Available via license: CC BY 4.0
Content may be subject to copyright.
THE
JOURNAL
OF
BIOLOGICAL
CHEMISTRY
6
1987
by
The American Society for Biochemistry and
Molecular
Biology,
Ine.
Vol.
262,
No.
26,
Ienue of September
6,
pp.
12064-12058 1987
Printed
in
ir.~.~.
Role
of
the COOH-terminal B-chain Domain in
Insulin-Receptor Interactions
IDENTIFICATION OF PERTURBATIONS INVOLVING THE INSULIN MAINCHAIN*
(Received for
publication,
April
23,
1987)
Satoe
H.
Nakagawa and Howard
S.
TagerS
From the
Department
of
Biochemistry
and
Molecular
Biology,
The
University
of
Chicago, Chicago,
ZUinois
60637
Previous studies have suggested that the COOH-ter-
minal pentapeptide of the insulin B-chain can play
a
negative role in ligand-receptor interactions involving
insulin analogs having amino acid replacements
at
po-
sition B26 (Nakagawa,
S.
H., and Tager, H.
S.
(1986)
J.
Biol.
Chern.
261,7332-7341).
We
undertook by the
current investigations to identify the molecular site in
insulin that induces this negative effect and to explore
further the importance of conformational changes that
might occur during insulin-receptor interactions. By
use of semisynthetic insulin analogs containing amino
acid replacements or deletions and of isolated canine
hepatocytes, we show here that
(a)
the markedly de-
creased affinity of receptor for insulin analogs in
which PheB2’
is
replaced by Ser
is
apparent for analogs
in which up to
3
residues of the insulin B-chain have
been deleted, but
is
progressively reversed in the cor-
responding des-tetrapeptide and des-pentapeptide
an-
alogs, and
(b)
unlike the case for deletion of TyrBae and
ThrBa7, replacement of residue TyPas or ThPa7 has
no
effect to reverse the decreased affinity of full length
analogs containing Ser for Phe substitutions
at
position
B26. Additional experiments demonstrated that intro-
duction of
a
cross-link between LysfBae and GlyPA1 of
insulin decreases the affinity of ligand-receptor inter-
actions whether or not PheBa6
is
replaced by
Ser.
We
conclude that the negative effect of the COOH-terminal
B-chain domain on insulin-receptor interactions
arises
in greatest
part
from the insulin mainchain near the
site of the TyPaa-Thfla7 peptide bond and that multiple
conformational perturbations
may
be necessary to in-
duce
a
high-affinity
state
of receptor-bound insulin.
While the interactions of ligands with plasma membrane
receptors are recognized to initate the processes by which
hormones and other effectors produce their biological actions,
little is known about the molecular events that may be in-
volved in ligand recognition, in the conferal of ligand-receptor
affinity or in the transduction of the signal resulting from
ligand-receptor interactions. Evidence for multiple
states
of
ligand-receptor conformations (conformations that may re-
flect intermolecular information transfer) has recently been
documented for the neutrophil formyl peptide receptor (1, 2)
*
These
studies
were
supported
by
Grants
DK18347 and
DK20595
from The
National
Institutes of Health.
The
costa
of
publication
of
this
article
were defrayed in part
by
the
payment
of
page
charges.
This
article
must therefore
be
hereby
marked
“advertisement”
in
accordance
with
18
U.S.C. Section
1734
solely
to
indicate
this
fact.
3
To
whom
correspondence
should
be
addressed Dept.
of
Biochem-
istry
and
Molecular
Biology, The University
of
Chicago,
920
East
58th St., Chicago,
IL
60637.
and for the
Torpedo
cholinergic receptor (3). In the case of
insulin, putative changes in ligand-receptor conformations
have been addressed by studies of the kinetics of ligand
association and dissociation
(4,
5),
by analysis of ligand-
receptor susceptibility to various agents (6,
7),
and by exam-
ination of a variety of chemically modified hormone analogs
(8).
Notwithstanding the fact that insulin is a globular peptide
hormone with relatively defined secondary and tertiary struc-
ture, the hormone is known to undertake multiple conforma-
tions. Notably,
(a)
the structures of molecules
l
and 2 of the
insulin dimer within the 2-zinc insulin hexamer differ some-
what in distant regions involving both the mainchains of the
NHz-terminal
A-
and B-chain domains and the sidechains in
the COOH-terminal B-chain domain (9-11),
(b)
the structure
of the NH2-terminal region of the B-chain in 4-zinc insulin
differs considerably from the structure of the same region in
2-zinc insulin (12, 13), and (c) the NH2-terminal domain of
the B-chain in des-pentapeptide insulin undertakes a confor-
mation quite different from that seen in either 2-zinc or
4-
zinc insulin (14-16). Most important, conformational restric-
tion in the insulin molecule has been shown in several ways
to decrease significantly affinity of receptor for ligand. First,
insulin cross-linked between residues LysrBZB and Gly”*’ by
the 2,7-diaminosuberoyl group shows only about 6% of the
normal affinity of insulin for receptor, although it exhibits no
major change in structure relative to uncross-linked insulin
(17,
18).
