(Received for publication, March 24, 1997)
From the Departments of Medicine and ** Physiology and
Biophysics, State University of New York at Stony Brook, Stony Brook,
New York 11794 and
Insulin Research, Novo Nordisk A/S,
2880-Bagsvaerd, Denmark
Insulin and insulin-like growth
factor 1 (IGF-1) are peptides that share nearly 50% sequence homology.
However, although their cognate receptors also exhibit significant
overall sequence homology, the affinity of each peptide for the
non-cognate receptor is 2-3 orders of magnitude lower than for the
cognate receptor. The molecular basis for this discrimination is
unclear, as are the molecular mechanisms underlying ligand binding. We
have recently identified a major ligand binding site of the insulin
receptor by alanine scannning mutagenesis. These studies revealed that
a number of amino acids critical for insulin binding are conserved in
the IGF-1 receptor, suggesting that they may play a role in ligand binding. We therefore performed alanine mutagenesis of these amino acids to determine whether this is the case. cDNAs encoding
alanine-substituted secreted recombinant IGF-1 receptors were expressed
in 293 EBNA cells, and the ligand binding properties of the expressed
proteins were evaluated. Mutation of Phe701 resulted in a
receptor with undetectable IGF-1 binding; alanine substitution of the
corresponding amino acid of the insulin receptor, Phe714,
produces a 140-fold reduction in affinity for insulin. Mutation of
Asp8, Asn11, Phe58, Phe692,
Glu693, His697, and Asn698 produces a
3.5-6-fold reduction in affinity for IGF-1. In contrast, alanine
mutation of the corresponding amino acids of the insulin receptor with
the exception of Asp12 produces reductions in affinity that are
50-fold or greater. The affinity of insulin for these mutants relative
to wild type receptor was similar to that of their relative affinity
for IGF-1 with two exceptions; the IC50 values for
insulin binding to the mutants of Arg10, which has normal
affinity for IGF-1, and His697, which has a 6-fold reduction in
affinity for IGF-1, were both at least 2 orders of magnitude greater
than for wild type receptor. The Kd values for
insulin of the corresponding alanine mutants of the insulin receptor,
Arg14 and His710, are 2-3 orders of magnitude
greater than for wild type receptor. However, in contrast, the
relative affinity of des(25-30)[PheB25
-carboxamide]insulin for these IGF-1 receptor mutants is reduced only 4- and 50-fold, respectively.
Insulin and insulin-like growth factor 1 are circulating serum peptides that share nearly 50% sequence homology (for a review, see Ref. 1). Conservation of the predicted major secondary structural elements of both peptides suggests that their tertiary structures are similar (2). Crystallographic and solution NMR structural studies have shown this to be the case (3, 4). Their cognate receptors also exhibit significant overall sequence homology (1). However, despite the overall homology of receptors and ligands, insulin and IGF-11 only bind weakly to each other's receptors; the affinity of each peptide for the non-cognate receptor is at least 3 orders of magnitude lower than for the cognate receptor (5). The molecular basis for this discrimination is at present unclear, as are the molecular mechanisms underlying ligand binding.
As a consequence of the availability of a large number of naturally occurring and chemically and biosynthetically modified analogs of insulin, an extensive body of information regarding its receptor binding determinants has accumulated. A current consensus is that A2 isoleucine, A3 valine, B12 valine, B24 and B25 phenylalanine, A19 tyrosine, A21 asparagine, and the partially buried residues A16 and B15 leucine are the major determinants of the receptor binding site, with A8 threonine, B9 serine, B10 histidine, B13 glutamate, and B16 tyrosine making minor contributions.2 This forms a patch on the surface of the molecule overlapping its dimerization surface. More recent studies also suggest that a small patch formed from the residues A13 and B17 on the hexamerization surface of insulin may represent a topologically distinct receptor binding site located on the opposite side of the molecule (6). Much less information is available with regard to the receptor binding determinants of IGF-1. Studies by Bayne and Cascieri (7, 8) implicate a role of the C region of the molecule in receptor binding; specifically, tyrosine 31 appears to be essential for high affinity binding (8). In addition, tyrosines 24 and 60, corresponding to B25 phenylalanine and A19 tyrosine of insulin, respectively, both essential for high affinity binding of insulin, play important roles in the receptor binding of IGF-1 (8).
