Peptides,Vol. 13, pp. 7-11, 1992 Printed in the USA.
0196-9781/92 $5.00 + .DO Copyright© 1992 PergamonPressLtd.
Fluorescent Peptide Hormones: Development of High Affinity Vasopressin Analogues GILLES GUILLON,* DAVID BARBEAU,t WITOLD NEUGEBAUER,'~ SYLVAIN GUAY,t LINE BILODEAU,'~ MARIE-NOELLE BALESTRE,* NICOLE GALLO-PAYETJ; A N D E M A N U E L E S C H E R ~ "t
*Centre CNRS-INSERM de Pharmacologie-Endocrinologie, rue de la Cardonille, BP No 5, Montpellier, France 34094 Departement of tPharmacology and of SMedecine, Faculty of Medecine, UniversitO de Sherbrooke, Sherbrooke, Quebec, Canada J1H 5N4 R e c e i v e d 5 F e b r u a r y 1991 GUILLON, G., D. BARBEAU, W. NEUGEBAUER, S. GUAY, L. BILODEAU, M.-N. BALESTRE, N. GALLO-PAYET AND E. ESCHER. Fluorescentpeptidehormones:Developmentof higha2~nityvasopressinanalogues.PEPT1DES 13(1) 7-11, 1992.-Highly potent and specific peptide hormone analogues with fluorescent reporter groups are current research goals. Until now, however, only moderately potent analogues have been described. We report here several types of vasopressin (VP) analogues with different fluorophores attached to the peptide, In a first series, fluorophores were attached to the free ~ amino function of [desamino~-lysineg]VP(dLVP), producing agonistic analogues. In a second series, reporter groups were added to the N-terminal of open-chain antagonist structures. The biological activities of these analogues were assessed by two different sets of experiments: 1) The measurement of their binding affinities towards the Vla-vasopressin receptor subtype from WRKI cells or rat liver membrane preparations; 2) Their ability to stimulate the phospholipase C activity in WRK~ cells. As expected, a simple acylation of fluorophores to dLVP resulted in a considerable loss of affinity. If, however, the Lys~ side chain was extended through double Schiff-base formation with glutaraldehyde-ethylenediamine followed by reduction to an aminoalkyl aminoalkylamine, single fluorophores could be added without loss of affinity compared to VP. The open-chain analogues, on the other hand, while displaying weak affnity, nevertheless exhibited pure antagonistic behavior. Vasopressin
Vasopressin analogues
Fluorescent reporter groups
PEPTIDE hormones with reporter functions have been a key to success in peptide hormone endocrinology and pharmacology. Several hormone-receptor systems are now well characterized biochemically, due to the availability of such functionalized ligands. Although vasopressin (VP) is one of the longest known peptide transmitters, it has not yet attained this level of knowledge, although a great deal of knowledge concerning its structureactivity relationship is currently available. Research on the VP receptor has always been hampered due to the lack of receptor ligand with reporter functions. Among the plenitude of hormone analogues, only tritiated VP is available as a radioactive peptide with full agonistic properties, and, until now, no fluorescent analogue of VP has been reported with more than 5-10% of the affinity of the native hormone despite the considerable efforts in this direction which have recently been made (2,3,11,13). Highly fluorescent analogues with affinity and selectivity similar to natural VP would greatly facilitate the search for the VP receptors. Binding studies could then be carried out without the presence of radioactivity, and histological studies could be made
without the necessity for film exposure; colocalization and internalization of hormonal receptors would also become feasible. Furthermore, such fluorescent analogues could greatly facilitate receptor cloning studies, and VP receptor-expressing cells could be selected from nonexpressing cells by cell sorter techniques in the presence of such analogues. As a starting point for the synthesis of the agonistic structure, we chose [des-aminol-lysineS]Vp (dLVP). For the antagonistic structures, the recently reported noncyclic structures seemed to be almost ideally suited due to the possibility of N-terminal acylation with hydrophobic groups such as adamantylacetate (14). As fluorophores, we tested the classical fluoresceine and the pyrene residues. METHOD The chemicals used in the present study were obtained from the following sources: myo-[3H]inositol (10-20 Ci/mmol) from Commisariat fi l'6nergie atomique, France; [3H][ArgS]VP (70 Ci/mmol) from New England Nuclear; minimal essential me-
1Requests for reprints should be addressed to Emanuel Escher, Ph.D., Department of Pharmacology, Faculty of Medicine, Universit6 de Sherbrooke, Sherbrooke, Qu6bec, Canada J 1H 5N4.