Second, insulin cross-linked between the same resi-
dues with the bis-adipimido group shows an equivalently
decreased binding affinity, although the uncross-linked bis-
acetimido derivative exhibits nearly the full potency of normal
insulin
(19).
Third, proinsulin (a natural analog cross-linked
between residues B30 and A1 by the connecting peptide) has
only 2% of normal affinity for the insulin receptor (20).
Fourth, the truncated proinsulin analogs des-(30-65)-proin-
sulin (21) and des-(23-65)-proinsulin (22) (while they may be
considerably strained) exhibit <0.1%
of
the affinity of insulin
for interaction with its receptor.
It
thus appears both that
insulin has the propensity for considerable conformational
change and that restrictions in the potential for such change
decrease considerably the affinity of insulin receptors for
ligand.
Our previous work on insulin structure-function relation-
ships and on conformational changes that may occur during
insulin-receptor interactions identified
(a)
the 100-fold de-
crease in ligand-receptor affinity that attends replacement of
PheB% by serine, leucine, or homophenylalanine,
(b)
the re-
versal of this decrease in affinity (by up to 40-fold) when the
COOH-terminal residues B26 to B30 are deleted, and (c) a
potential negative role of the COOH-terminal B-chain domain
in the interaction of hormone with receptor (23). We con-
12054
This is an Open Access article under the CC BY license.
Insulin-Receptor Interactions
12055
cluded
that
insulin-receptor
interactions normally involve
conformational changes requiring
movement
of
the
COOH-
terminal
B-chain
domain
(to
compensate
for
its
potential
negative effect)
and
that
such
changes normally
involve
in-
teraction of
the
PheBZs
sidechain
(a sidechain lying
outside
of
the
negative domain)
with
receptor
(23).
By
use
of
additional
semisynthetic
insulin
analogs,
we
now
identify
both
the
mo-
lecular
locus
that
is
involved
in
the
effect
of
the
COOH-
terminal
B-chain
domain
to
decrease
receptor
binding affinity
(and
thus
the
molecular
locus
involved
in
the
site
of
potential
B-chain
structural
change)
and
the
extent
of
conformational
change
that
appears
to
be
required
for high-affinity insulin-
receptor
interactions.
MATERIALS AND METHODS
Peptide Semisynthesis-Insulin analogs were prepared by methods
involving trypsin-assisted peptide bond formation (24) between the
a-carboxyl group of ArcB of
bis-t-butyloxycarbonyl-des-octapep-
tide(B23-B30)-insulin and the a-amino group of tri-, tetra-, penta-,
hexa-, hepta-, and &peptides prepared by solid-phase peptide syn-
thesis (25) on an Applied Biosystems model 430A Peptide Synthesizer
followed by treatment with
HF
(26). All peptides shorter than
8
residues in length were synthesized
as
their a-carboxamide deriva-
tives (by use of p-methylbenzhydrylamine resins)
to
ensure that
inappropriately placed, ionizable carboxyl groups were not introduced
into the final insulin analogs.
Bis-t-butyloxycarbonyl-des-octapep-
tide-insulin was prepared from Monocomponent porcine insulin
(Novo Pharmaceuticals, Copenhagen) by previously published meth-
ods
(23). Reagents for peptide synthesis were purchased from Applied
Biosystems (Foster City, CA), except for
9-fluorenylmethyloxycar-
bony1 glycine (27) which was from Bachem (Torrence, CA). Synthetic
peptides were purified by reverae-phase high performance liquid
chromatography by use of solvent mixtures containing 0.1% (v/v)
aqueous trifluoroacetic acid and 0.1% (v/v) trifluoroacetic acid pre-
pared in acetonitrile. Peptides containing LyeB" were synthesized as
their
N-a-9-fluorenylmethyloxycarbonyl-Gl~B
derivatives. Subse-
quent
to
cleavage from the resin, these peptides were converted
to
derivatives with t-butyloxycarbonyl groups at Lys"" and with free
a-amino groups
at
GIPm by use of
2-t-butyloxycarbonyloximino-2-
phenylacetonitrile
(28)
followed by treatment with 10% (v/v) piperi-
dine prepared in dimethylformamide. Peptide semisynthesis followed
the course previously described (23, 29). The products were depro-
teded by use of trifluoroacetic acid and were purified sequentially by
(a) gel filtration on Bio-Gel P-4 (by use of 3
M
acetic acid
as
the
solvent) and (b) ion-exchange chromatography on DEAE-Sephadex
A-25 or reverse-phase high performance liquid chromatography (30).
