(Received for publication, October 29, 1997)
From the Department of Molecular Pharmacology and Biochemistry, Merck Research Laboratories, Rahway, New Jersey 07065
Deletion of residues 252-259 within the putative
second intracellular loop of the human glucagon receptor results in a
protein with high affinity for glucagon but with attenuated agonist
activation of adenylyl cyclase. The 252-259 mutant has 4-fold
higher affinity for glucagon than does the wild type receptor. The
nonhydrolyzable GTP analog, guanosine 5
-(
,
-imido)triphosphate
(Gpp(NH)p), inhibits binding of 125I-glucagon to the
wild type receptor but not to the
252-259 mutant. Divalent cations
such as MgCl2 and CaCl2 stimulate the binding of 125I-glucagon to the wild type receptor by increasing
glucagon affinity. The rate of dissociation of
125I-glucagon is decreased 4-fold by MgCl2 and
increased 6-fold by Gpp(NH)p. However, divalent cations do not affect
the binding of 125I-glucagon to the
252-259 mutant. The
rate of dissociation of 125I-glucagon from the
252-259
mutant protein is equivalent to the rate of dissociation from the wild
type receptor in the presence of MgCl2. These data suggest
that at least three conformations of the glucagon receptor can exist in
the membrane based on their differing affinities for
125I-glucagon. Deletion of residues 252-259 appears to
lock the protein in the conformation promoted by divalent cations and
prevents the protein from normal coupling to Gs.
Glucagon is a major counterregulatory hormone that attenuates the inhibition of liver gluconeogenesis and glycogenolysis by insulin. Glucagon action is mediated by a G protein-coupled receptor (1) that stimulates cyclic AMP accumulation via activation of Gs. G protein-coupled receptors are characterized by the ability of agonists to promote the formation of a high affinity ternary complex between the agonist, the receptor, and the G protein (2). The affinity of the agonist for the receptor-G protein complex is higher than its affinity for the uncomplexed receptor, so that disruption of the ternary complex reduces the affinity of the receptor for the agonist. Thus, the affinity of agonists for G protein-coupled receptors is a function of the affinity of the receptor for G protein. The changes in affinity of agonist and G protein for receptor are thought to reflect changes in the conformation of the receptor. In contrast, most antagonists bind with the same affinity to the receptor in the presence or absence of G protein coupling.
G protein-coupled receptors such as the glucagon receptor are predicted
to have seven transmembrane domains linked by hydrophilic loops (1, 2).
Mutagenesis and modeling studies predict that the binding domain for
small molecules is within the transmembrane helical domains, although
peptide agonist binding also involves the hydrophilic extracellular
domains (3). Coupling of the -adrenergic receptors to Gs is
disrupted by deletion of either the amino-terminal or carboxyl-terminal
sections of the third intracellular loop (4). These mutant proteins
maintained high affinity for agonist. A similar phenotype has been
observed in the receptor for glucagon-like peptide 1, a receptor that
is 50% homologous to the glucagon receptor, when three residues
predicted to be at the amino-terminal of the third intracellular loop
are deleted (5). The second intracellular loop has also been shown to
be involved in G protein coupling (2). In the current work we have
investigated the properties of glucagon receptors with deletions within
the second and third intracellular loops.
Mutagenesis of a
cloned human glucagon receptor (6) was performed using the Bio-Rad
Mutagene Phagemid in vitro mutagenesis kit, version 2. The
putative second intracellular loop was deleted by modification of the
sequence of the wild type receptor cDNA to produce the 252-259
mutant protein using the oligonucleotide primer 5
CTG TAC CTG CAC AAC
GAG AGG AGC TTC TTC 3
. This primer sequence corresponds to nucleotides
739 through 753 and 778 through 792 of the coding sequence of the human
glucagon receptor. The deletion (i.e. nucleotides 754 through 777 of the coding sequence of the human glucagon receptor) was
confirmed by sequencing. The cDNA was ligated into the vector
PCI/neo (Promega) for expression studies. Deletion of the
amino-terminal end of the putative third intracellular loop (
334-340) was prepared using oligonucleotide primer 5
CTC GTG GCC AAG
CTG ACA GAC TAC AAG TTC 3
, and the deletion was confirmed by
sequencing. Deletion of the carboxyl-terminal end of the putative third
intracellular loop (
341-347) was prepared using oligonucleotide
primer 5
CGG CAG ATG CAC CAC GCC AAG TCC ACG CTG 3
, and the deletion
was confirmed by sequencing. Deletion of amino acid residues 332-334
was accomplished by recombinant polymerase chain reaction using the
oligonucleotide primer 5
-ATC GTT CAG CTG CTC GTG GCC GCA CGG CAG ATG
CAC CAC-3
and its complement. The cDNA was ligated into the vector
PCI/neo-wild type glucagon receptor for expression studies. The
deletion of amino acid residues 332-334 was confirmed by DNA
sequencing.
