(Received for publication, April 16, 1997, and in revised form, June 4, 1997)
From To study the vasopressin receptor domains
involved in the hormonal binding, we synthesized natural and modified
fragments of V1a vasopressin receptor and tested
their abilities to affect hormone-receptor interactions. Natural
fragments mimicking the external loops one, two, and three were able to
inhibit specific vasopressin binding to V1a receptor. In
contrast, the natural N-terminal part of the V1a
vasopressin receptor was found inactive. One fragment, derived from the
external second loop and containing an additional C-terminal cysteine
amide, was able to fully inhibit the specific binding of both labeled
vasopressin agonist and antagonist to rat liver V1a
vasopressin receptor and the vasopressin-sensitive phospholipase C of
WRK1 cells. The peptide-mediated inhibition involved specific
interactions between the V1a receptor and synthetic V1a vasopressin receptor fragment since 1) it was dependent
upon the vasopressin receptor subtype tested
(Ki(app) for the peptide: 3.7, 14.6, and 64.5 µM for displacing [3H]vasopressin
from rat V1a, V1b, and V2
receptors, respectively; 2) it was specific and did not affect sarcosin
1-angiotensin II binding to rat liver membranes; 3) it was not mimicked
by vasopressin receptor unrelated peptides exhibiting putative
detergent properties; and 4) no direct interaction between
[3H]vasopressin and synthetic peptide linked to an
affinity chromatography column could be observed. Such an inhibition
affected both the maximal binding capacity of the V1a
vasopressin receptor and its affinity for the labeled hormone,
depending upon the dose of synthetic peptide used and was partially
irreversible. Structure-activity studies using a serie of synthetic
fragments revealed the importance of their size and cysteinyl
composition. These data indicate that some peptides mimicking
extracellular loops of the V1a vasopressin receptor may interact with the vasopressin receptor itself
and modify its coupling with phospholipase C.
Vasopressin (AVP),1 a
small polypeptidic neurohypophysial hormone, exerts different
biological effects in mammals. At the periphery, its major
physiological role is played in regulating water and solute
excretion by the kidney. This hormone is also involved in blood
pressure control, platelet aggregation, corticotropin and aldosterone
secretion (by the adenohypophysis and the adrenals, respectively),
hepatic glycogenolysis, and uterine motility (for review, see Refs. 1
and 2). In the central nervous system, AVP is also involved in
interneuronal communication (3).
These distinct biological functions are mediated, in mammals, by at
least three distinct receptor subtypes: V2,
V1a, and V1b. V2 receptors,
involved in the antidiuretic response, are positively coupled to
adenylyl cyclase (4). V1a and V1b receptors,
involved in multiple peripheral responses and in corticotropin release, stimulate phospholipase C and activate protein kinase C (5, 6). Cloning
the different subtypes of receptors (7-9) confirmed that these
peptidic receptors belong to the family of the G-protein-coupled receptors.
Small amounts of vasopressin receptors in natural tissues and
difficulties in solubilizing and purifying vasopressin receptors (10)
led to restricted information on the topology of the hormonal binding
domain of AVP receptors. However, molecular biology and biochemical
approaches have recently headed the way to characterize the vasopressin
receptor binding domain. Using a tritium-labeled photoreactive
vasopressin agonist, Kojro et al. (11) demonstrated that
residues Arg106 and Thr102 present in the
second loop of the bovine kidney V2 receptor are involved
in AVP binding. Chini et al. (12) showed that
replacement of Tyr115 by an alanine, in the first loop of
the human V1a vasopressin receptor, greatly affects its
ligand selectivity toward a series of vasopressin analogues. More
recently, three-dimensional computer modeling of rat V1a
vasopressin receptor structure, combined with directed
site-mutagenesis experiments, has indicated that the transmembrane
domains of the receptor are also involved in the vasopressin binding
site (13). However, the precise location of vasopressin binding to the
three receptor isoforms remains incompletely characterized.
Another approach to the study of hormone-receptor interactions involves
the use of small synthetic peptides mimicking the sequence of the
supposed active region of the receptor. This approach allows the use of
pure and well characterized fragments, available in large amounts, to
determine their possible interaction with the specific ligand or with
the receptor itself. This approach has been successfully used in the
case of the thyrotropin-stimulating hormone receptor and luteotropin
human choriogonadotropin receptor, where synthetic peptides mimicking
the N-terminal extracellular sequence of the receptor were able to bind
the hormone (14, 15). Similarly, peptide fragments corresponding to the
N-terminal segment of the follitropin receptor were shown able to bind
this hormone (16). More recently, Howl and collaborators (17) also described the inhibitory properties of some synthetic peptides mimicking neurohypophysial hormone receptor. A 20-amino acid synthetic peptide derived from the tumor necrosis factor receptor inhibited the
binding and the cytolytic activity of the corresponding recombinant human hormone (18). This approach has also been successfully used in
characterizing hormonal receptor-G-protein interactions, such as
synthetic peptides mimicking the third intracellular loop of the
5HT1a receptor which prevent hormonal adenylyl cyclase inhibition (19), and also the interactions between other classes of
proteins like actin-tropomyosin and calponin (20).
