Insulin is thought to have two distinct receptor binding surfaces, with one site encompassing at least the residues G1, E4, Q5, and N21 of the insulin A-chain and the other site being made up of residues V12, Y16, F24, F25, and Y26 of the B-chain; other residues apart from those listed above are almost certainly involved in this interaction . Binding of insulin to the native, dimeric IR is characterized by curvilinear Scatchard plots and the phenomenon of negative cooperativity, both interpreted as showing the presence of two states of the receptor, one of high-affinity insulin binding and the other of low affinity. Half receptors, i.e., aP-monomers, formed by mild reduction of the dimeric (aP)2 receptor, do not show these phenomena and bind insulin with only low affinity , as does the expressed recombinant IR ectodomain. High-affinity binding is restored when the IR is truncated below the transmembrane domain or when a dimerization moiety such as the IgG-y domain or a leucine-zipper segment is fused to the C terminus of the expressed ectodomain . Clearly, not only is the dimeric state of the IR essential for high-affinity binding, but also the relative disposition of the two monomers in the
dimer is critical; it seems that the C termini of the two a-chains must be in close proximity for effective high-affinity binding to insulin.
The isolated a-chain of the IR binds insulin, although with an affinity lower than that of the wild-type receptor, and thus appears to have all the insulin-binding determinants of the receptor. By alanine-scanning mutagenesis, studies of chimeric receptors, and direct cross-linking of insulin to the IR, binding determinants on the receptor have been located in the L1 domain, in the Cys-rich region, and near the N terminus of the L2 domain. Additionally, important determinants are also found close to the C terminus of the a-chain . For both the IR and the IGF-1R, the L1-Cys-rich-L2 fragment does not bind the cognate ligand, despite the presence of many of the binding determinants and despite the fact that the horseshoe-like structure reveals a cavity of sufficient dimensions to partly encircle the ligand. Thus, the X-ray structure of this fragment does not yield details of the complete ligand-binding site. However, ligand binding can be restored by adding to the C terminus of this fragment a 16-residue peptide that includes the binding determinants identified at the C terminus of the a-chain. The peptide can be attached directly to the L1-Cys-rich-L2 fragment or by using linkers of varying length. This suggests that the mode of attachment is perhaps not critical, a view supported by the observation that the addition of the free peptide to the fragment also restores ligand binding . The 16-residue peptide itself probably does not directly bind insulin, as
C-terminal fragments of the IR comprising two or three FnIII domains and including this 16-residue segment of the a-chain do not bind insulin .
Although each a-chain of the IR has all the determinants necessary for binding of insulin, high-affinity binding requires two a-chains held in appropriate juxtaposition by at least two disulfide bonds between these chains and by a single disulfide bond linking each a-chain to a P-chain that is itself anchored to the cell membrane. This suggests that two aP monomers are involved in binding a single insulin molecule, thus providing a cross-link between the chains in addition to the disulfide bonds. A significant observation is that the IR undergoes an obvious conformation change (Stokes radius from 9.1 to 7.5 nm) on binding insulin . The nature of this change is unclear, but it may in some way bring together the two kinase domains of the homodimer so that the activation loop of one is accessible to the active site of the other. Such a movement may be equivalent to that observed for other receptor molecules, such as the EGFR, that undergo dimerization following ligand binding, resulting in activation of the receptor tyrosine kinase. Full details of the binding of insulin to its receptor and the subsequent conformation change will probably not be known until crystal structures of the receptor with and without bound ligand are available. Activation of the IR kinase is accompanied by autophosphorylation at up to six tyrosines which both further activates the kinase and creates binding sites for signaling proteins, which in turn become phosphorylated and bind their downstream targets.
Was this article helpful?