The ligand binding determinants of both receptors have also been the
subject of study. Studies with chimeric receptors have provided
insights into the regions of the molecules that appear to confer
specificity. A consensus of such studies is that determinants of
insulin specificity appear to reside between amino acids 1 and 68 and
between amino acids 450 and 524 of the insulin receptor (5, 9, 10), and
determinants of IGF-1 specificity appear to reside between amino acids
190 and 290 of the IGF-1 receptor (5, 11). Affinity labeling studies of
the insulin receptor confirm the involvement of its N terminus in
ligand binding (12, 13) and also implicate regions between amino acids
430 and 488 (14) and 704 and 718 of the subunit (15).
Recent studies from this laboratory, utilizing alanine scanning
mutagenesis, have identified and characterized a major ligand binding
site of the insulin receptor subunit (16, 17). This appears to be
composed of an N-terminal subdomain between amino acids 1 and 120 (16)
and a C-terminal subdomain between amino acids 704 and 716 (17). The
N-terminal subdomain is composed of four discontinuous polypeptide
segments widely separated in the primary sequence, composed of 14 amino
acids. In contrast, the C-terminal subdomain is composed of three
discontinuous segments composed of 12 amino acids. Insulin analog
binding studies of these mutants reveal that the free energy
contributions of these amino acids can fully account for the free
energy of binding of insulin to the secreted recombinant receptor (18).
It thus appears that the major determinants of insulin binding have
been identified.
In contrast, it is far from clear that all of the determinants of ligand binding by the IGF-1 receptor have been identified. As noted above, IGF-1 binds to the insulin receptor with very low but measurable affinity (5). This suggests that both receptors may have some common ligand binding determinants in addition to their specificity determinants. This notion receives further support from our scanning mutagenesis studies of the insulin receptor (16, 17); many of the amino acids critical for insulin binding are conserved in the IGF-1 receptor, suggesting that there may be evolutionary conservation of function. In the present study we have, therefore, investigated this possibility by examining the ligand binding properties of alanine mutants of conserved amino acids in the IGF-1 receptor that correspond to those in the insulin receptor that are critical for insulin binding.
All molecular biological procedures,
including agarose gel electrophoresis, restriction enzyme digestion,
ligation, bacterial transformation, and DNA sequencing, were performed
by standard methods (19). All oligonucleotides were purchased from
DNAgency (Malvern, PA). Restriction and DNA modifying enzymes were from New England Biolabs (Beverly, MA). Recombinant human IGF-1 was a gift
from Dr. Ron Chance (Lilly Research Labs, Indianapolis, IN). Human
insulin and
des(B25-30)[HisA8,AspB10,TyrB25
-carboxamide]insulin (X92) were synthesized as described previously (6). Des(B25-30)[PheB25
-carboxamide]insulin (20) was
kindly provided by Dr. Satoe Nakagawa (University of Chicago).
HPLC-purified monoiodinated IGF-1 (125I-IGF-1), insulin
radioiodinated at A14 tyrosine of the A chain ([125I-TyrA14]insulin) and carrier-free
125-iodine (IMS 30) were from Amersham Corp. Protease inhibitors were
from Boehringer Mannheim. 293 EBNA cells were purchased from In
Vitrogen (San Diego, CA). Medium for tissue culture was from Mediatech
(Herndon, VA), and serum was from Sigma. Anti-insulin receptor
monoclonal antibody 18-44 (21) and anti-IGF-1 receptor monoclonal
antibody 24-31 (22) were generously provided by Drs. M. Soos and K. Siddle (University of Cambridge, United Kingdom). A polyclonal antibody
directed toward the
subunit of the IGF-1 receptor (amino acids
31-50 according to Ref. 23) was from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA), and affinity-purified goat antimouse IgG was from
Pierce. cDNAs encoding the entire coding region of the IGF-1 receptor (23) were obtained by screening a human kidney cDNA library (a gift of Dr. G. I. Bell, Howard Hughes Medical Institute, University of Chicago) with oligonucleotide probes and by reverse transcriptase-polymerase chain reaction (24) of mRNA isolated from
HepG2 cells using oligonucleotide primers based on the sequence of
Ullrich et al. (23). A full-length cDNA was
reconstructed in the phagemid pTZ18U. The sequence was identical to
that reported by Ullrich et al. (23). A cDNA encoding a
recombinant secreted receptor was obtained by inserting a stop codon in
the position of the codon encoding the first amino acid of the
transmembrane domain (24). The amplified DNA was sequenced in its
entirety to exclude the possibility of nucleotide misincorporation.