All of the insulin analogs exhibited the expected amino acid compo-
sitions (determined after hydrolysis in 6
N
HCl for 24 h at 110 "C)
and are identified in Table I.
Cross-linked insulins (the a-G1p,c-LysB"-suberoyl and a-Glfl',
c-LysB"-succinoyl derivatives of porcine insulin and of [Se$16]insu-
lin) were prepared by treating the parent peptides (3 mg/ml, dissolved
in dimethylsulfoxide) with
1
eq of
bis-N-hydroxysuccinimidyl
suber-
ate (Pierce Chemical Co., cf. Ref. 31) or
bis-N-hydroxysuccinimidyl
succinate (prepared in the laboratory, cf. Ref. 32). The products were
purified by high performance liquid chromatography. End group
analysis of each derivative by use of dansyl' chloride (33) revealed
only dansyl-Phe, a result documenting that cross-linking had
oc-
curred between Glp' and LysBm.
Receptor-binding Studies-Canine hepatocytes were isolated
as
described (34) and were incubated in glass scintillation vials at a cell
density of 2
X
10' cells/ml in supplemented Krebs-Ringer bicarbonate
buffer under an atmosphere containing 95%
02,
5% Con. Cell viability
for these studies always exceeded 98% as assessed by exclusion of
the dye Trypan blue. Individual incubation vials contained [(lZ6I)-
iodoTyP] insulin purified by high performance liquid chromatog-
raphy (50,000 cpm, about 2 pmol, obtained from Ldly) and porcine
insulin or semisynthetic insulin analogs
at
the concentrations given
in the figures. Cells were incubated with insulin or analogs for 30 min
at 30 "C after which the cell suspensions were diluted with ice-cold
buffer, the cells pelleted by centrifugation for 2 min at 400
X
g
in the
cold, and the cell pellets counted for radioactivity. The ability of
'
The abbreviation used is: dansyl,
dimethylaminonaphthalene-l-
sulfonyl.
TABLE
I
Identification and receptor-binding potencies of insulin analogs
The semisynthetic insulin analogs prepared for this study and their
receptor-binding potencies (relative
to
the potency of porcine insulin)
are identified below. Details of the semisynthetic methods
used
and
of the cell preparation and incubation conditions are provided under
"Materials and Methods." Relative receptor-binding potency is
de-
half-maximal inhibition of binding of [('261)iodo~'']insulin
to
re-
fined for these purposes as
[
(concentration of porcine insulin causing
binding of
[('261)iodoTyr""]insulin
to receptor)]
X
100. All inhibitions
ceptor)/(concentration of analog causing half-maximal inhibition of
were complete and all curves describing the concentration dependence
for the inhibition of radiolabeled insulin binding to receptor were
parallel. Each value represents the mean
f
S.D. of multiple deter-
minations; the number of separate determinations is shown in par-
enthesis. The concentration of insulin causing half-maximal inhibi-
tion of radiolabeled insulin binding was
0.70
f
0.13 nM
(n
=
11,
range
=
0.55-0.90 nM).
-
Identifying
No.
Peptide
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Insulin
Des-TyP2s,ThP27,ProBoB28,LysBm,ThPao-
[PheB16-a-carboxamide] insulin
[SeP26]insulin
Des-ThP'"[Sep16,LysBm-a-carboxam-
idelinsulin
Des-LysB29,ThP30-[SeP26,ProB28-a-car-
boxamide]insulin
D~~-P~O~",L~S~~,T~~~~-[S~P~~,T~P~~-
a-carboxamide]insulin
Des-Thpz7,ProBZ8,LysBZB,ThPo-
[SeP",TyP28-a-~arboxamide]insulin
Des-Ty$2',ThP27,ProB",LysB",Th~-
[SeP26-a-carboxamide]insulin
[Se~26,AlaB26]insulin
Des-ThPZ7,ProB3,LysBm,ThPm-
[SeP26,AlaB26-a-~arboxamide]insulin
[SeP26,AlaB27]ins~lin
De~-Pro~~,Lys~~,ThP"-[SeP~,Ala~~~-
a-carboxamide]insulin
a-G1p',t-Lyse"-suberoyl-insulin
a-Glp',c-LysBm-succinoyl-insulin
a-G1~',~-LysB2e-suberoyl-
[SePZ6]insulin
a-Gly"',t-LysB"-succinoyl-
ISePlslinsulin
Relative
potency
100
109f 13 (6)
1.1
f
0.3 (6)
1.2
f
0.3 (3)
1.7
f
0.5
(3)
2.2
f
0.4 (5)
9.9
f
3 (5)
45f6 (5)
1.4
f
0.2 (2)
32fO (2)
2.3
f
0.1 (2)
3.5
f
0.8
(2)
7.5
f
0.6 (3)
0.90
f
0.1 (3)
0.17
f
0.04 (3)
0.065
f
0.002 (2)
insulin or insulin analogs
to
inhibit the binding of '261-labeled insulin
to
hepatocyte insulin receptors
was
tested by duplicate determina-
tions on at least two occassions, with similar results being obtained
in each case. Representative studies
are
shown. Table I identifies for
each analog the concentration of peptide (relative
to
that of insulin)
that caused half-maximal inhibition of binding of '161-labeled insulin
to
hepatocyte insulin receptors.