COS cells were transfected in T175 flask monolayers using 10 µg of cDNA and 8 mg of DEAE-dextran in Dulbecco's modified Eagle's medium containing 10% Nu-Serum (Collaborative Biomedical Products) and 100 µM chloroquine for 2 h at 37 °C in 5% CO2 incubation followed by 2-min shock in 10% Me2SO. Cells were maintained in fresh growth medium overnight then were harvested and seeded into T175 flasks for binding and 24-well dishes for cyclase activity measurement. Stable cell lines were isolated after transfection into CHO1 cells and selection with G418.
Characterization of the Binding and Functional ActivityAfter 2 days, COS cells were harvested from flasks by
scraping in 1 mM Tris, pH 7.4, homogenized, and the crude
membrane pellet was recovered by centrifugation.
125I-Glucagon (58 pM) binding to the membrane
preparation was measured in 20 mM Tris, pH 7.4, containing
2.5 mM MgCl2, 1 mM dithiothreitol, 5 µg/ml leupeptin, 10 µg/ml benzamidine, 40 µg/ml bacitracin, 5 µg/ml soybean trypsin inhibitor, and 3 µM
o-phenanthroline ± 1 µM glucagon for
1 h. Bound counts/min were recovered by filtration using a Tomtec
harvester and quantified in a -scintillation counter. Functional
responses in transfectants were measured by exposing the cell
monolayers to Dulbecco's modified Eagle's medium (basal), Dulbecco's
modified Eagle's medium with increasing concentrations of glucagon or
Dulbecco's modified Eagle's medium with 10 µM forskolin for 30 min at 22 °C to generate intracellular cAMP. Cells were lysed
in HCl, and cAMP in lysates was acetylated and quantified by
radioimmunoassay. In stable CHO cell lines expressing the receptors, cells were harvested from monolayers with enzyme-free cell dissociation solution (Specialty Media, Inc.) and were pelleted at 500 × g. The cells were resuspended at 100,000 cells/100 µl in
75 mM Tris-HCl, pH 7.4, containing 250 mM
sucrose, 12.5 mM MgCl2, 1.5 mM
EDTA, 0.1 mM Ro-20-1724 (Biomol, Inc.), leupeptin (5 µg/ml), benzamidine (10 µg/ml), bacitracin (40 µg/ml), soybean
trypsin inhibitor (5 µg/ml), and 0.02% bovine serum albumin. Cells
were incubated for 30 min at 22 °C with increasing peptide
concentrations followed by lysis by boiling. Lysates were analyzed for
cAMP content versus a nonacetylated cAMP standard curve
using the Amersham cAMP radioimmunoassay SPA kit. Data were analyzed
using Packard TopCount with RIASmart and GraphPad Inplot4 software.
Transient expression studies in COS cells were repeated at least twice
with each mutant protein with similar results. In addition, stable CHO
cell lines expressing the
252-259 mutant protein were isolated and
analyzed.
The amount of receptor protein in COS membranes was quantitated by binding to an antibody to the COOH terminus of the glucagon receptor (7) on a dot blot using various concentrations of membrane protein. Membranes (0.06-2 µg of protein) were applied to nitrocellulose, incubated with primary antibody in the presence or absence of 100 µg/ml of the antigenic peptide, incubated with 125I-protein A, and then exposed to a PhosphorImager screen overnight before quantitation.