Oligonucleotide-directed mutagenesis was performed by the method of Kunkel (25). A cDNA encoding a recombinant secreted IGF-1 receptor subcloned into the phagemid pTZ18U was used as the template for mutagenesis. Where possible, restriction sites were deleted or introduced with the specific mutation to permit enrichment of mutants by restriction selection or purification and to facilitate the screening of mutants. The IGF-1 receptor amino acids mutated and the corresponding insulin receptor residues are shown in Table I; the numbering of amino acids for the insulin receptor is according to Refs. 26 and 27, and the numbering for the IGF-1 receptor is according to Ref. 23. Insulin receptor mutants were as described previously (16, 17).
|
The mutant IGF-1 receptor cDNAs encoding a secreted receptor extracellular domain were subcloned into the plasmid pCDE3 (16, 17) for expression. These constructs were transiently expressed in 293 EBNA cells (an adenovirus-transformed human kidney cell line expressing Epstein-Barr virus nuclear antigen) by transfection with 2 µg of miniprep DNA using the commercially available lipofection reagent Lipofectamine (Life Technologies, Inc.) according to the manufacturer's directions. For analysis of transient expression, media and cells were harvested 1 week after transfection. Conditioned medium was concentrated prior to assay using Centriprep 100 centrifugal concentrators (Amicon, Beverly, MA). Detergent lysates of cells were prepared as described previously.
We chose to utilize the extracellular domain for these experiments, since it is expressed in large amounts by this expression system and eliminates interference due to the substantial background of endogenous IGF-1 receptors in the 293 EBNA cells.3
Insulin receptor mutant cDNAs were also transiently expressed in 293 EBNA cells as described previously (16, 17).
Radioiodination of Insulin AnalogsFor ligand binding studies, the insulin analog, X92, was iodinated with 125-iodine by the lactoperoxidase method. The monoiodo-A14 tyrosine isomer was separated from the iodination reactions by reverse phase HPLC using a C18 column (28).
Receptor Binding AssaysSoluble IGF-1 receptor binding assays were performed using a microtiter plate antibody capture assay. Microtiter plates (Immulon 4, Dynal Inc., Lake Success, NY) were incubated with affinity-purified goat antimouse IgG (50 µl/well of 20 µg/ml solution in 0.2 M sodium carbonate, pH 9.4) for 2 h at room temperature. After washing with phosphate-buffered saline and blocking for 15 min with 250 µl of SuperblockTM (Pierce), wells were incubated overnight at 4 °C with a 1:100 dilution in SuperblockTM of crude ascites of anti-IGF-1 receptor monoclonal antibody 24-31 (22). This antibody has been demonstrated to have minimal effects on the affinity of the receptor for IGF-1. After washing with phosphate-buffered saline, wells were incubated for 4 h at 4 °C with soluble receptor, diluted to give 10-20% 125I-IGF-1 binding in the absence of added unlabeled IGF-1 or 10-20% 125I-X92 in the absence of unlabeled insulin under assay conditions. After washing with wash buffer (0.15 M sodium chloride, 20 mM Hepes, pH 7.8, 0.1% (w/v) bovine serum albumin, 0.025% (v/v) Triton X-100, and 0.02% (w/v) sodium azide), wells were incubated for 48 h at 4 °C with 125I-IGF-1 (12 pM) or 125I-X92 (12 pM) and varying concentrations of unlabeled IGF-1 or insulin, in 100 µl of binding buffer (137.5 mM sodium chloride, 10 mM magnesium sulfate, 20 mM Hepes, pH 7.8, 0.5% (w/v) bovine serum albumin, 0.025% (v/v) Triton X-100, and 0.02% (w/v) sodium azide). To terminate the assay, wells were aspirated and washed three times with 300 µl of ice-cold wash buffer and then counted. Insulin binding was determined as described previously (16, 17).