8 dium (MEM) and Earle's salts from Gibco. Other chemicals were of A-grade purity.
WRKI Cell Culture WRK~ cells were established and cultured, with (inositol phosphates experiment) or without myo-[3H]inositol (binding experiment) as described previously (7,12). Cells were plated in 3.5 cm plastic dishes, at a density of(5-7) X 104 cells in minimal essential medium containing Earle's salts, fetal calf serum (5%, v/v), rat serum (2%, v/v), glutamine (2 mM), penicillin (100 units/ml), streptomycin (100 m g / m l ) a n d myo-[3H]inositol (2 uCi/ml). The medium was changed after 2 days and experiments were performed 3 days later. Before the experiment, the culture medium was discarded and the cells were incubated for 30 min in the same medium without serum and myo-[SH]inositol.
Chemical Synthesis of VP Analogues Peptides were synthesized by a classicial solid phase method. In brief, the BOC-TFA-HF strategy was employed for synthesis of dLVP, 6, and the linear antagonist molecule 7, HD-Tyr(OEt)-Phe-Val-Asn-Abu-Pro-Arg-Arg-NH2 (14). Purity was assessed by HPLC (Waters M-45, reversed-phase iz-Bondapack column, 10-90% acetonitrile gradient in 0.05% aqueous TFA) and identity was confirmed by fast atom bombardement mass spectroscopy FAB-MS. Fluorophore modification was made by direct acylation of the free amino functions of peptides 6 and 7 with fluoresceine-isothiocyanate (FITC). For the preparation of compound 1, a solution of 10 mg of dLVP in 5 ml of anydrous dimethylformamide (DMF) was acylated with 10 mg (3-fold excess) of FITC (Sigma) and left at room temperature with stirring for 4 hours. Gel filtration of the reaction mixture over Sephadex LH-20 (0.8 X 29 cm) and elution with methanol produced three fractions. The first eluate was further purified on a short silica gel column, eluted with CHC13:MeOH:H20 = 65:25:4 and finally reeluted with methanol from LH-20. The pure product was fluorescent, reacted positively on the chlorinetoluidine test and had an Rf of 0.50 on TLC (Merck silica gel precoated), eluted with n-butanol:acetic acid:water = 5:2:3. Starting material 6 had Rf = 0.38 and FITC Rf = 0.77. After dilution with water and lyophilization, followed by two more lyophilizations from water, peptide 1 as an HPLC-pure product was obtained (2.3 mg). An equivalent amount of peptide 7 was treated in an identical manner with FITC and, after purification, 3.8 mg of 4 were obtained (Rf = 0.65, Rf 7 = 0.52). Peptide 2 was obtained by acylation of 8 mg of 6 under identical conditions with 4 mg of pyrenylbutyric acid (Aldrich), which had been preactivated for 25 min by 2.4 mg ofdicyclohexyl carbodiimide (Sigma) in 1.5 ml anhydrous DMF. After filtration on celite, the mixture was purified by semiprep reversed-phase chromatography (Michel-Miller column 0.8 X 20 cm filled with 30 u nucleosil-Cl 8, Macherey-Nagel, Darmstadt, Germany) and was eluted with an acetonitrile gradient (25-60%) in aqueous 0.1% TFA. Peptide 2 (2.1 mg) was recovered as an HPLC-pure compound after three-fold lyophilization, with an Rf = 0.57 (chlorine-toluidine positive) and fluorescent (Rf pyrenylbutyric acid = 0.90, chlorine-toluidine negative). A solution of 10 mg of 6 in 5 ml of DMF:AcOH (1:1) was reacted with an aqueous solution of glutaraldehyde (50 ~1, 25%) at room temperature for 30 rain. To this reaction mixture were added 50 ul of ethylene° diamine (Aldrich) and it was stirred for an additional 60 min. The reaction mixture was reduced by the addition of 18 mg of sodium cyanoborohydride (17). The crude reaction mixture was filtered over LH-20 and the first peptide fraction was collected (Rf = 0.21) and lyophilized. After dissolving this product in 10
GUILLON ET AL. ml of anhydrous methanol, 10 mg of solid FITC was added and the mixture was stirred for 4 h at room temperature. The reaction mixture was partially evaporated and purified on LH-20-methanol. The first peptide fraction was repeatedly lyophilized and further purified on reversed-phase chromatography as for peptide 2. After lyophilization, 3.2 mg of pure 3 with an Rf = 0.38 (chlorine-toluidine positive and fluorescent) were obtained from the fractions containing the product. Peptide 5, a potential photoaffinity label, was obtained by acylation of 7 with nitrophenylacetic acid and subsequent modification to the azidopeptide 5(1).