Deletion mutations were made in the human glucagon receptor within
putative intracellular loops 2 (252-259) and 3 (
334-340,
341-347, and
332-334). The positions of these mutations are shown schematically in Fig. 1. After transfection into
COS cells, binding of 125I-glucagon to the
252-259,
341-347, and
332-334 mutant proteins is comparable (40-60%)
with binding to the wild type glucagon receptor. However, binding to
the
334-340 mutant protein is <5% of that observed to the wild
type receptor, suggesting that this mutant protein is not well
expressed.
After transfection into COS cells, deletion of residues 252-259
results in marked attenuation of the ability of glucagon to activate
adenylyl cyclase (Fig. 2). The EC50 for
glucagon activation of the wild type receptor in transiently
transfected COS cells is 84 ± 67 pM (Table
I). Under conditions where glucagon stimulates adenylyl
cyclase activity of the wild type receptor 6-fold, glucagon (100 nM) stimulates adenylyl cyclase activity of the 252-259 mutant protein 47% (Fig. 2). In several independent transfections into
COS cells, glucagon stimulates adenylyl cyclase activity of the wild
type glucagon receptor 819 ± 743% (n = 5). In
contrast, glucagon stimulation of activity of the
252-259 mutant
protein is 40 ± 20% (n = 3). This is 12 ± 3% of the glucagon activation of wild type receptors in these
experiments (Table I). Basal and forskolin-stimulated adenylyl cyclase
activities are comparable for the COS cells transfected with wild type
receptor and the
252-259 mutant protein (data not shown). A similar
phenotype is observed in three independent CHO cell lines expressing
the
252-259 mutant protein. Maximal glucagon activation of the
252-259 mutant protein is 56 ± 40% (n = 3)
over basal activity. This is 7 ± 5% of the activation observed
with wild type glucagon receptor expressed at similar levels
(approximately 500 fmol/mg of protein).
|
Deletion of residues 332-334 from the amino-terminal end of putative
intracellular domain 3 results in a 42-fold increase in the
EC50 for glucagon activation and a decrease of 39% in the maximal activation observed (Table I). Even though the 334-340 mutant protein is poorly expressed, glucagon (100 nM)
stimulates adenylyl cyclase activity to 29 ± 18% of the maximal
stimulation observed in cells expressing the wild type glucagon
receptor (Table I). However, glucagon is >100-fold less active at this
mutant compared with the wild type receptor. Deletion of residues
341-347 from the carboxyl-terminal end of the third intracellular loop increases the EC50 for glucagon activation 24-fold, and
maximal glucagon-stimulated adenylyl cyclase activity was 67% of the
stimulation observed with the wild type receptor (Table I). Little
effect on basal or forskolin-stimulated adenylyl cyclase activity is observed with these mutations (data not shown).
Membranes prepared from COS cells transfected with the wild type human
glucagon receptor, the 252-259 mutant protein or the
3
adrenergic receptor were applied to nitrocellulose paper by dot blot
and incubated with a polyclonal antibody raised against a peptide
corresponding to the carboxyl-terminal of the receptor. This antibody
has been previously demonstrated to specifically interact with the
glucagon receptor (7), and no bands are observed by Western analysis of
gels loaded with untransfected COS cell membranes (10 µg of protein).
Similarly, no spots are observed on the dot blots loaded with membranes
expressing
3-adrenergic receptors (4 µg of protein).
Densitometric analysis of the conjugates after incubation with
125I-protein A is linear versus membrane protein
added (Fig. 3). Coincubation with the antigenic peptide
attenuates the signal (data not shown). Analysis of the slopes of the
lines in Fig. 3 indicate that the membranes expressing wild type
glucagon receptor express 60% higher levels of protein than membranes
expressing the
252-259 mutant protein.
The nonhydrolyzable GTP analog Gpp(NH)p inhibits binding of
125I-glucagon to the wild type human glucagon receptor with
an IC50 = 22 nM (Fig. 4). The
maximal inhibition of binding at excess levels of Gpp(NH)p is 90% of
the specific binding of 125I-glucagon observed in the
absence of nucleotide. These data suggest that 90% of the observed
binding of 125I-glucagon is to receptor coupled to Gs.