To further delineate the vasopressin binding site of the
V1a vasopressin receptor, we elected to follow an approach
similar to that described above. Thus, we synthesized peptides
corresponding to extracellular regions of the rat V1a
vasopressin receptor (N-terminal sequence and hydrophilic extracellular
loops) and examined their ability to alter specific vasopressin binding
to V1a vasopressin receptor and to modify
vasopressin-stimulated phospholipase C activity. Our data indicate that
some synthetic peptides and particularly one fragment of the second
extracellular loop with an additional cysteine residue exhibited
antagonistic properties. We studied the mechanisms involved in such
inhibition.
Tritiated [Arg8]vasopressin (60 Ci/mmol) ([3H]AVP)and
myo-[3H]inositol (10-20 Ci/mmol) were
obtained from NEN Life Science Products; OH-LVA, a specific
V1a vasopressin antagonist (2000 Ci/mol), and Sar-AngII, an
angiotensin II antagonist (2000 Ci/mmol), were radioiodinated using the
IODO-GEN technique as described previously (21, 22). [Arg8]Vasopressin was obtained from Bachem. The peptides
Nt-(50-79) (deriving from the N-terminal part of the bovine endothelin
A receptor) and GLP (a glucagon-like-peptide) were synthesized and characterized in the laboratory by solid phase peptide synthesis as
described previously (23, 24). Dowex AG1-X8 (100-200-mesh), chloride
form, was obtained from Fluka. Affinity chromatography was performed
with Affi-GelR 10 gel from Bio-Rad. All other chemicals
were of A-grade purity.
All
peptides were obtained by solid phase peptide synthesis, using two
different procedures. The fluoromethoxycarbonyl strategy consisted in
using a continuous flow procedure in an automated solid-phase
synthesizer (PepSynthesizer 9050, from Perseptive Biosystems,
Millipore) (25). The resin was Fmoc-Amino-Acid-PAL-PEG-PSR
from Perseptive Biosystems, Millipore (1 g of 0.2 mmol
NH2/g). We used free Peptides were purified by reverse phase HPLC. Analytical HPLC was
performed using a C18 cartridge and linear
CH3CN/H2O/0.1% CF3COOH (v/v)
gradients as described previously (28). Preparative runs were performed
using linear gradients of CH3CN in H2O,
acidified (pH 2-3) by 0.1% CF3COOH. The purified peptide
structure was checked via amino acid analysis, electrospray ionization
mass spectrometry (ESI-MS, VG FISONS, TRIO 2000) as described
previously (29) or by fast atom bombardment (FAB-MS, JEOL, SX 102). For
the ESI-MS method, determination of the peptide true molecular weight
from the raw m/z data was performed using VG TRIO 2000 mass
spectrometer with Maxent software (29). For the FAB-MS, determination
of the true molecular weight was derived from the (MH+)
peak of the fragmentation profile. Peptide quantification was performed
by weighing, assuming CF3COOH as counter-ion in the lyophilized compounds, but without considering any hydration of the
peptides.
The e2-(194-218)C peptide
was treated with iodoacetamide, as described previously (30), to block
its cysteinyl residues. This technique was preferred to
N-ethylmaleimide blockade, since it introduced a reduced
steric modification with no additional charge on the native
peptide.
Animals used in this study
were female Wistar rats (160-180 g body weight). Purified plasma
membranes from liver were prepared according to the procedure of
Neville (31) up to step 11. They were stored in liquid nitrogen. Crude
plasma membranes from the inner kidney medulla or from the
adenohypophysis were prepared according to Butlen (4) and Jard (6),
respectively. Briefly, the tissues were homogenized at 4 °C in an
isotonic buffer containing 250 mM sucrose, 5 mM
Tris-HCl, pH 7.4, 3 mM MgCl2, 1 mM
EDTA, 0.1 mM phenymethylsulfonyl fluoride, using a
glass/glass Potter-Elvehjem. The extracts were then centrifuged at
4 °C for 15 min at 3500 rpm. The supernatants were discarded and
pellets resuspended in a large volume of hypotonic buffer, the
composition of which was similar to the one described above, but
without sucrose. After a 20-min incubation at 0 °C, extracts were
centrifuged under the same conditions. Resulting pellets were
resuspended in the hypotonic medium and centrifuged again. These
pellets were collected, resuspended in hypotonic medium (1 mg of
protein/ml), and used immediately for binding experiments.
WRK1 cells, a rat mammary tumor cell line,
were cultured as described previously (32). Briefly, cells were plated
at a density of 105 cells/dish in a modified minimum
essential medium, containing 5% calf serum, 2% rat serum, 290 mg/ml
glutamine, 100 units/ml penicillin, and 100 mg/ml streptomycin. They
were cultured at 37 °C in a humidified atmosphere (5%
CO2, 95% air). Two days after plating, the culture medium
was removed and replaced by a fresh one containing 1 µCi/ml
myo-[3H]inositol and cells were used 2 days
thereafter.
[3H]AVP binding to plasma
membrane preparations was performed as described previously (4, 6, 33).