Determination of the Binding Properties of the VP Analogues The binding experiments were performed using partially purified preparations of plasma membranes derived from WRK~ cells (8) and from rat liver cells. As described earlier (7), the WRK~ cells exhibit a VP receptor of the V l a subtype. Both preparations represent convenient testing models for new VP analogues since VP receptors are present in large amounts in both of these preparations. Briefly, WRKI cells were washed 5 times with phosphate-buffered saline (PBS) without calcium and magnesium and were resuspended in an homogenization medium containing 10 mM Tris-HC1, pH 8.0, 0.5 mM EDTA, and PMSF 0.1 mM PMSF. The cells were disrupted at 0°C using a Dounce homogenizer equipped with a loose pestle. Partially disrupted cells were eliminated by a preliminary centrifugation (100 X g, 5 min at 0°C). The supernatant was then centrifuged for 30 min at 30,000 X g and 0°C. The pellet was resuspended in the homogenization buffer and used immediately. Rat liver cell membranes were prepared according to Neville (16) and were used as a microsomal fraction (preparation step 11). Binding experiments (WRK~ membranes) were initiated by adding plasma membranes (30 to 50 ug protein in a 10 mM Tris-HCl medium, pH 8.0, containing 1 mM MgC12, 0.1 mM PMSF, 1 mg/ml BSA, 0.5 mg/ml bacitracin, 0.005% w/v soybean trypsin inhibitor and the tritiated [ArgS]VP and the modified VP peptides to be tested. The reaction was allowed to proceed for 35 min at 37°C [condition under which VP binding reaches equilibrium, (9) and unpublished results (G.G.)]. The mixture was then filtered (GF/C for WRK~, Whatmann 934-AH for rat liver) and rinsed 3 times with 3 ml of an ice-cold buffer containing: 10 mM TrisHC1, pH 8.0, and 10 mM MgCI/. The binding experiments with rat liver membranes were carried out in a similar manner using 30 mM Tris-HC1, pH 7.4, with 3 mM MgCI without PMSF and soybean inhibitor, and the incubation was carried out at room temperature. The radioactivity retained on the filter (total binding) was measured by liquid scintillation counting. Nonspecific binding was determined by incubating the membrane preparation in the same incubation medium plus an excess of unlabeled [Arga]VP (1 t~M). Specific binding was calculated as the difference between total and nonspecific binding. The binding dissociation constants for unlabeled peptides (Ki) were deduced by fitting the experimental data with the expected linear relationship: log {[Bo/B - 1](H/I~) + 1]} = log I - log Ki, in which Bo is the specific binding measured in the absence of unlabeled peptide; B is the specific binding measured in the presence of a given concentration of unlabeled peptide I; H is the concentration of [SH][ArgS]VP used in the assay (0.5 nM) and I ~ is its affinity measured on these membranes (Ka = 0.29 + 0.09 nM) [see (6)].