RESULTS AND DISCUSSION
Previous
studies
have
already shown
that
(a)
replacement
of
insulin
PheBz6
by
a
residue
having
a
@-aromatic
ring
results
in
no
major change
in
the
affinity
of
hepatic insulin
receptors
for ligand,
(b)
replacement of insulin
PheBzs
by
a
residue
12056
Insulin-Receptor Interactions
lacking a 8-aromatic ring (whether
or
not that residue is in
fact aromatic
or
whether it is hydrophobic
or
hydrophilic)
results in a decrease in receptor binding potency (most often
a decrease of about 100-fold), and
(c)
deletion of a portion of
the COOH-terminal domain of the insulin B-chain containing
residues B26 to B30 results in a partial
or
complete reversal
of the diminished receptor binding potencies of insulins con-
taining PheB25 substitutions (23). Fig. la shows results for
insulins substituted with serine at position B25, a replacement
that can serve as a archetype for others. While [SeP25]insulin
(peptide 3) has an apparent affinity for receptors only
1%
as
great as that of the natural hormone, and while deletion of
residues B26-B30 has little effect on the binding of the
truncated analog containing Phe at position B25 (peptide 2),
deletion of residues B26-B30 in [Sep25]insulin (peptide
8)
increases by 40-fold the receptor binding potency of the
substituted insulin analog. It thus appears that residues B26-
B30 can play a negative role in the interaction of insulin
analogs with receptors.
To define more clearly the site in the COOH-terminal B-
chain domain of insulin that participates in a negative way,
we synthesized and analyzed insulins containing Ser at posi-
tion B25 and containing deletions of 1, 2, 3, 4,
or
5 residues
from the B-chain COOH terminus. Fig.
lb
shows that deletion
of the COOH-terminal 1,
2,
or
3 residues of the B-chain in
[SePZ5]insulin (peptides 4, 5, and 6, respectively) has only a
'"t
O-I
I
Log
molar
peptide concentration
FIG.
1.
Inhibition of
[('261)iodoTyrA14]insulin
binding to iso-
lated canine hepatocytes by full length and truncated insulin
analogs containing single amino acid substitutions.
Analogs
and radiolabeled insulin were incubated with canine hepatocytes as
described under "Materials and Methods." Quantitative information
is provided in Table
I;
identifying numbers in Table
I,
rather than
a:
1,
inhibition by insulin;
2,
inhibition by des-TyPz6,ThPZ7,
symbols, are used to indicate the peptide under consideration.
Panel
ProB28,LysB",ThP30-[PheBz6-a-carboxamide]insulin;
3,
inhibition by
[SePZ5]insulin;
8,
inhibition by
des-TyP26,ThP27,ProB28,LysB",
ThP30-[Se~25-a-carboxamide]insulin.
Panel
b
4,
inhibition by
des-Th$"-[SeP25,LysB29-a-carboxamide]insulin;
5, inhibition by
des-LysB29,Th$30-[Se~~,ProB28-a-carboxamide]insulin;
6,
inhibi-
tion by
des-ProB28,LysB",Th~30-[Se~26,Th~27-~-~arboxamide]in-
sulin;
7,
inhibition by
des-ThrB27,ProB28,LysB2g,ThrB30-[SerB25,
TyrB26-~-carboxamide]insulin;
8,
inhibition by des-TyrBz8,ThrBz7,
ProB2s,LysB2s,ThrB30-[SerB26-a-carboxamide]insulin.
very slight effect to reverse the lowered binding potency that
attends substitution of PheB25 by Ser. Deletion of the COOH-
terminal
4
residues of the B-chain (peptide
7),
however,
increases the receptor binding potency of [SePZ5]insulin by
9-fold, and (as noted above) deletion of the COOH-terminal
5
residues of the B-chain increases the receptor binding
potency of [Sep25]insulin by an additional 4.5-fold. (Further
deletions, deletions involving removal of Sep25, cause the
markedly decreased association of insulin with receptors (23).)
