Limits...
Structure of the tandem fibronectin type 3 domains of neural cell adhesion molecule.

Carafoli F, Saffell JL, Hohenester E - J. Mol. Biol. (2008)

Bottom Line: The two putative FGFR1-binding segments, one in each NCAM FN3 domain, are situated close to the domain interface.Surface plasmon resonance experiments demonstrated only a very weak interaction between the NCAM FN3 tandem and soluble FGFR1 proteins expressed in mammalian cells (dissociation constant >100 muM).Thus, the NCAM-FGFR1 interaction at the cell surface is likely to depend upon avidity effects due to receptor clustering.

View Article: PubMed Central - PubMed

Affiliation: Department of Life Sciences, Biophysics Section, Blackett Laboratory, Imperial College London, London SW7 2AZ, UK.

ABSTRACT
Activation of the fibroblast growth factor receptor (FGFR) by neural cell adhesion molecule (NCAM) is essential for NCAM-mediated neurite outgrowth. Previous peptide studies have identified two regions in the fibronectin type 3 (FN3)-like domains of NCAM as being important for these activities. Here we report the crystal structure of the NCAM FN3 domain tandem, which reveals an acutely bent domain arrangement. Mutation of a non-conserved surface residue (M610R) led to a second crystal form showing a substantially different conformation. Thus, the FN3 domain linker is highly flexible, suggesting that it corresponds to the hinge seen in electron micrographs of NCAM. The two putative FGFR1-binding segments, one in each NCAM FN3 domain, are situated close to the domain interface. They form a contiguous patch in the more severely bent conformation but become separated upon straightening of the FN3 tandem, suggesting that conformational changes within NCAM may modulate FGFR1 activation. Surface plasmon resonance experiments demonstrated only a very weak interaction between the NCAM FN3 tandem and soluble FGFR1 proteins expressed in mammalian cells (dissociation constant >100 muM). Thus, the NCAM-FGFR1 interaction at the cell surface is likely to depend upon avidity effects due to receptor clustering.

Show MeSH

Related in: MedlinePlus

SDS-PAGE analysis of recombinant NCAM and FGFR1 proteins. Coomassie blue-stained gel of His-tagged soluble proteins expressed in 293-EBNA cells. The positions of molecular mass standards (in kilodaltons) are indicated on the left. The FGFR1 proteins are modified by extensive glycosylation; the NCAM proteins are not glycosylated.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2267215&req=5

fig5: SDS-PAGE analysis of recombinant NCAM and FGFR1 proteins. Coomassie blue-stained gel of His-tagged soluble proteins expressed in 293-EBNA cells. The positions of molecular mass standards (in kilodaltons) are indicated on the left. The FGFR1 proteins are modified by extensive glycosylation; the NCAM proteins are not glycosylated.

Mentions: We wanted to map the FGFR1 binding site on the NCAM FN3 tandem by structure-based mutagenesis and first sought to establish a suitable binding assay. A solid-phase assay with immobilised NCAM and Fc-tagged FGFR1 proteins did not show any appreciable interaction (data not shown). We therefore used surface plasmon resonance (SPR) to analyse the binding of NCAM 1FN3–2FN3 to two FGFR1 ectodomain constructs. The FGFR1 D1–D3 construct used (residues 22–364) spans essentially the full ectodomain and contains the acid box situated between domains D1 and D2. The FGFR1 D2–D3 construct used (residues 151–364) lacks D1 and the acid box but retains the binding site for FGFs; this construct is similar to the construct previously used by Kiselyov et al. in SPR studies.28 Both soluble FGFR1 proteins were produced by the 293-EBNA cells in good yields. Due to the presence of multiple N-linked glycosylation sites in FGFR1 (see below), the purified recombinant proteins migrate as diffuse bands of higher-than-calculated molecular mass on SDS-PAGE (Fig. 5). In a first set of experiments, the two FGFR1 constructs were immobilised on a CM5 sensor chip [8000 resonance units (RU) of D1–D3 and 3850 RU of D2–D3]. Recombinant FGF1 injected at a concentration of 100 nM produced sensorgrams characteristic of a high-affinity interaction, confirming that the immobilised proteins are functional (Fig. 6a and b). In contrast, wild-type NCAM 1FN3–2FN3 up to a concentration of 70 μM did not produce a signal on the FGFR1 D1–D3 surface and showed only very weak binding to FGFR D2–D3 (Fig. 6c and d). In a second set of experiments, the order of proteins was reversed. NCAM 1FN3–2FN3 proteins were immobilised on a CM4 sensor chip (1800 RU of wild-type protein and 1900 RU of M610R mutant), and the two soluble FGFR1 constructs were used as analytes up to a concentration of 100 μM. We again observed only weak interactions for all pairings (Fig. 6e–h). Wild-type and M610R NCAM 1FN3–2FN3 behaved almost identically in these experiments, and, as before, it appeared that the affinity of NCAM for FGFR1 D2–D3 was higher than that for FGFR1 D1–D3. The fast association and dissociation steps in the sensorgrams prevented the fitting of kinetic constants. We used the plateau values at equilibrium to estimate a dissociation constant of > 100 μM for the interaction of NCAM 1FN3–2FN3 with FGFR1 D2–D3 (not shown), but we emphasise that this value is very approximate given the weak resonance signals obtained. In view of the weakness of the NCAM–FGFR1 interaction in our assay, we were unable to pursue our initial plans of mapping the binding site(s) by mutagenesis.


