May 13, 1992 - Use of mutant insulin receptors in structure-function studies of insulins. Lauge Schaffer,* Jan Markussen, Thomas Kjeldsen. Asser S. Andersen ...
Biochemical Society Transactions
Use of mutant insulin receptors in structure-function studies of insulins Lauge Schaffer,* Jan Markussen, Thomas Kjeldsen. Asser S. Andersen, Finn C. Wiberg and Howard Tagert Novo Nordisk NS, Bagsvaerd, Denmark and +University of Chicago, USA. 842
Since the publication of the sequence of the human insulin receptor, hIR, in 1985 much work has been done to increase our understanding of the molecular mechanism of insulin action. Our interest has been mainly focused on the first step in the process, i.e. binding of insulin to the receptor. The insulin receptor is a membrane glycoprotein which is synthesized from a single-chain precursor. The processing of the receptor involves dimerization of the proreceptors, cleavage of the proreceptors to a disulphide-bridged a& structure and cell-specific modification of the carbohydrate. The a-subunits (molecular weight 125 kDa) are completely extracellular, whereas the P-subunits (molecular weight 90 kDa) contain a hydrophobic a-helix typical of a membrane spanning domain. Crosslinking studies indicate that insulin binds to the a-subunit. The binding of insulin is thought to induce a conformational change in the receptor which is somehow transmitted through the cell membrane, giving rise to autophosphorylation of a number of tyrosine residues and activation of the ability of the receptor to act as a kinase towards other substrates. This in turn gives rise to a cascade of effects on transport and metabolism of glucose and other metabolic substrates. In our efforts to gain an understanding of the binding process we are pursuing two strategies. The first is to express the soluble extracellular domain of the insulin receptor, SIR, in quantities for crystallization and X-ray structure determination, and for characterization by standard chemical techniques. The other is to locate the binding site by construction of chimeric receptors (Fig. 1). Soluble insulin receptor, SIR, has been produced in milligram quantities and purified to > 98% purity by affinity chromatography on insulinagarose [ 11. The affinity purification procedure has been developed specifically for this purpose since existing methods all suffered from too low capacity. W e coupled insulin to the activated agarose through the R1-amino group as this is located at the oppo-
site end of the molecule from the amino acids that are important for the receptor binding. The resulting affinity resin had a capacity of several milligrams of receptor per millilitre of gel. The affinity purified SIR was characterizsed by SDS-PAGE and quantitative amino acid analysis, and was found to have two identical, non-interacting binding sites for insulin. This is in contrast with the holoreceptor which usually displays curvilinear Scatchard plots. Crystallization experiments with SIR are in progress, but so far the crystals obtained have been too small for X-ray studies. The chimeric receptor approach is based on the close similarity between insulin and insulin-like growth factor 1 (IGF1) and between their receptors. The two hormones bind, albeit weakly, to each other’s receptors, and the homology between the receptors (including the conservation of most of the cysteines) suggest an overall similarity in the threedimensional organization. Exchange of the Nterminal 294 amino acids of the insulin receptor a-subunit with the corresponding sequence from the IGFl receptor, hIGFlR, resulted in a receptor that displayed high-affinity binding of IGFl instead of insulin, indicating that the binding site mainly resides in this region [Z]. Further subdivision of this region gave the surprising result that the region conferring insulin specificity on the insulin receptor and the region conferring IGFl specificity on the IGFl receptor appear to be located in different parts of the receptor [ 3 ] .Thus, an hIGFlR with the N-terminal 68 amino acids from the insulin receptor, hIGF1R 1-68, bound both hormones with high affinity, whereas the insulin receptor could be made to bind both hormones by inserting the sequence 191-290 from the hIGF1R. Further subdivision of the 1-68 region of the insulin receptor has indicated the importance of a few amino acids in the region 38-68 [4]. It appears that although insulin and Fig. I
Schematic diagram of some of the receptors used Abbreviations used: hIK, human insulin receptor; SIR, soluble insulin receptor; IGF 1, insulin-like growth factor 1; hlGF1K. human insulin-like growth factor 1 receptor. *Correspondence to: Lauge Schaffer, Biophysical Chemistry Department? Insulin Research, Novo Nordisk A/S, Novo Alle, DK-2880 Hagsvaerd, Denmark.