Briefly, kidney, adenohypophysis or liver membrane preparations (30-60
µg of protein/assay) were incubated 1 h at 37 °C in 200 µl
of a buffer containing 50 mM Tris-HCl, pH 7.4, 3 mM MgCl2, 1 mg/ml bovine serum albumin, 0.1 mM phenymethylsulfonyl fluoride, 1 nM
[3H]AVP, and increasing amounts of vasopressin receptor
fragments (total binding determinations). An excess of unlabeled AVP (1 µM) was also added in the incubation medium to determine
the nonspecific binding. The reaction was initiated by adding the
membranes, stopped by addition of 4 ml of cold washing solution (10 mM Tris-HCl, pH 7.4, 3 mM MgCl2),
and immediately filtered onto Whatman glass fiber (0.8-1.2 µm)
previously presoaked 2 h with 10 mg/ml BSA. The filters were then
rinsed three times with 4 ml of the washing solution and the remaining
radioactivity, measured by liquid scintillation spectrometry. The
specific binding was calculated as the difference between the total and
nonspecific values. 125I-OH-LVA binding was performed as
described previously (21). Briefly, 1-2 µg of rat liver membranes
were incubated 1 h at 37 °C in 200 µl of the same incubation
medium as the one described above. [3H]AVP was replaced
with a single concentration of 125I-OH-LVA
(concentration-displacement experiments) or increasing amounts of
125I-OH-LVA (concentration-dependent binding
experiments). The concentration of 125I-OH-LVA used in the
displacement experiments was around 20 pM, a value
corresponding to the Kd of this labeled analogue for
the rat liver V1a vasopressin receptor (21). Binding
experiments were carried out as described for [3H]AVP,
except that radioactivity was measured using a To verify if our binding assay conditions outlined the kinetic
equilibrium conditions between the synthetic peptides and the hormonal
receptor, we tested the influence of a preincubation between the
synthetic peptides and the plasma membranes, prior addition of
125I-OH-LVA. As shown in Fig. 1A, a 30-min
preincubation at 37 °C did not modify the ability of e2-(194-218)C
to inhibit 125I-OH-LVA specific binding. These results
validate the experimental protocol chosen for measuring the
interactions between the synthetic peptides and the hormonal receptor.
To further characterize such interactions, we also calculated the Hill
coefficient for the e2-(194-218)C peptide, as described previously
(33). As illustrated on Fig.
1B, the Hill coefficient was
found independent from the preincubation conditions. Its value
(1.14 ± 0.07, n = 7) was near to unit, suggesting
a single site binding interaction. Similar results were also obtained
with the same synthetic peptide on rat liver plasma membrane using
[3H]AVP as radioligand (Hill coefficient = 1.11 ± 0.09, n = 5) (data not shown).
125I-Sar-AngII binding was performed on
rat liver membranes as described for 125I-OH-LVA. The
concentration of labeled hormone used in this assay was around 0.2 nM, a value similar to the Kd of this analogue for rat liver AngII receptor (22).
As described
previously (34), myo-[3H]inositol-prelabeled
WRK1 cells were incubated for 45 min, at 37 °C in a culture medium deprived of myo-[3H]inositol and sera, washed
twice with 1.5 ml of phosphate-buffered saline (PBS) and then
preincubated 15 min at 37 °C with 0.7 ml of PBS supplemented with: 1 g/liter glucose, 10 mM LiCl, 1 mg/ml BSA, and the
vasopressin receptor fragments tested. AVP (1 nM) was then
added to the incubation medium for an additional 6-min period.
Perchloric acid (5% final concentration) together with 0.1 ml of BSA
(20 mg/ml) were added to the incubation medium to stop the reaction.
Cells were scraped and cellular extracts centrifuged. Labeled mono-,
bis-, and triphosphate inositol present in the supernatant were
separated from inositol and glycerophosphoinositides using Dowex AG1-X8
columns and measured by liquid scintillation spectrometry as described
previously (34).
Data presented are the mean of triplicate
determinations. The standard error (S.E.) associated to each
experimental value never exceeded 10% both for binding and inositol
phosphate accumulation measurements. The number of distinct experiments
performed (n) is indicated in parentheses. Determination of
the apparent dissociation constant
(Ki(app)) for the vasopressin receptor
fragments was calculated as described previously (33), assuming a
direct competition between the vasopressin receptor fragment and
labeled AVP or OH-LVA molecules for the vasopressin receptor binding
sites. Briefly, the concentration of vasopressin fragment leading to half-maximal specific binding inhibition of labeled hormone
(ED50) was determined by concentration-displacement
experiments and its Ki(app), calculated
according to the Cheng-Prusoff equation: Ki(app) = ED50 × Kd/(Kd + [H*]), where Kd is the dissociation constant of the labeled
ligand used for the vasopressin receptor considered and [H*] the
concentration of the labeled hormone used in the assay.
We prepared synthetic fragments of the rat
V1a vasopressin receptor from the primary sequence
published (8). These peptides (from 13 to 51 residues) mainly
corresponded to hydrophilic extracellular sequences of the rat
V1a receptor. The C termini of the synthetic peptides, for
which length did not exceed 26 residues, were amidated to suppress the
carboxyl negative charge not present in the corresponding domain of the
native protein. The putative importance of cysteine residues in the
receptor binding domain (35), led us to synthesize "modified
peptides." They were either cysteinylated (addition of a C-terminal
extra-cysteinyl residue) or alkylated by iodoacetamide (blockade of
cysteinyl residues).