Structure of the tandem fibronectin type 3 domains of neural cell adhesion molecule.

Carafoli F, Saffell JL, Hohenester E - J. Mol. Biol. (2008)

SDS-PAGE analysis of recombinant NCAM and FGFR1 proteins. Coomassie blue-stained gel of His-tagged soluble proteins expressed in 293-EBNA cells. The positions of molecular mass standards (in kilodaltons) are indicated on the left. The FGFR1 proteins are modified by extensive glycosylation; the NCAM proteins are not glycosylated.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2267215&req=5

fig5: SDS-PAGE analysis of recombinant NCAM and FGFR1 proteins. Coomassie blue-stained gel of His-tagged soluble proteins expressed in 293-EBNA cells. The positions of molecular mass standards (in kilodaltons) are indicated on the left. The FGFR1 proteins are modified by extensive glycosylation; the NCAM proteins are not glycosylated.
Mentions: We wanted to map the FGFR1 binding site on the NCAM FN3 tandem by structure-based mutagenesis and first sought to establish a suitable binding assay. A solid-phase assay with immobilised NCAM and Fc-tagged FGFR1 proteins did not show any appreciable interaction (data not shown). We therefore used surface plasmon resonance (SPR) to analyse the binding of NCAM 1FN3–2FN3 to two FGFR1 ectodomain constructs. The FGFR1 D1–D3 construct used (residues 22–364) spans essentially the full ectodomain and contains the acid box situated between domains D1 and D2. The FGFR1 D2–D3 construct used (residues 151–364) lacks D1 and the acid box but retains the binding site for FGFs; this construct is similar to the construct previously used by Kiselyov et al. in SPR studies.28 Both soluble FGFR1 proteins were produced by the 293-EBNA cells in good yields. Due to the presence of multiple N-linked glycosylation sites in FGFR1 (see below), the purified recombinant proteins migrate as diffuse bands of higher-than-calculated molecular mass on SDS-PAGE (Fig. 5). In a first set of experiments, the two FGFR1 constructs were immobilised on a CM5 sensor chip [8000 resonance units (RU) of D1–D3 and 3850 RU of D2–D3]. Recombinant FGF1 injected at a concentration of 100 nM produced sensorgrams characteristic of a high-affinity interaction, confirming that the immobilised proteins are functional (Fig. 6a and b). In contrast, wild-type NCAM 1FN3–2FN3 up to a concentration of 70 μM did not produce a signal on the FGFR1 D1–D3 surface and showed only very weak binding to FGFR D2–D3 (Fig. 6c and d). In a second set of experiments, the order of proteins was reversed. NCAM 1FN3–2FN3 proteins were immobilised on a CM4 sensor chip (1800 RU of wild-type protein and 1900 RU of M610R mutant), and the two soluble FGFR1 constructs were used as analytes up to a concentration of 100 μM. We again observed only weak interactions for all pairings (Fig. 6e–h). Wild-type and M610R NCAM 1FN3–2FN3 behaved almost identically in these experiments, and, as before, it appeared that the affinity of NCAM for FGFR1 D2–D3 was higher than that for FGFR1 D1–D3. The fast association and dissociation steps in the sensorgrams prevented the fitting of kinetic constants. We used the plateau values at equilibrium to estimate a dissociation constant of > 100 μM for the interaction of NCAM 1FN3–2FN3 with FGFR1 D2–D3 (not shown), but we emphasise that this value is very approximate given the weak resonance signals obtained. In view of the weakness of the NCAM–FGFR1 interaction in our assay, we were unable to pursue our initial plans of mapping the binding site(s) by mutagenesis.

Bottom Line: The two putative FGFR1-binding segments, one in each NCAM FN3 domain, are situated close to the domain interface.Surface plasmon resonance experiments demonstrated only a very weak interaction between the NCAM FN3 tandem and soluble FGFR1 proteins expressed in mammalian cells (dissociation constant >100 muM).Thus, the NCAM-FGFR1 interaction at the cell surface is likely to depend upon avidity effects due to receptor clustering.

View Article: PubMed Central - PubMed

Affiliation: Department of Life Sciences, Biophysics Section, Blackett Laboratory, Imperial College London, London SW7 2AZ, UK.

ABSTRACT
Activation of the fibroblast growth factor receptor (FGFR) by neural cell adhesion molecule (NCAM) is essential for NCAM-mediated neurite outgrowth. Previous peptide studies have identified two regions in the fibronectin type 3 (FN3)-like domains of NCAM as being important for these activities. Here we report the crystal structure of the NCAM FN3 domain tandem, which reveals an acutely bent domain arrangement. Mutation of a non-conserved surface residue (M610R) led to a second crystal form showing a substantially different conformation. Thus, the FN3 domain linker is highly flexible, suggesting that it corresponds to the hinge seen in electron micrographs of NCAM. The two putative FGFR1-binding segments, one in each NCAM FN3 domain, are situated close to the domain interface. They form a contiguous patch in the more severely bent conformation but become separated upon straightening of the FN3 tandem, suggesting that conformational changes within NCAM may modulate FGFR1 activation. Surface plasmon resonance experiments demonstrated only a very weak interaction between the NCAM FN3 tandem and soluble FGFR1 proteins expressed in mammalian cells (dissociation constant >100 muM). Thus, the NCAM-FGFR1 interaction at the cell surface is likely to depend upon avidity effects due to receptor clustering.

Show MeSH
Related in: MedlinePlus