Volume 20
I
hlR
SIR
hlGFlR
hlGFlR 1-68
Sequence-Specific Peptides
Fig. 2 Schematic view of the binding pockets of three of the receptors with their preferred ligands
W hlR
+ Inwlln
hlOFlR + IOFI
hlOFlR 1-08 + T l
better than that of either native hormone. A schematic view of these results is presented in Fig.
2. In conclusion, a full understanding of the binding of insulin to the insulin receptor will probably not be reached until structural data from crystals of the insulin receptor or an insulin-binding fragment thereof are available. Meanwhile, however, there is a lot of information available from studies of receptor mutations and insulin analogues. 1. Markussen, J., Halstrom, Wiberg, I;. C. & Schaffer, 1,. (1991)J. Hiol. Chem. 266, 18814-18818 2. Andersen. A. S., Kjeldsen, T., Wiberg, I;. C., Christensen, 1’. M., Kasmussen, J. S.. Norris, K., Moller, K. H. & Mdler, N. P. H. (1990) Biochemistry 29,7363-7366 3. Kjeldsen, T., Andersen, A. S.. Wiberg, I;. C., Kasmussen, J. S., Schaffer, L., Balschmidt, P., Moiler, K. H. & Moiler, N. 1’. H. (1991) Proc. Natl. Acad. Sci. U S A . 88,4404-4408 4. Andersen. A. S., Kjeldsen, T., Wiberg, I;. C., Vising, H., Schaffer, I,.? Kasmussen, J. S.. De Meyts, P. & M ~ i l e r , N. P. H. (1992) J. Biol. Chem., 267, 13681-13686 5. Cara, J. F.,Mirmira, K.G., Nakagawa, S. H. & Tager, H. S. ( 1990)J. Biol. Chem. 265, 17820- 17825 J.?
IGFl may have a common binding site, the mode of binding may differ somewhat between the two hormones. This is in agreement with the finding that the C-domain of IGFl (which is not present in insulin) is important for binding to the hIGF1R. To confirm this, we have examined the binding of an insulin analogue, T1, which has a sequence derived from the C-domain of IGFl attached to the framework of the insulin molecule [ S ] . Whereas insulin binds approximately 1000-fold less than IGFl to the hIGFlR, T 1 has an affinity for the hIGF1R which is only approximately 10-fold lower than that of IGFl itself. As might be predicted for a molecule which has both insulin and IGFl properties, its binding to the chimeric receptors, which show high affinity binding to both insulin and IGF1, is indeed
Received 13 May 1992
Identification of antibodies to bovine growth hormone binding to defined, continuous epitopes within the protein James Beattie Hannah Research Institute, Ayr KA6 SHL, Scotland
Introduction
using antibodies directed to these sites. In this way
Growth hormone (GH) is a 22 kDa polypeptide secreted by the anterior pituitary which acts on a variety of target tissues to produce diverse biological effects [ 11. Previously, our laboratory has demonstrated that it is possible to neutralize the activity of GH by raising anti-sera to the hormone and subsequently passively immunizing rats with anti-GH antibodies [2, 31. There is evidence that the different biological and chemical activities of GH (e.g. insulin-like activity, dimerization domain) reside in discrete parts of the molecule [4, 51. If this is the case, then it may be possible to refine our previous studies and selectively inhibit the activities of the hormone by
it may be possible to alter the endocrine profile of
Abbreviations used: GH, growth hormone; rbGH, recombinant bovine growth hormone.
the hormone such that one particular function of GH may be enhanced over another. As a prelude to this it was important to identify the predominant immunogenic regions on the molecule in order to determine whether it would be feasible to produce these antibodies and indeed whether these epitopes coincided with putatively biologically active domains. T o that end we used the technique of multiple pin peptide synthesis first described by Geysen and co-workers [6] to identify the continuous epitopes recognized by five polyclonal anti-sera to bovine GH. We present this data and discuss the results obtained in relation to solved secondary and tertiary structure of GH and the possible biological significance of these epitope-specific antibodies.
I992
843