Taken together, the results of Fig.
1
demonstrate that the
negative effect of the COOH-terminal B-chain domain on the
interactions of SerB25 -substituted insulins with receptor arises
from the presence of residues Ty$26 and Th$27. More distal
residues (ProBm, LysB29, and Thp3") apparently play no role,
either negative
or
positive, in determining the affinity of
insulin-receptor interactions.
Since the importance of TypZ6 and ThP7 in determining
the affinity of insulin-receptor interactions could arise from
either an effect of the length of the peptide mainchain
or
an
effect of structural features of the relevant amino acid side-
chains, insulin derivatives selectively assessing mainchain
length and sidechain structure were constructed and subjected
to study. Fig. 2a illustrates results for analogs designed to
assess the importance of sidechain structure at position
B26; i.e. for analogs containing AlaBz6 for TyF6 and SePZ5
for PheBz5 replacements. Fig. 2a shows that replacing TyP6
by Ala (in peptide 10) indeed increases the apparent receptor
binding potency of
de~-Th$~~,Pro~~~,Lys~~,Thp~~-[Sep~~,
"i,
I
I
I I I
-10
-9
-8
-7
-6
Log
molar
peptide concentration
FIG.
2.
Inhibition of
[('261)iodoTyrA'4]insulin
binding to
iso-
lated canine hepatocytes
by
full length and truncated insulin
analogs containing two amino acid substitutions.
Analogs and
radiolabeled insulin were incubated with canine hepatocytes as de-
scribed under "Materials and Methods." Quantitative information is
provided in Table
I;
identifying numbers in Table
I,
rather than
symbols, are used to indicate the peptide under consideration.
Panel
a:
1,
inhibition by insulin;
3,
inhibition by [Se8'25]insulin;
7,
inhibition
by
des-Th~27,ProB28,LysBzg,Th$30-[Se$25,Ty$26-~-~arboxamide]in-
sulin;
9.
inhibition by [SerB26,AlaB26]insulin;
10,
inhibition by
des-Th~27,ProB28,LysB29,Th~-[Se~s,AlaB26-~-carboxamide]insulin.
Panel
b
3,
inhibition by [Se$25]insulin;
6,
inhibition by des-
ProB28,LysB29,Th~30-[Se~25,Th$Z7-a-~arboxamide]insulin;
11,
inhi-
bition by [Se$z5,AlaB27]insulin;
12,
inhibition by des-ProBm,
LysBz9,ThPQ-[SeP25,AlaB27-a-carboxamide]insulin.
Insulin-Receptor
Interactions
12057
Tyl.BZ6-~-carboxamide]insulin
(peptide 7) about 3-fold. AI-
though this result suggests the importance of the tyrosinyl
sidechain in inducing the negative effect observed for the
COOH-terminal B-chain domain in insulin-receptor interac-
tions, replacement of TyPz6 by Ala in full length [SePz5]
insulin (peptide
9)
causes only a very slight increase in the
receptor binding potency of the parent insulin analog. It thus
appears that the mass of the tyrosinyl sidechain indeed plays
a role in decreasing the affinity of the truncated insulin analog
for receptor, but that this role is of minor consequence when
dealing with a more severe effect arising from mainchain
length rather than sidechain structure in analogs of larger
size.
Since the results of Fig.
lb
had suggested that the presence
of ThPZ7, as well as that of TyPZ6, might contribute to the
decreased receptor binding potency of [Se$25]insulin, we fur-
ther assessed the potential role of the threoninyl sidechain
(perhaps through hydrogen bonding) in contributing to the
decreased potencies of Se~-‘j~~-substituted insulin analogs. Fig.
2b
shows, however, that replacement of Thp27 by Ala in both
the truncated analog
de~-Pro~~,Lys~~,Thr~~~-[Se~~~,Ala~~~-
a-carboxamide]insulin (peptide
12)
and in full length [SePZ5]
insulin (peptide
11)
fails to increase receptor binding potency
significantly. We can thus conclude from Fig. 2, a and
b
that,
for the most part (and certainly for full length, 51-residue
insulins), mainchain length rather than sidechain structure
determines the decreased receptor binding potency of insulin
analogs containing the replacement of PheBZ5 by Ser.