Determination of lnositol Phosphate Accumulation The characterization of the agonistic properties of the VP analogues studied was performed by measuring their ability to
F L U O R E S C E N T PEPTIDE H O R M O N E S
9
TABLE 1 STRUCTURESOF REPORTERGROUP LABELEDPEPTIDES VP 1 2 3 4 5 6 7
H-Cys-Tyr-Phe-Gln-Asn-Cys-Pro-Arg-Gly-NH2 Mpa-Tyr-Phe-Gln-Asn-Cys-Pro-Lys(N-e-fluoresceinylthiocarbamoyl)-Gly-NH2 Mpa-Tyr-Phe-Gln-Asn-Cys-Pro-Lys(N-e-pyrenylbutyryl)-Gly-NH2 Mpa-Tyr-Phe-G~n-Asn-Cys-Pr~-Lys(N-~-~u~resceiny~thi~carbam~y~-amid~ethy~-amin~penty~)-G~y-NHz N-a-fluoresceinyl-thiocarbamoyl-D-Tyr(OEt)-Phe-Val-Asn-Abu-Pro-Arg-Arg-NHz N-a(4-azidophenylacetyl)-D-Tyr(OEt)-Phe-Val-Asn-Abu-Pro-Arg-Arg-NH2 Mpa-Tyr-Phe-Gln-Asn-Cys-Pro-Lys-Glu-NH2 (dLVP) H-D-Tyr(OEt)-Phe-Val-Asn-Abu-Pro-Arg-Arg-NH2
Mpa:/3-mercaptopropionic acid, Abu: aminobutyric acid.
increase the intracellular concentration of inositol phosphate. This represents a convenient technique for such studies since we have previously demonstrated on these cells that the VP effects were triggered by the stimulation of a PIP2 phospholipase C (12). Briefly, myo-[aH]inositol prelabeled cells (0.5 × 10 6 per assay) were preincubated 15 m i n at 37°C in a PBS medium containing 10 m M LiCI and 1 mg/ml BSA [see (7) for composition]. The cells were then further incubated 6 min at 37°C in the same medium with increasing a m o u n t s of the VP agonists to be tested or in the presence of 3 n M of [ArgS]Vp, plus increasing amounts of VP antagonists. The incubation was stopped by aspirating the medium and adding 5.0% (w/v) HCIO4. The cells were scraped off and the resulted extracts neutralized with Hepes-KOH. Accumulated inositol phosphates were separated by batch elution on Dowex columns (1X-10, 200-400 mesh, formate from ) as earlier described (12)). Previous studies performed on these cells demonstrated that VP and agonistic analogues generated the accumulation of the inositol phosphates InsP~, InsP2 and InsP3 (12,15). As the concentrations of VP analogues leading to half-maximal inositol phosphate accumulation were similar, whatever the inositoi phosphate isomer, we have not separated the different inositol phosphates in this study. RESULTS Modifications of dLVP (6) and the linear antagonist 7 were relatively easily obtained, although the yields were rather moderate ( 18% to 27%) due to the small a m o u n t s of starting material used, the often incomplete modification necessary to avoid side
reactions, and the repeatecl purification procedures. The binding experiments on W R K t and liver cell membranes showed poor binding potencies for the simple acylation peptides I, 2, 4 and 5; all were far below the necessary 10% relative affinity compared to VP. On the other hand, c o m p o u n d 3, which displayed free secondary a m i n o functions in the side chain ofpos. 8, had reasonable affinity. The activity profile on the phospholipase C activation resembled very much the binding studies with, again, only peptide 3 having good agonistic properties. As expected, the linear peptides derived from peptide 7 had weak but distinct antagonistic properties. The results of the binding assays and the phospholipase C activation are presented in Table 1. DISCUSSION It has been pointed out (11) that direct acylation of the amino function in the lateral side chain of [Lysa]vp results in a drastic loss of its binding affinity, and therefore it has been the strategy to preserve the positive charge of the Lys or Arg -residue, e.g., by the use ofimidate-esters (9) for the introduction of prosthetic groups into VP. However, the chemically rather labile nature of the resulting imido-amides, together with the reduced biological activity of the modified VP's, left much room for improvement. In our approach we extended the epsilon-amino function through a Schiff-base formation with glutaraldehyde and a second Schiffbase formation with ethylenediamine, and after reduction with hydroborocyanide, the resulting dLVP analogue, containing in position 8 N-epsilon-(amino-ethylaminopentyl)-Lys8, was easily acylated by activated fluoresceine (fluoresceine-isothiocyanate), presumably on the distal primary a m i n o function. This product
TABLE 2 BIOLOGICALPROPERTIESOF VP ANALOGUES
VP I 2 3 4 5 6
PhospholipaseC Activation
WRKt Membranes
Rat Liver Membranes Kd (nM)
RA%
n
Kd (nM)
RA%
n
Km (nM)
0.64 + 0.04 >2000 40.8 _+ 2.21 1.49 + 0.60 ~1000 75.4 _+ 15.0 2.62 + 0.62
100
0196-9781/92 $5.00 + .DO Copyright© 1992 PergamonPressLtd.