However, Gpp(NH)p does not inhibit the binding of
125I-glucagon to the 252-259 mutant protein, suggesting
that this protein is not coupled to Gs (Fig. 4).
Glucagon inhibits binding of 125I-glucagon to the
252-259 mutant protein with 4-fold higher affinity than to the wild
type protein in the absence of added divalent cations. The
IC50 (mean ± S.D.) for three experiments in COS cells
and three experiments in CHO cells (i.e. n = 6) is
3.5 ± 2.0 and 0.76 ± 0.39 for the wild type receptor and
the
252-259 mutant protein, respectively.
Binding of 125I-glucagon to the wild type human glucagon
receptor expressed in CHO cells is stimulated 2-3-fold by divalent
cations such as MgCl2, MgSO4, and
CaCl2 (Fig. 5, top). The EC50
for this stimulation is ~0.3 mM. The primary effect of
divalent cation is to increase the affinity for glucagon 4-fold.
Unlabeled glucagon inhibits binding of 125I-glucagon to the
wild type glucagon receptor with an IC50 = 2.6 ± 1.5 nM (n = 3). Inclusion of 5 mM
MgCl2 in the incubation decreases the IC50 for
glucagon to 0.65 ± 0.13 nM (n = 3).
In contrast, divalent cations have no effect on binding of
125I-glucagon to the 252-259 mutant protein (Fig. 5,
bottom). Glucagon inhibits the binding with IC50
values of 0.76 ± 0.39 nM (n = 3) and
0.53 ± 0.15 nM (n = 3) in the absence
or presence of 5 mM MgCl2, respectively.
The rate of dissociation of 125I-glucagon for the wild type
receptor is decreased in the presence of divalent cation (Fig.
6). The dissociation rate of 125I-glucagon
in the absence or presence of divalent cation is 0.04 ± 0.01 min1 (n = 2) or 0.013 ± 0.001 min
1 (n = 2), respectively. In the
presence of excess Gpp(NH)p, the rate of dissociation is increased to
0.27 ± 0.04 min
1 (n = 2). In
contrast, divalent cation has no effect on the dissociation rate from
the
252-259 mutant protein (Fig. 6). The rates of dissociation of
125I-glucagon from the
252-259 mutant protein are
0.013 ± 0.002 min
1 (n = 3) and
0.010 ± 0.004 min
1 (n = 3) in the
absence and presence of 5 mM MgCl2,
respectively.
Mutagenesis experiments with the adrenergic receptor family have
shown that deletions or replacements within the third intracellular loop attenuate agonist-mediated signal transduction and, in many cases,
result in a protein that has high affinity for agonist ligands in the
absence of G protein coupling (4). Similar data have subsequently been
obtained for other G protein-coupled receptors (2). More recently,
deletion of a Lys-Leu-Lys tripeptide sequence predicted to be adjacent
to the fifth transmembrane domain of the glucagon-like peptide-1
receptor produces a similar phenotype, displaying high affinity binding
of glucagon-like peptide (7-37) amide, but only 10% of the maximal
activation of adenylyl cyclase as the wild type receptor (5). However,
our data show that deletion of the homologous Lys-Leu-Arg sequence of
the related glucagon receptor (332-334 mutant protein) increases
the EC50 for glucagon 42-fold, but does not completely
attenuate the maximal activation observed. Deletion of residues
334-340 results in a protein that is poorly expressed as determined by
its low binding of labeled glucagon and low potential for glucagon
activation. Deletion of the seven amino acids predicted to form the
carboxyl-terminal end of the third intracellular loop of the human
glucagon receptor also results in a 24-fold increase in
EC50 with only a 33% decrease in the maximal receptor
activation. These data suggest that the third intracellular loop of the
glucagon receptor is involved in receptor activation and that
impairment of coupling efficiency results from deletions in this
region.