The sequences and the abbreviations of the peptides are summarized in
Table I. Peptide purity was determined
by analytical HPLC and evaluated to be higher than 98%, as can be
shown for e2-(194-218)C peptide (Fig. 2,
inset). Moreover, amino acid analysis of each peptide is in
good agreement with its sequence (data not shown). Finally, the mass of
each peptide was confirmed using mass spectrometry. As shown in Table
II, the experimental mass values found
for each peptide corresponded to the calculated mass. Fig. 2
illustrates a typical result of mass spectrometry for e2-(194-218)C synthetic fragment.
Table I.
Positions and sequences of natural and modified synthetic fragments of
rat V1a vasopressin
receptor
Table II.
Mass spectrometry characterization of synthetic peptides
INSERM U469 and ¶ CNRS UPR 9023,
CNRS UPR 9008,
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
Chemicals
CO2H amino acids (0.6 mmol) with temporary protection on the
NH2 and the
side-chain protections as follows: pentamethylchromanesulfonyl for Arg;
trityl for Cys, Asn, and Gln; tertiobutyloxycarbonyl for Lys;
tertiobutyl ether for Ser, Thr, and Tyr; tertiobutyl ester for Asp and
Glu. The coupling agent used was
2-(1H-benzotriazole-1-yl)-1,1,3,3-tetramethyluronium tetrafluoroborate
(0.6 mmol) and diisopropylethylamine as base (1.2 mmol). After the
completion of the synthesis, the peptides were deprotected and cleaved
from the support using K reagent (CF3COOH/phenol/water/thioanisole/ethanedithiol,
85/5/2.5/2.5, v/v) for 0.5-4 h, depending on the presence or not of
Arg(pentamethylchroman sulfonyl) in the sequence. The
tertiobutyloxycarbonyl strategy was carried out using either an
automated solid-phase synthesizer (Applied Biosystems 430A) or a manual
device as described previously (26). The resin used was a
p-methylbenzhydrylamine polystyrene from Bachem (1 g of 1.1 mmol NH2/g for automatic synthesis, and 3 g of 0.6 mmol NH2/g of titratable amine for manual synthesis). We
used
CO2H amino acids (2.2 and 3.6 mmol for automatic
and manual synthesis, respectively) with the following side-chain protections: p-toluenesulfonyl for Arg;
p-methylbenzyl for Cys, benzyl for Ser and Thr, cyclohexyl
for Glu and Asp, o,o-dichlorobenzyl for Tyr, and
o-chlorobenzyloxycarbonyl for Lys with
(benzotriazolyloxy)tripyrrolidinophosphonium hexafluorophosphate as
coupling agent (2.2 or 3.6 mmol for automatic or manual synthesis,
respectively) (27) and diisopropylethylamine as base (4.4 or 7.2 mmol
for automatic or manual synthesis, respectively). The coupling
efficiency was usually evaluated by the colorimetric Kaiser test. In
this strategy, the deprotection was performed by
CF3COOH/CH2Cl2/ethanedithiol
(50/47/3, v/v) for 2 and 28 min, and finally 60 min in
HF/anisole/dimethylsulfide (9/0.8/0.2, v/v).
counter.
Fig. 1.
Influence of preincubation on the interaction
between V1a vasopressin receptor and e2-(194-218)C
peptide. Panel A, 1.5 µg of rat plasma membrane were
preincubated 30 min at 37 °C with increasing amounts of
e2-(194-218)C synthetic peptide () or vehicle (control). Then,
either 20 pM 125I-OH-LVA (total binding) or 20 pM labeled ligand plus 1 µM unlabeled AVP
(nonspecific binding) were added in the incubation medium and the
reactions allowed to proceed for another 60-min period at the same
temperature. In another set of experiment (
), rat liver membranes
were preincubated 30 min without synthetic peptide. Then, increasing
amounts of e2-(194-218)C peptide were added together with 20 pM 125I-OH-LVA (total binding) or 20 pM 125I-OH-LVA plus 1 µM
unlabeled AVP and incubation proceeded for 60 additional min.
Radioactivity found associated to plasma membrane was determined by
filtration as described under "Materials and Methods." Specific
binding were calculated in each condition (total-nonspecific binding)
and expressed as percent of control specific binding (100% = 2900 ± 150 dpm/assay for membrane preincubated with or without
e2-(194-218)C peptide). Results are the mean of triplicate determinations from a single experiment representative of three. Panel B, data from panel A were plotted as
described previously (33) for the determination of the Hill
coefficient. Bo, control specific binding;
B, specific binding measured in the presence of a given
concentration (I) of e2-(194-218)C peptide; H,
concentration of 125I-OH-LVA used in the assay;
Kd, dissociation constant of the rat liver
vasopressin receptor for 125I-OH-LVA; [I],
concentration of synthetic peptide.
[View Larger Version of this Image (16K GIF file)]
Sequences and Purity of V1a Vasopressin Receptor
Fragments
Position Abbreviations Sequences
Fig. 2.
Purity and mass measurement of e2-(194-218)C
peptide. The deconvoluted electrospray mass spectrum of
e2-(194-218)C peptide is illustrated. The calculated mass
corresponding to the principal peak was 3002 Da. The inset
illustrates the reverse-phase HPLC profile of the same synthetic
peptide performed on a C18 column, monitored at 214 nm,
using a 0-40% acetonitrile gradient in 0.1% trifluoroacetic
acid.