Results described above identify a region in the COOH-
terminal B-chain domain involving residues TyrBZ6 and
ThrBZ7, a region most probably limited to the mainchain in
the area of the TyPZ6-ThPz7 peptide bond, that
(a)
contrib-
utes strongly to the decreased receptor binding affinity of
[SerBZ5]insulin relative to that of the natural hormone, and
(b)
probably participates in necessary conformational change
to achieve high-affinity ligand-receptor interactions when
position B25
is
filled by Phe or by another residue containing
a &aromatic ring (23). While the sequence GlyBZ3-PheBz4-
PheBZ5-TyrBz6 has remained nearly invariant during animal
evolution, more distal residues involving positions B27 to B30
have indeed undergone multiple evolutionary changes. In the
extraordinary cases of both coypu (35) and dogiish (36) in-
sulins, residue B25 (normally Phe) has apparently been de-
leted, with the result that Tyr (normally appearing at position
B26) occurs at position
B25
and that residues B26-B30 have
become shifted in frame relative to most insulins. It is the
case that the appearance of Tyr at position B25 is in no way
detrimental (since TyPz5 fulfills the criterion of a residue
with a @-aromatic ring appearing at position B25, Ref.
23),
a
tyrosine residue is not actually required at position B26
(23),
and the ordering of residues B26-B30 is not critical to the
biological activity of the resultant insulin.
As
stated previously, the 40-fold increase in the receptor
binding potency of
des-Ty$26,ThrBZ7,ProBz8,LysBzg,ThrB30-
[SerBz5-a-carboxamide]insulin,
when compared to the potency
of full length [SerBZ5]insulin, can only be rationalized by terms
involving a concerted conformational change in ligand and
receptor which normally requires the presence of a @-aromatic
side chain at position B25 and which alleviates a potential
negative effect of the COOH-terminal B-chain domain in
limiting high-affinity interaction of ligand with receptor (23).
This conformational change, however, need not be the only
one to affect the potency of insulin-receptor interactions. The
use of insulin analogs covalently cross-linked between LystBz9
and GlyaA’ (analogs that have been designed to restrict the
extent of possible conformational changes within the hormone
by linking regions of the insulin A- and B-chain which are in
close proximity)
(9-11,
17,
18,
37) suggests, in fact, that
multiple conformational changes must occur during insulin-
receptor interactions. Fig. 3a shows that insulin when cross-
linked by the suberoyl group (peptide
13)
exhibits a receptor
binding potency 7.5% of that of native insulin, and that
further restriction of potential conformational mobility by the
introduction of the succinoyl group as the cross-linker (pep-
tide
14)
decreases apparent affinity of receptor for ligand
another 8-fold.
Fig. 36 shows that [Se8‘25]insulin (an analog which already
has limited ability to undergo receptor-induced conforma-
tional change, Ref. 23) when cross-linked by the suberoyl
group (peptide 15) exhibits a receptor-binding potency only
15% of that of [SePZ5]insulin and only about 0.2% of that of
the unmodified hormone. Further restriction of conforma-
tional mobility by introduction of the succinoyl cross-link, in
this case, decreases receptor-binding potency to 0.065% of
that of insulin.
It
is
apparent that restriction of potential
conformational change in insulins lacking a @-aromatic ring
at position B25, as well as in native insulin, decreases sub-
stantially the affinity of receptor for ligand. The fact that the
receptor binding potency of cross-linked [SePZ5]insulin is
nearly the same as that
of
des-octapeptide(B26-B30)-insulin
(23) suggests that, once PheBZ5 has been replaced, and once
conformational mobility of the insulin analog has been
further
restricted by cross-linking, the COOH-terminal region of the
insulin B-chain apparently plays no significant role in deter-
mining the affinity of insulin-receptor interactions.
Taken together, our results identify (a) the site of the
negative effect of the COOH-terminal B-chain domain of
insulin in position B25-substituted analogs as arising from
the insulin mainchain in the region of the TyP26-ThrB27
peptide bond,
(b)
the site of the inferred conformational
b
-9
-8 -7
-6
-5
Log
molar peptide concentration
FIG.
3.
Inhibition
of
[(12”I)iodoTyrA’4]insulin
binding to ca-
nine hepatocytes by native and cross-linked insulin and in-
sulin analogs.
Analogs and radiolabeled insulin were incubated with
canine hepatocytes as described under “Materials and Methods.”
Quantitative information is provided in Table
I;
identifying numbers
in Table
I,
rather than symbols, are used to indicate the peptide
under consideration.