Binding data were analyzed by the LIGAND program (29) to obtain the Kd of the expressed protein. Transfection and binding assays were repeated at least once to confirm the Kd of each mutant. Each result is the mean of two experiments.
ImmunoblottingImmunoblotting of the IGF-1 receptor subunit in conditioned medium and detergent lysates of transfected
cells was performed according to standard methods (30) using a
polyclonal antibody directed toward the IGF-1 receptor
subunit
(Santa Cruz Biotechnology). Blots were visualized by enhanced
chemiluminescence (ECL, Amersham).
Mutant IGF-1 receptor cDNAs were transiently expressed in 293 EBNA cells. To evaluate expression and post-translational processing of
the receptor, conditioned medium from transfected cells and detergent
lysates of the cells were analyzed by Western blotting with an antibody
directed toward the N terminus of the subunit of the IGF-1
receptor. In conditioned medium, a Mr 135,000 protein representing IGF-1 receptor
subunit was detectable for all
transfections (see Fig. 1 for a representative
immunoblot) with the exception of those with cDNAs encoding alanine
mutations of Leu32 and Glu38 (data not shown);
the epitope for the antibody used for these blots is directed toward
amino acids 31-50 of the receptor
subunit, and hence it would not
be expected to recognize these mutants. In detergent lysates of cells,
a protein of Mr 160,000, corresponding to the
predicted mobility of the secreted receptor precursor (23), was
detected in all transfections with the exception of the
Leu32 and Glu38 mutants. The
Mr 135,000 protein observed in the lysates was
present in all transfections and probably represents a protein
cross-reacting with the secondary antibody, since we have seen a
protein of the same mobility in lysates of transfected cells blotted
with anti-insulin receptor antibodies (17). For each transfection, the
amount of detectable
subunit relative to precursor was similar (see Fig. 1 for a representative example), suggesting that the mutations did
not cause any major perturbation of post-translational processing of
the receptor. While the relationship between post-translational processing and protein folding has not been systematically examined for
the IGF-1 receptor, studies with the closely related insulin receptor
(31, 32) would suggest that folding of the precursor of the IGF-1
receptor into its native structure is a prerequisite for its export
from the endoplasmic reticulum and completion of post-translational
processing. Thus, it is unlikely that these mutations result in major
perturbations of receptor structure.
Equilibrium binding studies were performed on conditioned media to
characterize IGF-1 binding to wild type and mutant receptors. As we
have previously described (16, 17) for insulin binding to recombinant
insulin receptor extracellular domain expressed in 293 EBNA cells,
IGF-1 binding to recombinant secreted wild type IGF-1 receptor
displayed simple kinetics, with a linear Scatchard plot (data not
shown). Analysis with the LIGAND program (29) indicated a single
population of binding sites with a Kd of 0.86 ± 0.04 × 1010 M (mean ± S.E.,
n = 4). Since studies utilizing alanine scanning mutagenesis have demonstrated that meaningful changes in affinity, produced by a single alanine substitution, range from 2- to 100-fold (33), in the experiments described below, we regarded any mutant with a
greater than 2-fold increase in Kd, i.e.
Kd greater than 1.8 × 10
10
M, as exhibiting a significant disruption of IGF-1-receptor
interactions.
The results of our analyses of the alanine mutants of the IGF-1 secreted receptor are shown in Table II. Data are expressed as a ratio of the dissociation constant (determined by the LIGAND program (29) using a single binding site model) of the mutant to that of wild type receptor; results for the corresponding insulin receptor mutants are shown for comparison. It should be noted that values for the mutants in the N-terminal subdomain of the insulin receptor differ from those that we originally reported (16), reflecting the modification in the assay conditions. In the N-terminal subdomain of the IGF-1 receptor, mutations of 3 amino acids, aspartate 8, asparagine 11, and phenylalanine 58, to alanine have small but significant effects on the affinity of the receptor for IGF-1; each mutation results in a 3-4-fold reduction in affinity (Table II, N-terminal subdomain). All other mutations were without significant effect. In the C-terminal subdomain, mutations appear to have similar, profound effects, with the exception of phenylalanine 701 to alanine; mutation of phenylalanine 692, glutamate 693, histidine 697, and asparagine 698 to alanine each produce a 3.5-6-fold reduction in affinity for IGF-1, and mutation of phenylalanine 701 produces a receptor with an affinity too low to be accurately determined (Table II, C-terminal subdomain). Immunoblotting (Fig. 1) revealed that this mutant is secreted into the medium in quantities comparable with the secretion of wild type receptor. Also, the amount of mutant receptor precursor detectable in the lysate, relative to the amount of mature receptor in conditioned medium, is similar to that observed for wild type receptor, indicating that post-translational processing appears to be normal. Thus, it is probable that the mutation has not perturbed the structure of the receptor in a major way and that its effects on ligand binding are due to direct perturbation of the IGF-1-receptor interaction. This mutation has an even greater effect on IGF-1 binding than the corresponding insulin receptor mutation on insulin binding (Table II). None of the other mutations in either domain produce effects on affinity of ligand binding comparable with those of the corresponding insulin receptor mutation (Table II).