In contrast, deletion of residues 252-259 within the putative second
intracellular loop of the human glucagon receptor results in a protein
that is well expressed and that has high affinity for glucagon with
drastically attenuated glucagon-mediated signal activation. Its degree
of activation is even lower than that for the poorly expressed
334-340 protein. This phenotype more closely resembles that
observed with intracellular loop 3 deletions in the adrenergic
receptors, although reductions in coupling efficiency have been
observed in both biogenic amine receptors and peptide receptors after
second intracellular domain deletions (2).
The glucagon receptor and other members of this structural family (i.e. glucagon-like peptide-1, pituitary adenylyl cyclase-activating peptide, calcitonin, parathyroid hormone, and corticotropin-releasing factor receptors) lack the sequence markers that are conserved in other families of G protein-coupled receptors (8). Therefore, it has been more difficult to predict the beginning and ends of the transmembrane domain helices in this family of proteins. However, the similarity of the phenotypes of these putative second and third intracellular domain mutations in the glucagon receptor to the phenotypes of the analogous mutations in biogenic amine receptors is consistent with the assignment of the helices of the glucagon receptor as described by MacNeil et al. (6).
Analysis of binding isotherms shows that the 252-259 mutant protein
has 4-fold higher affinity for glucagon as compared with the wild type
glucagon receptor. This increased affinity is reflected in a 4-fold
slower rate of dissociation of 125I-glucagon from the
252-259 mutant protein as compared with the wild type receptor in
the absence of divalent cation.
Divalent cations have been shown to allosterically enhance the binding of agonists to many types of G protein-coupled receptors. Previous data suggest that divalent cations stimulate binding of 125I-glucagon to receptors in chicken liver membranes (9) but not from mammalian liver membranes (9, 10). Subsequently, Post et al. (11) reported that Mg2+ cations stimulated binding of 125I-glucagon to saponin-permeabilized canine hepatocytes 3-fold. In contrast to our data, calcium did not mimic the effects of magnesium in this system.
Both calcium and magnesium stimulate binding of
125I-glucagon to the human glucagon receptor. The increased
binding is due to a 3-fold increase in affinity for glucagon in the
presence of divalent cation. Inclusion of divalent cation decreases the
rate of dissociation of 125I-glucagon from the receptor
from 0.04 to 0.01 min1 at room temperature. The receptor
in the absence of magnesium does not simply reflect the affinity of
agonist in the absence of G protein coupling, since inclusion of
Gpp(NH)p in the incubation increases the rate of dissociation of
125I-glucagon to 0.29 min
1. This conclusion
is also supported by the effects of calcium, a cation that does not
alter the affinity of G proteins for receptors. These data suggest that
the primary effect of divalent cation is not on G protein coupling, but
on the conformation of the receptor. These data also suggest that at
least three conformations of the glucagon receptor exist in the
membrane, reflected in the differences in affinity observed in the
absence or presence of divalent cation, and in the fully uncoupled
state (i.e. in the presence of excess Gpp(NH)p).
Gpp(NH)p does not inhibit the binding of 125I-glucagon to
the 252-259 mutant protein, and these data together with the lack of glucagon-activated adenylyl cyclase activity are consistent with a
lack of G protein coupling to this protein. In biogenic amine
receptors, deletions within the second intracellular domain alter
activation, and chimeric receptors in this region have altered specificity for G proteins, suggesting that this region is directly involved in coupling of receptor and G protein (4). However, divalent
cation does not alter the affinity of 125I-glucagon for
this mutant as measured by binding isotherms or dissociation kinetics.
Since the affinity of the
252-259 mutant protein for
125I-glucagon and the rate of dissociation of
125I-glucagon from this protein are identical to the
affinity and rate of dissociation of 125I-glucagon from the
wild type receptor in the presence of divalent cation, these data
strongly suggest that this deletion locks the protein into the
conformation that is normally promoted by the binding of divalent
cation to the receptor. These data strongly suggest that the glucagon
receptor contains a divalent cation binding site that allosterically
regulates glucagon binding and receptor coupling. As such, the
252-259 mutant protein will be a useful tool for assessing the
differences in binding interactions among the different affinity states
of the glucagon receptor.