Panel
a:
1,
inhibition by insulin;
13,
inhibition
by
a-GlyA1,~-LysB2@-suberoyl-insulin;
14,
inhibition by a-G1fll,c-
LysBm-succinoyl-insulin.
Panel
b:
3,
inhibition by [Se$z5]insuIin;
15,
inhibition by
a-G1~,~-LysB29-suberoyl-[Se~25]-insulin;
16,
inhibition
by
a-G1~1,c-LysB29-succinoyl-[Se~25]-insulin.
12058
Insulin-Receptor Interactions
change resulting in the attainment of a high-affinity ligand-
receptor state for insulins containing a residue with a
B-
aromatic sidechain at position B25 as arising from the insulin
mainchain rather than from sidechains in the COOH-terminal
B-chain domain
(cf.
Ref. 23), and
(c)
the requirement for a
conformational change that is apparent from the analysis of
cross-linked insulin and [SeP2']insulin and that is quite sep-
arate from the conformational change identified above as
requiring the presence of a residue
at
position
B25
containing
a
sidechain with a /3-aromatic ring. It thus appears that, by
increments, the receptor binding affinity of the conformation-
ally restricted analog
a-Gl~',t-LysB29-s~~cinoyl-[Se~2']in-
sulin (peptide
16),
an analog that retains only
0.065%
of the
receptor binding potency of native insulin, can be increased
by 17-fold by deletion of the cross-link (to achieve the struc-
ture of [Se~"]insulin) and then can be increased an addi-
tional 40-fold by deletion of residues B26-B30 (to achieve the
structure of
de~-Typ~~,Th$~~,Pro~~~,Lys~~,Th$30_[Sep~~-a-
carboxamide]insulin. While the
full
extent of participation of
the insulin mainchain in these individual conformational
changes is not known with assurance, our results provide a
framework for understanding both the multiple components
which result in the attainment of a high-affinity state of
insulin-receptor interactions and the ways in which hormone-
receptor associations (with their need for conformational
changes concomitant with transmembrane signaling) differ
from simple events involving, for example, the binding of
ligand to carriers or transport proteins.
Acknowledgments-We
thank Dr. Yasushi Nakagawa for perform-
ing the amino acid analyses, Hector Escoffie for help in preparing
isolated canine hepatocytes, and Crystal Sherman-Jones for assist-
ance in the preparation of the manuscript.
REFERENCES
1.
Sklar, L. A., Bokoch, G. M., Button, D., and Smolen,
J.
E.
(1987)
2.
Yuli,
I.,
and Oplatka, A.
(1987)
Science
235,340-342
3.
Covarrubias, M., Prinz, H., Meyers, H.-W., and Maelicke,
A.
(1986)
J. Biol. Chem.
261,14955-14961
4.
Donner, D.
B.,
and Corin, R. E.
(1980)
J.
Biol.
Chem.
265,9005-
9008
5.
Corin, R. E., and Donner, D. B.
(1982)
J. Biol. Chem.
257,104-
110
6.
Donner, D. B., and Yonkers, K.
(1983)
J. Biol. Chem.
258,9413-
9418
7.
Lipson, K.
E.,
Yamada, K., Kolhatkar,
A.
A., and Donner, D.
B.
(1986)
J.
Biol.
Chem.
261,10833-10838
8.
Dodson,
E.
J.,
Dodson, G. G., Hubbard, R. E., and Reynolds, C.
D.
(1983)
Biopolymers
22,281-291
9.
Blundell,
J.
T., Cutfield,
J.
F.,
Cutfield,
S.
M., Dodson, E.
J.,
Dodson, G. G., Hodgkin, D. C., Mercola, D. A., and Vijayan,
M.
(1971)
Nature
231,506-511
10.
Blundell, T., Dodson, G., Hodgkin, D., and Mercola, D.
(1972)
Adu. Protein Chem.
26,279-402
J. Bwl. Chem.
262,135-139
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
23.
22.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
Chothia, C., Leske, A. M., Dodson, G. G., and Hodgkin, D. C.
(1983)
Nature
302,500-505
Bentley, G., Dodson,
E.,
Dodson, G., Hodgkin, D., and Mercola,
D.
(1976)
Nature
261,166-168
Smith, G.
D.,
Swenson, D.
C.,
Dodson,
E.
J., Dodson, G. G., and
Reynolds, C. D.
(1984)
Proc.
Natl. Acad.
Sci.
U. S.
A.
81,7093-
7097
Bi,
R.
C., Dauter, Z., Dodson, E., Dodson, G., Giordano, F., and
Reynolds, C.
(1984)
Bwpolymers
23,391-395
Bi, R.-C., Cutfield,
S.