|
To characterize the insulin binding properties of these mutants, the
concentration dependence of insulin displacement of
125I-IGF-1 and 125I-X92 were studied. The
latter, which is a superanalog of insulin (affinity for the insulin
receptor 200-fold greater than that of insulin (18)), was used because
the tracer binding of 125I-insulin was too low to permit
accurate determination of insulin binding (data not shown). The
IC50 for insulin displacement of IGF-1 from the wild type
receptor was 3.3 ± 0.05 × 107 M
(mean ± S.E., n = 4). The IC50 for
insulin displacement of X92 was 5.1 ± 0.12 × 10
7 M (mean ± S.E., n = 4). Since the IC50 values for displacement of labeled IGF-1
and X92 by unlabeled IGF-1 from wild type IGF-1 receptor were 1.5 ± 0.03 × 10
10 M (mean ± S.E.,
n = 4) and 3.1 ± 0.06 × 10
10
M (mean ± S.E., n = 4), respectively,
the relative affinity of insulin for the receptor is 1000-2000-fold
lower than that of IGF-1, in good agreement with the findings of
Kjeldsen et al. (5).
Results for the mutant receptors are shown in Table III
and are expressed as ratios of the IC50 for the mutant to
the IC50 of the wild type receptor. In the N-terminal
subdomain, only one mutation, arginine 10 to alanine, had a significant
effect on insulin displacement of IGF-1 and X92. The IC50
for IGF-1 was increased at least 300-fold; insulin failed to produce
50% displacement at a concentration of 104
M. The IC50 for X92 was increased 118-fold.
These results are in reasonable agreement with the effect of the
corresponding mutation in the insulin receptor, which results in a
700-fold decrease in affinity for insulin (18), but they contrast
markedly with its negligible effect on the affinity for IGF-1 (Table
II, N-terminal subdomain).
|
In the C-terminal subdomain, similar results were obtained with mutation of histidine 697 to alanine. The IC50 values for insulin displacement of IGF-1 and X92 were increased more than 300- and 200-fold, respectively (Table III, C-terminal subdomain) despite only a 5.6-fold increase in Kd for IGF-1 (Table II, C-terminal subdomain). Again, these results are reasonably consistent with the effects of the corresponding mutation in the insulin receptor, which produces a 1000-fold decrease in affinity for insulin (18). Mutation of valine 702 to alanine, while being without effect on affinity for IGF-1, increased the IC50 for insulin 4-fold with both tracers, an effect that was half that of the corresponding mutation in the insulin receptor on its affinity for insulin (Tables II and III, C-terminal subdomain). Mutation of glutamate 693 and asparagine 698 to alanine produced reductions in relative affinity for insulin with both tracers that were comparable with the reductions in affinity for IGF-1 (4-fold versus 5-fold and 6-fold versus 4-fold, respectively; see Tables II and III, C-terminal subdomain). In contrast, the corresponding mutations in the insulin receptor, glutamate 706 and asparagine 711 to alanine, produced much more profound decreases in affinity for insulin (Table II, C-terminal subdomain). Mutation of phenylalanine 692 to alanine produced a 16-fold decrease in relative affinity for insulin with both tracers (Table III, C-terminal subdomain). This is some 4-fold greater than its effect on affinity for IGF-1 but not comparable with the effect of the corresponding mutation in the insulin receptor, phenylalanine 705 to alanine, on affinity for insulin (Table II, C-terminal subdomain). Mutation of leucine 696 to alanine had no effect on IC50 for insulin with either tracer (Table III, C-terminal subdomain) despite the finding that the comparable mutation in the insulin receptor produces a 151-fold decrease in affinity for insulin (Table II, C-terminal subdomain). It was not possible to characterize insulin displacement of IGF-1 from the valine 702 to alanine mutant receptor because of its extremely low tracer IGF-1 binding.