[View Larger Version of this Image (23K GIF file)]
Abbreviations
Mass spectrometry data
Theoretical mass value
Experimental mass value
ESI-MS
FAB-MS
Nt-(1-51)
5434
5432
e1-(111-124)
1827
1828
e3-(319-335)
2111
2112
e2-(206-218)
1595
1596
e2-(205-218)
1697
1698
e2-(194-218)
2899
2895
e2-(194-218)S
2883
2885
e2-(194-218)C
3002
3004
e2-(203-218)C
2043
2045
e2-(201-218)C
2272
2274
e2-(194-218)C(cam)2
3118
3117
HPLC control experiments performed on the synthetic fragment e2-(194-218)C, containing cysteinyl residues, showed that it was not able to dimerize or to be cyclized (data not shown).
Influence of V1a Vasopressin Receptor Fragments on the Iodinated OH-LVA Specific Binding to Rat V1a Vasopressin ReceptorFig. 3 summarizes the
effects of some receptor fragments on the specific binding of
125I-OH-LVA on rat liver plasma membranes, a preparation
that contained only the V1a vasopressin receptor subtype
(33). Short fragments of the first loop (e1-(111-124)), of the second
loops (e2-(205-218), e2-(206-218), e2-(194-218)), and of the third
loop (e3-(319-335)) inhibited, in a
concentration-dependent fashion, the specific binding of
125I-OH-LVA with an apparent inhibition constant ranging
between 7 and 61 µM (see Table
III). Finally, the N-terminal part of the V1a vasopressin receptor, Nt-(1-51) peptide, had no effect
even up to 300 µM. From these results, it appears that
among the V1a vasopressin receptor fragments tested, the
peptides corresponding to the second extracellular loops were the most
active. More interestingly, their size and cysteine composition seemed
to represent crucial parameters for their activities. Thus, increasing
the size of peptide e2-(205-218) resulted in peptide e2-(194-218)
with a reduced activity (see Table III). The importance of the cysteine
205 residue in e2-(194-218) is further evidenced in Fig.
4. Its replacement by a serine in
e2-(194-218)S reduced by at least 10-fold the peptide activity.
Moreover, peptide e2-(205-218), which contained only one N-terminal
cysteine more than e2-(206-218), was found more potent (see Table
III).
|
Influence of Cysteinyl V1a Vasopressin Receptor Fragments on the Iodinated OH-LVA Specific Binding to Rat V1a Vasopressin Receptor
As the e2-(205-218) peptide was the most potent in inhibiting iodinated OH-LVA specific binding and as its activity was due in part to its N-terminal cysteine, we synthesized new peptides differing by their size or cysteine composition. The sequences of these new fragments are illustrated in Table I. The most striking finding is illustrated in Fig. 4. The elongation of e2-(194-218) by a single cysteine residue in the C-terminal position (e2-(194-218)C) increased its activity by 60-fold (Ki(app) of 0.55 µM) (Table III). The importance of the two cysteines of e2-(194-218)C peptide is further illustrated in Fig. 4. e2-(194-218)C(cam)2, deriving from e2-(194-218)C by iodoacetamide treatment which blocked its cysteinyl residues, exhibited a reduced activity (Ki(app) larger than 80 µM). Reducing its sequence on its N-terminal part led to e2-(201-218)C and e2-(203-218)C, which were found to be 120- and at least 200-fold less active than e2-(194-218)C despite a similar cysteinyl residue composition (Table III).
Influence of Active V1a Vasopressin Receptor Fragments on the [3H]AVP Specific Binding to Rat V1a Vasopressin ReceptorAs the agonist and antagonist binding sites
are probably distinct, we compared the activity of some of the most
active vasopressin receptor fragments on the specific binding of
125I-OH-LVA (a vasopressin linear antagonist) and
[3H]AVP (the natural agonist) to the V1a
vasopressin receptor. e2-(194-218), e2-(194-218)C, and e2-(205-218)
peptides were able to fully inhibit the specific [3H]AVP
binding to rat liver membranes in a concentration-dependent manner (Fig. 5). The rank order of
potency to inhibit [3H]AVP binding was found to be
similar to that observed for the inhibition of 125I-OH-LVA.
However, the e2-(194-218)C peptide was found to be more active in
displacing the iodinated OH-LVA than the tritiated AVP (Ki(app) = 0.55 and 3.7 µM, respectively) (Table III). In contrast, e2-(205-218)
and e2-(194-218) peptides exhibited a similar efficiency in
inhibition of both antagonist and agonist specific binding (Table
III).
Selectivity of the e2-(194-218)C Fragment on the [3H]AVP Specific Binding to Various Vasopressin Receptor Subtypes
To further test the specificity of the e2-(194-218)C
peptide, we tested its ability to inhibit specific hormone binding on the two other vasopressin receptor subtypes, earlier characterized: V1b present in rat adenohypophysis (6) and V2
present in rat kidney inner medulla (4). We also compared these values
to those obtained with the rat liver V1a vasopressin
receptor. For these studies, we used [3H]AVP as
radioligand, since this hormone exhibited the same affinity for all
receptor subtypes studied (2, 6). As illustrated in Fig.
6, the e2-(194-218)C peptide was found
to be more active to inhibit [3H]AVP binding to the
V1a vasopressin receptor than to the other subtypes.