Fluorescent Peptide Hormones: Development of High Affinity Vasopressin Analogues GILLES GUILLON,* DAVID BARBEAU,t WITOLD NEUGEBAUER,'~ SYLVAIN GUAY,t LINE BILODEAU,'~ MARIE-NOELLE BALESTRE,* NICOLE GALLO-PAYETJ; A N D E M A N U E L E S C H E R ~ "t
*Centre CNRS-INSERM de Pharmacologie-Endocrinologie, rue de la Cardonille, BP No 5, Montpellier, France 34094 Departement of tPharmacology and of SMedecine, Faculty of Medecine, UniversitO de Sherbrooke, Sherbrooke, Quebec, Canada J1H 5N4 R e c e i v e d 5 F e b r u a r y 1991 GUILLON, G., D. BARBEAU, W. NEUGEBAUER, S. GUAY, L. BILODEAU, M.-N. BALESTRE, N. GALLO-PAYET AND E. ESCHER. Fluorescentpeptidehormones:Developmentof higha2~nityvasopressinanalogues.PEPT1DES 13(1) 7-11, 1992.-Highly potent and specific peptide hormone analogues with fluorescent reporter groups are current research goals. Until now, however, only moderately potent analogues have been described. We report here several types of vasopressin (VP) analogues with different fluorophores attached to the peptide, In a first series, fluorophores were attached to the free ~ amino function of [desamino~-lysineg]VP(dLVP), producing agonistic analogues. In a second series, reporter groups were added to the N-terminal of open-chain antagonist structures. The biological activities of these analogues were assessed by two different sets of experiments: 1) The measurement of their binding affinities towards the Vla-vasopressin receptor subtype from WRKI cells or rat liver membrane preparations; 2) Their ability to stimulate the phospholipase C activity in WRK~ cells. As expected, a simple acylation of fluorophores to dLVP resulted in a considerable loss of affinity. If, however, the Lys~ side chain was extended through double Schiff-base formation with glutaraldehyde-ethylenediamine followed by reduction to an aminoalkyl aminoalkylamine, single fluorophores could be added without loss of affinity compared to VP. The open-chain analogues, on the other hand, while displaying weak affnity, nevertheless exhibited pure antagonistic behavior. Vasopressin
Vasopressin analogues
Fluorescent reporter groups
PEPTIDE hormones with reporter functions have been a key to success in peptide hormone endocrinology and pharmacology. Several hormone-receptor systems are now well characterized biochemically, due to the availability of such functionalized ligands. Although vasopressin (VP) is one of the longest known peptide transmitters, it has not yet attained this level of knowledge, although a great deal of knowledge concerning its structureactivity relationship is currently available. Research on the VP receptor has always been hampered due to the lack of receptor ligand with reporter functions. Among the plenitude of hormone analogues, only tritiated VP is available as a radioactive peptide with full agonistic properties, and, until now, no fluorescent analogue of VP has been reported with more than 5-10% of the affinity of the native hormone despite the considerable efforts in this direction which have recently been made (2,3,11,13). Highly fluorescent analogues with affinity and selectivity similar to natural VP would greatly facilitate the search for the VP receptors. Binding studies could then be carried out without the presence of radioactivity, and histological studies could be made
without the necessity for film exposure; colocalization and internalization of hormonal receptors would also become feasible. Furthermore, such fluorescent analogues could greatly facilitate receptor cloning studies, and VP receptor-expressing cells could be selected from nonexpressing cells by cell sorter techniques in the presence of such analogues. As a starting point for the synthesis of the agonistic structure, we chose [des-aminol-lysineS]Vp (dLVP). For the antagonistic structures, the recently reported noncyclic structures seemed to be almost ideally suited due to the possibility of N-terminal acylation with hydrophobic groups such as adamantylacetate (14). As fluorophores, we tested the classical fluoresceine and the pyrene residues. METHOD The chemicals used in the present study were obtained from the following sources: myo-[3H]inositol (10-20 Ci/mmol) from Commisariat fi l'6nergie atomique, France; [3H][ArgS]VP (70 Ci/mmol) from New England Nuclear; minimal essential me-
1Requests for reprints should be addressed to Emanuel Escher, Ph.D., Department of Pharmacology, Faculty of Medicine, Universit6 de Sherbrooke, Sherbrooke, Qu6bec, Canada J 1H 5N4.
