loss of other effectors that bind to FRS2, such as Grb2. Signaling pathways likely to be influenced by VT splicing include activation of MAPK through recruitment of Grb2-Sos to FRS2 (44,46); recruitment of PKC and PKC, which have also been shown to activate MAPK (31)(32)(33); and Gab1-dependent recruitment of phosphoinositol 3-kinase (28,34). However, other signaling channels dependent on receptor kinase activity such as recruitment of PLC␥ described here remain intact in the VTϪ isoform.
No significant differences in levels of expression of the two alternatively spliced Fgfr1 isoforms were detected between different mouse tissues using RT-PCR. The VTϩ and VTϪ Fgfr1 isoforms were co-expressed in all tissues studied. However, differential expression of VTϩ and VTϪ isoforms of Fgfr1 has been identified during mesoderm formation in Xenopus embryos (47), and a more extensive survey in the mouse may FIG. 4. Direct binding of GST-FRS2 to VT؉ Fgfr1 juxtamembrane region. VTϩ and VTϪ Fgfr1 juxtamembrane regions purified from bacteria were coated on enzyme-linked immunosorbent assay plates and incubated with purified GST-FRS2. Direct binding of GST-FRS2 to the juxtamembrane regions was detected using an anti-GST antibody.
FIG. 5. Tissue alternative splicing of Fgfr1. a, schematic representation of Fgfr1 cDNA exons 6 -8, 10, and 11. Exon 8 (IIIb) is alternatively spliced to exons 7 and 10. VT, valine and threonine amino acids. RT-PCR product 22L-27R, detecting the IIIb form, was 338 bp (VTϪ) or 344 bp (VTϩ). If the VT motif is absent, the PCR product (338 bp) can be digested with PshAI to produce two bands of 227 and 111 bp. b, the VTϩ (344 bp) and VT-(227 bp) fragments are indicated with arrows on the right. Both products were present in every tissue tested, but their quantities varied. The larger bands indicated with an asterisk contain both IIIb and IIIc exons, and their physiological significance is not clear. A similar experiment using a forward primer on exon 9 (IIIc) and the same reverse primer was performed and showed the same results (data not shown). reveal cell types in which differential expression of the VT isoforms does occur. If the relative levels of mRNA are reflected in expression of receptor protein, this indicates that alternative splicing of the juxtamembrane region represents bifurcation in Fgfr signaling in which different signal outputs may be quantitatively regulated. Combined with the tissue-specific expression of the FRS2 adaptor, this contributes to a mechanism by which different signaling channels may be selectively activated during Fgfr signaling in different cell types.