Recent analog binding studies of insulin receptor mutants from this
laboratory have shown that the compromised affinity of the
Arg14 and His710 to alanine mutants is
dependent on the presence of the C-terminal pentapeptide of the B chain
of the insulin molecule (18). Thus, to determine whether the same
mechanism is involved in the compromised affinity of the
Arg10 and His697 to alanine mutants of the
IGF-1 receptor for insulin, competition studies with
des(25-30)[PheB25 -carboxamide]insulin were performed
with these mutant receptors. The IC50 for displacement of
125I-IGF-1 from wild type IGF-1 receptor by
des(25-30)[PheB25
-carboxamide]insulin was 0.9 × 10
7 M (Table IV). The
IC50 values for the Arg10 and
His697 to alanine mutants were 3.4 × 10
7 M and 4.5 × 10
6
M, respectively (Table IV), which are in marked contrast
with the >300-fold increase in IC50 observed for
insulin with these mutants (Table III).
|
In the present study we have characterized the ligand binding
properties of alanine mutants of conserved amino acids of the IGF-1
receptor corresponding to amino acids in the major ligand binding site
of the insulin receptor. We identified a number of mutants that
compromised IGF-1 binding, but only one, phenylalanine 701 to alanine,
produced a decrease in affinity similar to or greater than that
observed for the corresponding mutation in the insulin receptor. We
also examined insulin displacement of tracer IGF-1 from these mutant
receptors. Two mutants, arginine 10 and histidine 697, behaved in a
strikingly similar manner to the corresponding mutations of the insulin
receptor, arginine 14 and histidine 710; their affinity for insulin was
unmeasurably low, but in contrast their relative affinities for
des(25-30)[PheB25 -carboxamide]insulin were reduced
only 3- and 50-fold, respectively. In the remainder of the mutants,
either the displacement of IGF-1 by insulin was unaffected by the
mutation or the observed decrease in affinity was similar to the
decrease in affinity for IGF-1 produced by the mutation.
This mutational analysis of conserved amino acids in the subdomains of
the IGF-1 receptor corresponding to the major ligand binding site of
the insulin receptor clearly demonstrates that certain of these
residues play a role in ligand binding. Although the limitations of the
sensitivity of the receptor binding assay do not permit us to determine
the quantitative contribution of these domains of the IGF-1 receptor to
the overall free energy of the ligand-receptor interaction, it is
possible to provide some reasonable estimates by comparison of these
data with those we have recently obtained from analog binding studies
of insulin receptor mutants (18). Since the affinities of both
receptors for their cognate ligands are similar and the analysis of
binding data indicates the maximum amount of each receptor bound in the microtiter wells is very similar under assay conditions,3
the sensitivities of each assay must also be very similar. Our binding
studies of X92 to insulin receptor mutants suggest that the lower limit
for the accurate determination of affinity of the insulin receptor for
insulin, using this assay, is a 500-fold reduction (18). This would
therefore suggest that the Phe701 mutant of the IGF-1
receptor has an affinity for IGF-1 that is at least 500-fold lower than
that of the wild type receptor. Thus, the free energy contribution of
the interaction of this residue of the receptor with IGF-1, calculated
from the differences in binding free energy
(G)4 between wild type
receptor and the alanine-substituted mutant, may account for up to 25%
of the free energy of the ligand-receptor interaction (
3.4
versus 12.7 kcal/mol). Furthermore, since it has been shown
that, in the absence of perturbation of molecular structure and
intramolecular interactions, the free energy changes of protein-protein
interactions resulting from multiple mutations in a protein to binding
are the sum of those produced by the individual mutations (for a
review, see Ref. 34), the calculated free energy contributions of all
the residues whose mutation is disruptive of ligand binding can
potentially account for up to 75% of the free energy of the
ligand-receptor interaction (
8.8 versus 12.8 kcal/mol).
