Fibroblast Growth Factors (FGF's): Kinemages

Fibroblast Growth Factors

Their Receptors:

A Kinemage Collection

Larry P. Taylor, Ph. D. (retired)

Feedback appreciated; please send comments to:

Email: lpt

Molecular & Behavioral Neuroscience Institute

The University of Michigan

Ann Arbor, MI

This site is best viewed with

a screen resolution of 1280 x 1024.

Fibroblast Growth Factors (FGF's) represent a homologous family of at least 18 proteins with a beta trefoil (three-fold repeat of a four-stranded sheet assembly without extended alpha helix strands possessing a pseudo three-fold axis of symmetry) motif associated with cell proliferation and differentiation. They are a prime component of angiogenesis associated with organogenesis, tumor growth, and wound healing. Imbalances in FGF levels and/or mutations of the gene encoding for FGF or its receptors have been associated with numerous diseases, cancers and other pathological states. Although the FGF ligands share similar sequences and three-dimensional motif, there are subtle sequence-dependent backbone differences than can rationalize much of the observed FGF to FGF receptor binding behavior. FGF 1 is considered the universal FGF ligand since it binds with high affinity to all four of the major human FGF receptors. Each of the FGF ligands is summarized here.

A sub-family of the 23 homologous FGF related proteins, the FHF family (FGF 12 (FHF 1)) of protein ligands was recently removed (lack of functional similarity) from the FGF family. This change in "status" of the FHF sub-family of FGF ligands means that the FGF protein family is now composed of at least 18 homologous proteins that bind FGF receptors.

FGF molecules function by interacting with FGF receptors (FGFR's). These receptors typically contain three Ig-like binding domains (with domains D2 and D3 being most involved in FGF-FGF receptor binding interactions), a short transmembrane spanning domain and a cytoplasmic component that possesses tyrosine kinase activity. Between receptor protein Domains D1 and D2 is a short span of acidic residues called the "Acid Box." A cartoon schematic of FGF Receptor Domains is shown below:

The D1 and the linker region between D1-D2 have much sequence variability, multiple splice variants, are structurally disordered (even in the presence of bound FGF ligand), and are considered not necessary for ligand binding ( FGF binds to FGFR even in the total absence of protein receptor domain D1). It is hypothesized that the "Acid Box" can participate in the regulation of FGF binding to FGFR's in that the multiple acidic sites of the "Acid Box" can mimic the negative potential surface of heparin-like compounds and thus, bind at the heparin-binding sites located on the surface of the D2 region of the FGF receptor without initiating a biological response (an auto-inhibition mechanism). A cartoon of this inhibition is shown below:

There are 4 major FGF receptors, but multiple splice variants, especially in exon III of the receptors (corresponding to FGF receptor protein domain D3) add to the complexity of the ligand-receptor binding event The third exon contains three parts: (IIIa, IIIb, and IIIc) and gene splicing events lead to D3 domain transcribed from the invariant IIIa portion of the gene followed by either b or c. This splice variation in the ligand binding domain defined by protein receptor domains D2 and D3 splits the four main receptors into a total of seven key human FGF receptors: FGFR 1b , FGFR 1c, FGFR 2b , FGFR 2c, FGFR 3b, FGFR 3c, and FGFR 4. The alternative splicing in D3 apparently is a tissue-specific process with isoform c associated with mesenchymal and isoform b predominate in epithelial cells.

An essential component of biological activity is the heparin mediated dimerization of the FGF receptor.

Both FGF ligands and receptors are characterized by descriptions of their beta sheets. (see nomenclature).

Since chemical structure and physiological function are intimately related, examination of FGF and FGFR molecular architectures (especially comparing similar molecules with different physiological responses) may provide a better understanding of how the FGF ligands interact with their receptors to provide a biological effect. Although static pictures and FGF and FGFR Sequence Alignments of a molecular family are helpful, dynamic images better facilitate delineation of molecular shapes and ligand-receptor interactions responsible for biological activity. So, this site uses a protein visualization tool, KiNG , to facilitate real-time on-line manipulation of kinemage renderings of FGF related molecules. Kinemages for the molecules discussed and the KiNG Manual (pdf) and the are also available as separate downloads for off-line use. These kinemages were prepared from crystal structure coordinates found in the Brookhaven Data Base (References).

The real-time visualization using KiNG of the structures on this site requires a java-enabled (JRE from Java) browser.