However, it entirely suppressed the specific binding to the
V1b receptor, but with a lower potency
(Ki(app) = 3.7 ± 1.4, n = 3 and 14.6 ± 4.5 µM,
n = 4 for V1a and V1b
receptors, respectively). It also weakly inhibited the
[3H]AVP binding to the V2 vasopressin
receptor (Ki(app) = 64.5 ± 3.5 µM, n = 4).
Nature of the Interaction between the e2-(194-218)C and the V1a Vasopressin Receptor
To further study the
mechanisms by which e2-(194-218)C peptide inhibited the hormone
specific binding, we analyzed the effects of this peptide on the
binding parameters of 125I-OH-LVA to rat liver membranes.
As shown in Fig. 7, and as described previously (21), 125I-OH-LVA interacts with a single class
of vasopressin receptors, present in rat liver membranes. The addition
of e2-(194-218)C in the incubation medium drastically reduced the
specific binding whatever the concentration of iodinated antagonist
used. Scatchard representations of dose-dependent binding
experiments indicated that e2-(194-218)C affected both the affinity of
125I-OH-LVA for the V1a vasopressin receptor
and its maximal specific binding capacity. As illustrated in Table
IV, when membranes were incubated with
3.16 µM e2-(194-218)C peptide, the apparent affinity was
reduced by 2.3-fold. The maximal binding capacity
(Bmax) was slightly reduced, but this effect was
not statistically significant. Increasing the amounts of synthetic
peptide in the incubation medium up to 10 µM greatly
increased the Kd by 7-fold and reduced the
Bmax significantly (by 2-fold). Such results
indicate that e2-(194-218)C peptide inhibits 125I-OH-LVA
in an uncompetitive fashion, which suggests interaction between the
receptor and the peptide. We also tested the reversible nature of the
interaction between e2-(194-218)C fragment and the V1a
vasopressin receptor. For this purpose, rat liver membranes were first
preincubated with or without a concentration of synthetic peptide
allowing an almost maximal inhibitory effect and washed to eliminate
the synthetic peptide. The kinetic binding parameters of the
V1a vasopressin receptor were then measured in control and
preincubated membranes. As illustrated in Fig.
8A, the inhibitory effect of
e2-(194-218)C peptide still persisted even after washing the membranes
previously incubated with the synthetic receptor fragment. However, at
variance with membrane directly incubated with the synthetic receptor
fragment, only the effect on the maximal binding capacity was observed.
Fig. 8B also showed that membranes preincubated with
e2-(194-218)C peptide and further washed were still sensitive to the
synthetic receptor fragment. A new addition of e2-(194-218)C peptide
always strongly reduced the remaining specific binding of
125I-OH-LVA. Altogether, these results indicate that the
interactions between e2-(194-218)C peptide and the V1a
vasopressin receptor are partly irreversible.
|
Influence of V1a Vasopressin Receptor Fragments on Vasopressin-stimulated Inositol Phosphate Accumulation
As the
V1a vasopressin receptor subtype is positively coupled to
phospholipase C activation, we tested the activity of some of the most
active V1a vasopressin receptor fragments characterized above. Such studies were performed on WRK1 cells, which expressed, as
rat liver membranes only, a V1a vasopressin receptor
tightly coupled to phospholipase C (32, 34). As illustrated in Fig. 9, e2-(194-218), e2-(194-218)C, and
e2-(205-218) peptides up to 300 µM did not stimulate
basal inositol phosphate accumulation. However, they were able to fully
inhibit in a concentration-dependent manner the
vasopressin-stimulated inositol phosphate accumulation. As expected,
e2-(194-218)C peptide was the most potent in inhibiting the
vasopressin effect (Ki(app) = 0.44 ± 0.07 µM, 3 distinct experiments). The two other
peptides, e2-(205-218) and e2-(194-218), were less active
(Ki(app) = 23 and 84 µM,
respectively). This rank order of potency corresponded to those found
for binding studies (Table III).
Specificity of V1a Vasopressin Receptor Fragment Activity
To further demonstrate that peptide e2-(194-218)C
specifically inhibited the binding of 125I-OH-LVA to
V1a vasopressin receptor, we performed the following control experiments. First, we tested the ability of a 30-amino acid
peptide derived from the N-terminal part of the endothelin A receptor
(peptide Nt-(50-79)) to inhibit specific vasopressin binding. This
peptide, containing a cysteinyl residue and a size similar to
e2-(194-218)C fragment with no sequence homology did not alter the
binding of 125I-OH-LVA on rat liver membranes, even tested
at 30 µM. Under the same experimental conditions,
e2-(194-218)C peptide almost completely inhibited the
125I-OH-LVA binding (Fig.