8 dium (MEM) and Earle's salts from Gibco. Other chemicals were of A-grade purity.
WRKI Cell Culture WRK~ cells were established and cultured, with (inositol phosphates experiment) or without myo-[3H]inositol (binding experiment) as described previously (7,12). Cells were plated in 3.5 cm plastic dishes, at a density of(5-7) X 104 cells in minimal essential medium containing Earle's salts, fetal calf serum (5%, v/v), rat serum (2%, v/v), glutamine (2 mM), penicillin (100 units/ml), streptomycin (100 m g / m l ) a n d myo-[3H]inositol (2 uCi/ml). The medium was changed after 2 days and experiments were performed 3 days later. Before the experiment, the culture medium was discarded and the cells were incubated for 30 min in the same medium without serum and myo-[SH]inositol.
Chemical Synthesis of VP Analogues Peptides were synthesized by a classicial solid phase method. In brief, the BOC-TFA-HF strategy was employed for synthesis of dLVP, 6, and the linear antagonist molecule 7, HD-Tyr(OEt)-Phe-Val-Asn-Abu-Pro-Arg-Arg-NH2 (14). Purity was assessed by HPLC (Waters M-45, reversed-phase iz-Bondapack column, 10-90% acetonitrile gradient in 0.05% aqueous TFA) and identity was confirmed by fast atom bombardement mass spectroscopy FAB-MS. Fluorophore modification was made by direct acylation of the free amino functions of peptides 6 and 7 with fluoresceine-isothiocyanate (FITC). For the preparation of compound 1, a solution of 10 mg of dLVP in 5 ml of anydrous dimethylformamide (DMF) was acylated with 10 mg (3-fold excess) of FITC (Sigma) and left at room temperature with stirring for 4 hours. Gel filtration of the reaction mixture over Sephadex LH-20 (0.8 X 29 cm) and elution with methanol produced three fractions. The first eluate was further purified on a short silica gel column, eluted with CHC13:MeOH:H20 = 65:25:4 and finally reeluted with methanol from LH-20. The pure product was fluorescent, reacted positively on the chlorinetoluidine test and had an Rf of 0.50 on TLC (Merck silica gel precoated), eluted with n-butanol:acetic acid:water = 5:2:3. Starting material 6 had Rf = 0.38 and FITC Rf = 0.77. After dilution with water and lyophilization, followed by two more lyophilizations from water, peptide 1 as an HPLC-pure product was obtained (2.3 mg). An equivalent amount of peptide 7 was treated in an identical manner with FITC and, after purification, 3.8 mg of 4 were obtained (Rf = 0.65, Rf 7 = 0.52). Peptide 2 was obtained by acylation of 8 mg of 6 under identical conditions with 4 mg of pyrenylbutyric acid (Aldrich), which had been preactivated for 25 min by 2.4 mg ofdicyclohexyl carbodiimide (Sigma) in 1.5 ml anhydrous DMF. After filtration on celite, the mixture was purified by semiprep reversed-phase chromatography (Michel-Miller column 0.8 X 20 cm filled with 30 u nucleosil-Cl 8, Macherey-Nagel, Darmstadt, Germany) and was eluted with an acetonitrile gradient (25-60%) in aqueous 0.1% TFA. Peptide 2 (2.1 mg) was recovered as an HPLC-pure compound after three-fold lyophilization, with an Rf = 0.57 (chlorine-toluidine positive) and fluorescent (Rf pyrenylbutyric acid = 0.90, chlorine-toluidine negative). A solution of 10 mg of 6 in 5 ml of DMF:AcOH (1:1) was reacted with an aqueous solution of glutaraldehyde (50 ~1, 25%) at room temperature for 30 rain. To this reaction mixture were added 50 ul of ethylene° diamine (Aldrich) and it was stirred for an additional 60 min. The reaction mixture was reduced by the addition of 18 mg of sodium cyanoborohydride (17). The crude reaction mixture was filtered over LH-20 and the first peptide fraction was collected (Rf = 0.21) and lyophilized. After dissolving this product in 10
GUILLON ET AL. ml of anhydrous methanol, 10 mg of solid FITC was added and the mixture was stirred for 4 h at room temperature. The reaction mixture was partially evaporated and purified on LH-20-methanol. The first peptide fraction was repeatedly lyophilized and further purified on reversed-phase chromatography as for peptide 2. After lyophilization, 3.2 mg of pure 3 with an Rf = 0.38 (chlorine-toluidine positive and fluorescent) were obtained from the fractions containing the product. Peptide 5, a potential photoaffinity label, was obtained by acylation of 7 with nitrophenylacetic acid and subsequent modification to the azidopeptide 5(1).