M., Dodson, E.
J.,
Dodson, G. G., Giordano,
F.,
Reynolds,
C.
D., and Tolley,
S.
P.
(1983)
Acta Crystauogr.
Dongeai, L., Stuart, D., Jinbi, D., Todd,
R.,
Junming, Y., and
Meizhen,
L.
(1985)
Sci. Sin.
B28,472-483
Brandenburg, D., Schermutki,
W.,
and Zahn, H.
(1973)
2.
Physiol.
Chem.
354,1521-1524
Dodson, G. G., Cutfield,
S.,
Hohenjet, E., Wollmer, A., and
Brandenburg, D.
(1980)
in
Insulin: Chemistry, Structure and
Function
of
Insulin and Related
Hormones
(Brandenburg, D.,
and Wollmer, A., eds) pp.
17-26.
Walter de Gruyter and Co.,
New York
Friesen, H.-J., Saunders, D.
J.,
Naithani, V. K., Diaconescu, C.,
and Rosen,
P.
(1979)
in
Peptides: Proceedings
of
the 6th Amer-
ican Peptide Symposium
(Gross, E., and Meienhofer, J., e&)
pp.
511-518.
Pierce Chemical Co., Rockford IL
Peavy, D.
E.,
Brunner, M. R., Duckworth, W. C., Hooker, C.
S.,
and Frank, B. H.
(1985)
J. Biol. Chem.
260,13989-13994
Markussen,
J.,
Jorgensen, K. H., Sorensen, A. R., and Thim, L.
(1985)
Znt. J. Peptide Protein Res.
26, 70-77
Kubiak,
T.,
and Cowburn, D.
(1986)
Biochem.
J.
234,665-670
Nakagawa,
S.
H., and Tager, H.
S.
(1986)
J. Bwl. Chem.
261,
Inouye,
K.,
Watanabe,
K.,
Morihara, K., Tochino, Y., Kanaya,
T.,
Emura,
J.,
and Sakakibara,
S.
(1979)
J.
Am. Chem.
SOC.
B39,90-98
7332-7341
101,751-752
Merrifield, R. B.
(1963)
J.
Am. Chem.
SOC.
85,2149-2154
Sakakibara,
S.,
Shimonishi, Y., Kishida, Y., Okada, M., and
Sugihara, H.
(1967)
Bull. Chem.
SOC.
Jpn.
40,2164-2167
Chang, C.-D., Waki,
M.,
Ahmad, M., Meinhofer, J., Lundell, E.
O.,
and Haug,
J.
D.
(1980)
Znt.
J. Pept. Protein Res.
15,59-66
Itoh, M., Hagiwara, D., and Kamiya,
T.
(1975)
Tetrahedron
Lett.
4393-4394
Inouye,
K.,
Watanabe, K., Tochino,
K.,
Kobayashi, M., and
Shigeta, Y.
(1981)
Bwpolymers
20,1845-1858
Shoelson,
S.,
Haneda,
M.,
Blix,
P.,
Nanjo,
A.,
Sanke,
T.,
Inouye,
K., Steiner, D., Rubenstein, A., and Tager, H.
(1983)
Nature
Brandenburg, D., Busse, W.-D., Gattner, H.-G., Zahn, H., Woll-
mer,
A.,
Gliemann,
J.,
and
Puls,
W.
(1972)
in
Proceedings
of
the 12th European Peptide Symposium
(Hanson, H., and Jak-
ubke. H.-D.. e&)
DD.
270-283.
Elsevier Scientific Publishing
302,540-543
.
"
CO.,
Amsterdam
-
Lindsav. D. G.
(1971)
FEBS Lett.
21,
105-108
Gray,
W.
R.
(1969)
kethods Enzymoi
11, 139-151
Bonnevie-Nielsen, V., Polonsky, K.
S.,
Jaspan,
J. J.,
Rubenstein,
A. H., Schwartz,
T.,
and Tager, H.
S.
(1982)
Proc.
Natl. Acad.
Sci.
U.
S.
A.
79,2167-2171
Smith, L.
F.
(1972)
Diabetes
2l(suppl.
2), 457-460
Bejaj, M., Blundell,
T.
L.,
Pitts,
J.
E., Wood,
S.
P., Tatnell, M.
A.,
Falkmer,
S.,
Emdin,
S.
O.,
Gowan, L. K., Crow, H., Schwabe,
C., Wollmer,
A.,
and Strassburger, W.
(1983)
Eur. J. Biochem.
136,535-542
Brandenburg, D.
(1972)
2.
Physiol. Chem.
353,869-873