10A). Second, we also tested
the influence of GLP, known to exhibit amphiphilic properties due to
its helicoidal structure (36), to exclude any nonspecific putative
detergent properties of the hormonal receptor fragments. As illustrated on Fig. 10A, GLP did not inhibit 125I-OH-LVA
binding at a maximal dose of 30 µM. Third, we checked that e2-(194-218)C peptide was not able to alter the binding of 125I-Sar-AngII to the angiotensin II receptor also present
in rat liver membranes. As seen on Fig. 10B, under
experimental conditions where 125I-OH-LVA specific binding
was completely suppressed (30 µM e2-(194-218)C), no
significant modification of 125I-Sar-AngII was observed in
the same plasma membrane sample. Fourth, to verify that there is no
direct interaction between the e2-(194-218)C peptide and the labeled
hormone, we compared the elution profiles of [3H]AVP
either on an Affi-GelR column where e2-(194-218)C peptide
was covalently fixed by its N-terminal part or on a control
Affi-GelR column. Whatever the column used, the two elution
profiles were identical. All the radioactivity was recovered in a
single symmetrical peak coeluting with blue dextran, a marker of the
column void volume (data not shown). Control experiments that validate
this negative experiment have been previously published by Pradelles and collaborators. They demonstrated that the well known
protein/protein interactions between AVP and neurophysin still persist
even if neurophysin was covalently linked to the Sepharose column (37). This property was routinely used in the laboratory to purify tritiated vasopressin.
Vasopressin receptors belong to the G-protein-coupled receptor family exhibiting an N-terminal part, seven transmembrane domains connected by three extracellular loops and three intracellular loops, and possessing a C-terminal intracellular tail (38). Only recently, some data have helped to identify the receptor functional domain underlying binding of vasopressin. As was earlier established for the many G-protein-coupled receptors, some specific regions of transmembrane helices of the V1a vasopressin receptor are responsible for vasopressin anchoring (39, 40). However, additional studies have also demonstrated that the external loops of V1a vasopressin receptors may also play a role in these mechanisms. Mutation of one amino acid of the first extracellular loop of the V1a vasopressin receptor deeply affects its selectivity for two series of vasopressin analogues (12), photoaffinity labeling of V2 vasopressin receptor pointed to the second extracellular loop as a potential binding site for the hormone (11), and a peptide strategy suggested that the first extracellular loop of the V1a vasopressin receptor may represent a putative binding site for radioligands (17).
To further characterize the interaction between the vasopressin and the extracellular loops of the V1a vasopressin receptor, we synthesized a series of peptides mimicking these hydrophilic loops and tested their abilities to interfere with hormonal binding. Results presented in this study indicate that the peptide mimicking the N-terminal part of the V1a receptor is unable to alter specific agonist or antagonist binding to this receptor. This suggests either that this receptor domain is not concerned in binding or the structural determinants such as glycosylation or three-dimensional organization are lacking for this longer peptide. In contrast, fragments of V1a receptor exhibiting sequences similar to those of the external loops one, two, and three are active. They inhibit in a concentration-dependent manner the specific [3H]AVP and 125I-OH-LVA binding to the rat V1a vasopressin receptor with Ki(app) ranging between 10 and 60 µM (Table III).
To study structure-activity relationships of the most active peptide found, e2-(205-218), we increased its size and modified its cysteinyl composition by adding a C-terminal cysteine residue. These modifications were based upon the following arguments. 1) Hydropathy profile of the primary receptor sequence indicated that the second extracellular loop, determined by hydrophilic amino acid composition, is larger than the e2-(205-218) sequence. 2) Cysteinyl residues of rat V2 and, to a lesser extent, those of rat V1a vasopressin receptors play an important role in AVP binding (35). 3) Most of our active fragments exhibited a cysteinyl residue. 4) Recent representations of the vasopressin receptor favored the existence of a disulfide bridge between the cysteinyl residues of the extracellular loops one and two (13). Results obtained indicate that increasing the size of the most active peptide, e2-(205-218), leads to e2-(194-218), 5-fold less active on 125I-OH-LVA binding assays on rat liver membranes. In contrast, increasing the size of the inactive modified e2-(203-218)C peptide leads to e2-(201-218)C and e2-(194-218)C peptides with enhanced activity. Addition of 2 or 9 amino acids to the N-terminal part of the inactive fragment, e2-(203-218)C, led to peptides active at the 50 µM and 1 µM concentration range, respectively. Moreover, the presence of cysteine residues in the synthetic fragments also greatly improve their activities; e2-(194-218)C peptide, which differs from e2-(194-218) peptide by only one cysteinyl residue in the C-terminal part, is 60-fold more active. On the opposite, blocking its cysteinyl residues leads to a nearly inactive fragment. All together, these data suggest that the size and the cysteinyl composition of the most active peptide, e2-(194-218)C, constitute crucial parameters for its activity.
At least two reasons may explain why e2-(194-218)C peptide inhibits specific vasopressin binding. This receptor fragment may interact either with the receptor itself or with the hormone. In both cases, such interactions prevent the hormonal binding to its specific receptor. Experimental data presented in this study favor the first hypothesis for the following reasons. 1) No direct interaction between AVP and e2-(194-218)C peptide could be observed by an affinity chromatography approach. [3H]AVP was not retained on an Affi-Gel column on which e2-(194-218)C peptide was covalently linked, as neurophysin was retained, using the same experimental approach (37). 2) A direct interaction between [3H]AVP and e2-(194-218)C peptide would lead to identical inhibitory effects on the three vasopressin receptor subtypes, since they possessed the same affinity for [3H]AVP (Kd = 3.5 ± 0.2, 1.0 ± 0.2, and 1.5 ± 0.3 nM, for V1a, V1b, and V2 vasopressin receptors present on rat liver, rat pituitary, and rat kidney membranes, respectively). Data presented in Fig. 6 clearly indicate that it is not the case. e2-(194-218)C peptide preferentially inhibited the binding of [3H]AVP on the V1a vasopressin receptor and is 4- and 17-fold less active on the V1b and the V2 vasopressin receptors, respectively. 3) The inhibitory effects of e2-(194-218)C peptide on specific 125I-OH-LVA binding were partially irreversible. They still persist even if rat liver plasma membranes preincubated with the synthetic receptor fragment were washed before the hormonal binding assay (Fig. 8).