Determination of the Binding Properties of the VP Analogues The binding experiments were performed using partially purified preparations of plasma membranes derived from WRK~ cells (8) and from rat liver cells. As described earlier (7), the WRK~ cells exhibit a VP receptor of the V l a subtype. Both preparations represent convenient testing models for new VP analogues since VP receptors are present in large amounts in both of these preparations. Briefly, WRKI cells were washed 5 times with phosphate-buffered saline (PBS) without calcium and magnesium and were resuspended in an homogenization medium containing 10 mM Tris-HC1, pH 8.0, 0.5 mM EDTA, and PMSF 0.1 mM PMSF. The cells were disrupted at 0°C using a Dounce homogenizer equipped with a loose pestle. Partially disrupted cells were eliminated by a preliminary centrifugation (100 X g, 5 min at 0°C). The supernatant was then centrifuged for 30 min at 30,000 X g and 0°C. The pellet was resuspended in the homogenization buffer and used immediately. Rat liver cell membranes were prepared according to Neville (16) and were used as a microsomal fraction (preparation step 11). Binding experiments (WRK~ membranes) were initiated by adding plasma membranes (30 to 50 ug protein in a 10 mM Tris-HCl medium, pH 8.0, containing 1 mM MgC12, 0.1 mM PMSF, 1 mg/ml BSA, 0.5 mg/ml bacitracin, 0.005% w/v soybean trypsin inhibitor and the tritiated [ArgS]VP and the modified VP peptides to be tested. The reaction was allowed to proceed for 35 min at 37°C [condition under which VP binding reaches equilibrium, (9) and unpublished results (G.G.)]. The mixture was then filtered (GF/C for WRK~, Whatmann 934-AH for rat liver) and rinsed 3 times with 3 ml of an ice-cold buffer containing: 10 mM TrisHC1, pH 8.0, and 10 mM MgCI/. The binding experiments with rat liver membranes were carried out in a similar manner using 30 mM Tris-HC1, pH 7.4, with 3 mM MgCI without PMSF and soybean inhibitor, and the incubation was carried out at room temperature. The radioactivity retained on the filter (total binding) was measured by liquid scintillation counting. Nonspecific binding was determined by incubating the membrane preparation in the same incubation medium plus an excess of unlabeled [Arga]VP (1 t~M). Specific binding was calculated as the difference between total and nonspecific binding. The binding dissociation constants for unlabeled peptides (Ki) were deduced by fitting the experimental data with the expected linear relationship: log {[Bo/B - 1](H/I~) + 1]} = log I - log Ki, in which Bo is the specific binding measured in the absence of unlabeled peptide; B is the specific binding measured in the presence of a given concentration of unlabeled peptide I; H is the concentration of [SH][ArgS]VP used in the assay (0.5 nM) and I ~ is its affinity measured on these membranes (Ka = 0.29 + 0.09 nM) [see (6)].