The role of these conserved residues in IGF-1-receptor interactions seems to be very different from the role of the corresponding amino acids in insulin-receptor interactions. With the exception of phenylalanine 701, mutation of the IGF-1 receptor residues corresponding to those that appear to be critical for insulin binding to the insulin receptor appears to have only relatively small effects on the affinity for IGF-1. Similarly, with the exception of the mutations of arginine 10 and histidine 697, they also only produce small effects on the affinity of the IGF-1 receptor for insulin, suggesting that insulin and IGF-1 use some common contacts to bind to the IGF-1 receptor and that those contacts differ from those that insulin utilizes to bind to the insulin receptor. One possible reason for this difference in interaction is the presence of the putative IGF-1 specificity element between amino acids 190 and 290 of the IGF-1 receptor (5, 11). Studies with chimeric receptors indicate that its substitution for the corresponding region of the insulin receptor can influence the affinity of the insulin receptor for insulin (5). It is possible that this region of the IGF-1 receptor contains structural elements that alter insulin-receptor interactions. However, preliminary studies of the binding properties of the C-terminal alanine mutants of the insulin receptor in an insulin-IGF-1 receptor chimera, in which amino acids 190-290 of the IGF-1 receptor have been substituted for the corresponding region of the insulin receptor, indicate that they retain the compromised affinities for insulin exhibited by the corresponding insulin receptor mutants.3
Interaction of insulin with arginine 10 and histidine 697 of the IGF-1 receptor and arginine 14 and histidine 710 of the insulin receptor appears to be an obligatory step in its binding to both receptors. The decrease in affinity of the IGF-1 receptor mutants for insulin, like that of the insulin receptor mutants, is dependent on the presence of the C-terminal region of the B chain of the insulin molecule; despentapeptide insulin analogs appear to have an affinity for these mutants, which are at maximum 60-fold lower than for the wild type receptor, although the affinity of insulin for them is unmeasurably low (Table IV and Ref. 18). The effects of these mutations on insulin-insulin receptor and insulin-IGF-1 receptor interactions appear to mirror the effects of substitutions of B25 phenylalanine in the insulin molecule on the affinity of insulin-insulin receptor interactions (20, 35). Nakagawa and Tager (20, 35) have demonstrated that substitutions in this position compromise the affinity of insulin for the receptor only in the full-length insulin molecule and have suggested that the interaction of this side chain with the receptor initiates a conformational change in the C-terminal main chain that is essential for high affinity binding. This is supported by the finding that insulin analogs that are cross-linked between A1 glycine and B29 lysine, in which movement of the C terminus of the B chain is constrained, exhibit varying decreases in biological activity, depending on the length and flexibility of the cross-link (36, 37), despite the lack of major perturbation of their structure. Solution NMR studies of the structure of B24 glycine insulin provide further insights into possible structural mechanisms underlying these findings (38). In this analog, the C terminus of the B chain is unfolded and disordered, exposing residues in the core of the molecule, in particular A3 valine, which has been shown to be essential for receptor binding (39). It is, therefore, possible that interaction of B25 phenylalanine with the receptor produces a similar change in conformation of the insulin molecule, allowing completion of the interaction with the receptor. Our studies suggest that, for insulin-insulin receptor and insulin-IGF-1 receptor interactions to take place, B25 phenylalanine of insulin must interact with the arginine and histidine residues of either receptor to initiate a conformational change in the C terminus of the B chain essential for receptor interactions. Further functional and high resolution structural studies will be necessary to confirm this hypothesis.
In summary, we have demonstrated that a limited number of conserved residues in the IGF-1 receptor, corresponding to amino acids in the insulin receptor critical for ligand binding, play a role in IGF-1 binding. While, with the exception of phenylalanine 701, they only make a small contribution to the free energy of binding, these findings are suggestive of involvement of these regions of the receptor in ligand binding. More extensive mutational analyses of these regions will be necessary to fully evaluate their functional role in IGF-1 binding by the receptor.
We are grateful to Drs. K. Siddle and M. Soos (University of Cambridge, Cambridge, UK) for providing monoclonal antibodies to the insulin and IGF-1 receptors.