The effects we observed with the synthetic peptides mimicking the rat V1a vasopressin receptor on vasopressin binding are specific and could not be explained by their putative detergent properties or cysteinyl composition. Indeed, we demonstrated that 30 µM e2-(194-218)C peptide, which completely inhibited 125I-OH-LVA binding on rat liver plasma membranes, did not affect, under the same experimental conditions, the specific binding of 125I-Sar-AngII (Fig. 10B). Similarly, we tested the effects of unrelated V1a vasopressin receptor peptides exhibiting either cysteine residue (Nt-(50-79), a fragment of the endothelin A receptor) or an helicoidal structure with potential detergent properties (GLP, glucagon-like peptide) on 125I-OH-LVA binding. As observed on Fig. 10A, neither Nt-(50-79) nor GLP peptide inhibited specific 125I-OH-LVA binding to rat liver membranes even tested at 30 µM.
The mechanisms by which e2-(194-218)C peptide specifically inhibits
hormonal binding to rat V1a vasopressin receptor are
probably complex, since we observed both an irreversible loss of the
maximal binding capacity and a reversible alteration of the affinity of the vasopressin receptor for its ligand. The substantial loss of
specific binding site observed for high concentrations of
e2-(194-218)C peptide (Table IV) is probably due to covalent
interactions between the receptor itself and the synthetic peptide via
cysteinyl residues since 1) both the V1a vasopressin
receptor and the e2-(194-218)C peptide exhibited cysteinyl residues
(Ref. 35 and Table I), 2) replacement of the Cys205 residue
of the natural receptor fragment e2-(194-218) by a serine led to a
compound (e2-(194-218)S) more than 10-fold less active, 3) blockade of
the two cysteinyl residues of e2-(194-218)C peptide led to peptide
e2-(194-218)C(cam2) exhibiting a Ki of at least 150-fold higher, and 4) the effect of e2-(194-218)C peptide on 125I-OH-LVA binding to rat liver membranes is partly
irreversible (Fig. 8). The modulation of the affinity of the
V1a vasopressin receptor for its ligand induced by
e2-(194-218)C fragment also probably involved non-covalent
protein-protein interactions between the synthetic peptide and the
receptor, since these effects were reversible (Fig. 8). The fact that
synthetic peptides devoid of any cysteinyl residue like e2-(206-218)
were active reinforces this assumption (see Table III). Probably, we
have synthesized a cysteinyl site-directed peptide able to interact
with the V1a vasopressin receptor and as a consequence to
block its ability to recognize its natural hormone and to transduce
intracellular second messenger generation. However, as shown in Fig.
10, the cysteine is not sufficient to confer activity to a peptide,
since a fragment of endothelin receptor of size similar to
e2-(194-218)C contains cysteine and yet did not inhibit
125I-OH-LVA binding. This indicates that sequence-specific
interactions are involved in the biological activity of the peptide
studied. Such results would imply that synthetic peptides mimicking
some specific regions of the extracellular domains of the
V1a vasopressin receptor are able to interact with
unidentified part of the V1a vasopressin receptor and block
its activity. This suggests that intramolecular protein-protein
interactions may represent a crucial parameter for vasopressin receptor
activity. Such a hypothesis was already verified by Ridge et
al. (41), who showed that co-expression of two or three
complementary fragments of rhodopsin in COS cells led to the expression
of a protein that reproduced the spectral properties of native
rhodopsin. Alternatively, our data would be also consistent with the
proposed importance of intermolecular protein-protein interactions
between two receptors domains of distinct proteins. Such specific
interactions may explain the formation of hormonal receptor dimers as
described previously for the muscarinic, the -adrenergic and the
glucagon receptors (42, 43). Similarly, the V1a vasopressin
receptor photolabeling, in rat liver membranes showed, by SDS-PAGE
autoradiography, higher molecular weight structures suggesting the
possibility of dimer formation (44). More interestingly, they may
represent a new mechanism to regulate their activities (45). Thus,
Bouvier and collaborators confirmed that
2-adrenergic
receptor may dimerize. They also demonstrate that a peptide mimicking
the transmembrane domain VI of this receptor prevents dimerization and
also inhibits
-adrenergic stimulation of adenylyl cyclase activity
(46). Maggio et al. (47) also demonstrated that
co-expression in COS cells of two muscarinic/adrenergic receptor
chimera, individually devoid of any binding activity, led to specific
active muscarinic and adrenergic binding sites.
In conclusion, the specific interaction evidenced in the paper between synthetic peptide mimicking the sequence of the V1a vasopressin receptor and the receptor itself strongly suggests the importance of intra- or intermolecular interactions between two receptor molecules and provides a new molecular basis to study their regulations.
We thank M. Passama and D. Bellenoue for drawings, M. C. Maraval and M. Chalier for English editing, Dr. J. Marie for 125I-Sar-AngII, and Dr. M. Bouvier, Dr. C. Chevillard, and Dr. R. Pascal for fruitful discussions.