Determination of lnositol Phosphate Accumulation The characterization of the agonistic properties of the VP analogues studied was performed by measuring their ability to
F L U O R E S C E N T PEPTIDE H O R M O N E S
9
TABLE 1 STRUCTURESOF REPORTERGROUP LABELEDPEPTIDES VP 1 2 3 4 5 6 7
H-Cys-Tyr-Phe-Gln-Asn-Cys-Pro-Arg-Gly-NH2 Mpa-Tyr-Phe-Gln-Asn-Cys-Pro-Lys(N-e-fluoresceinylthiocarbamoyl)-Gly-NH2 Mpa-Tyr-Phe-Gln-Asn-Cys-Pro-Lys(N-e-pyrenylbutyryl)-Gly-NH2 Mpa-Tyr-Phe-G~n-Asn-Cys-Pr~-Lys(N-~-~u~resceiny~thi~carbam~y~-amid~ethy~-amin~penty~)-G~y-NHz N-a-fluoresceinyl-thiocarbamoyl-D-Tyr(OEt)-Phe-Val-Asn-Abu-Pro-Arg-Arg-NHz N-a(4-azidophenylacetyl)-D-Tyr(OEt)-Phe-Val-Asn-Abu-Pro-Arg-Arg-NH2 Mpa-Tyr-Phe-Gln-Asn-Cys-Pro-Lys-Glu-NH2 (dLVP) H-D-Tyr(OEt)-Phe-Val-Asn-Abu-Pro-Arg-Arg-NH2
Mpa:/3-mercaptopropionic acid, Abu: aminobutyric acid.
increase the intracellular concentration of inositol phosphate. This represents a convenient technique for such studies since we have previously demonstrated on these cells that the VP effects were triggered by the stimulation of a PIP2 phospholipase C (12). Briefly, myo-[aH]inositol prelabeled cells (0.5 × 10 6 per assay) were preincubated 15 m i n at 37°C in a PBS medium containing 10 m M LiCI and 1 mg/ml BSA [see (7) for composition]. The cells were then further incubated 6 min at 37°C in the same medium with increasing a m o u n t s of the VP agonists to be tested or in the presence of 3 n M of [ArgS]Vp, plus increasing amounts of VP antagonists. The incubation was stopped by aspirating the medium and adding 5.0% (w/v) HCIO4. The cells were scraped off and the resulted extracts neutralized with Hepes-KOH. Accumulated inositol phosphates were separated by batch elution on Dowex columns (1X-10, 200-400 mesh, formate from ) as earlier described (12)). Previous studies performed on these cells demonstrated that VP and agonistic analogues generated the accumulation of the inositol phosphates InsP~, InsP2 and InsP3 (12,15). As the concentrations of VP analogues leading to half-maximal inositol phosphate accumulation were similar, whatever the inositoi phosphate isomer, we have not separated the different inositol phosphates in this study. RESULTS Modifications of dLVP (6) and the linear antagonist 7 were relatively easily obtained, although the yields were rather moderate ( 18% to 27%) due to the small a m o u n t s of starting material used, the often incomplete modification necessary to avoid side
reactions, and the repeatecl purification procedures. The binding experiments on W R K t and liver cell membranes showed poor binding potencies for the simple acylation peptides I, 2, 4 and 5; all were far below the necessary 10% relative affinity compared to VP. On the other hand, c o m p o u n d 3, which displayed free secondary a m i n o functions in the side chain ofpos. 8, had reasonable affinity. The activity profile on the phospholipase C activation resembled very much the binding studies with, again, only peptide 3 having good agonistic properties. As expected, the linear peptides derived from peptide 7 had weak but distinct antagonistic properties. The results of the binding assays and the phospholipase C activation are presented in Table 1. DISCUSSION It has been pointed out (11) that direct acylation of the amino function in the lateral side chain of [Lysa]vp results in a drastic loss of its binding affinity, and therefore it has been the strategy to preserve the positive charge of the Lys or Arg -residue, e.g., by the use ofimidate-esters (9) for the introduction of prosthetic groups into VP. However, the chemically rather labile nature of the resulting imido-amides, together with the reduced biological activity of the modified VP's, left much room for improvement. In our approach we extended the epsilon-amino function through a Schiff-base formation with glutaraldehyde and a second Schiffbase formation with ethylenediamine, and after reduction with hydroborocyanide, the resulting dLVP analogue, containing in position 8 N-epsilon-(amino-ethylaminopentyl)-Lys8, was easily acylated by activated fluoresceine (fluoresceine-isothiocyanate), presumably on the distal primary a m i n o function. This product
TABLE 2 BIOLOGICALPROPERTIESOF VP ANALOGUES
VP I 2 3 4 5 6
PhospholipaseC Activation
WRKt Membranes
Rat Liver Membranes Kd (nM)
RA%
n
Kd (nM)
RA%
n
Km (nM)
0.64 + 0.04 >2000 40.8 _+ 2.21 1.49 + 0.60 ~1000 75.4 _+ 15.0 2.62 + 0.62
100
