a new probe for affinity labelling pancreatic cholecystokinin receptor with minor modification of...

7
Eur. J. Biochem. 185, 397-403 (1989) 0 FEBS 1989 A new probe for affinity labelling pancreatic cholecystokinin receptor with minor modification of its structure Daniel FOURMY Stephen P. POWERS3 and Nicole VAYSSE' ' Groupe de Recherche de Biologie et Pathologie Digestine, INSERM U 151, CHU Rangueil, Toulouse Patrick LOPEZ', Sandrine POTROT ', Juan JIMENEZ', Marline DUFRESNE', Luis MORODER'. Max-Planck-Institut fur Biochemie, Munchen Mayo Clinic, Rochester (Received April 17/July 11, 1989) - EJB 89 0479 Biochemical studies on receptors for peptides are most often carried out on affinity-labelled (peptide-receptor) complexes. Necessarily, the assumption is made that a covalent (peptide-receptor) complex behaves as the native receptor. The validity of this assumption is dependent on both the affinity-labelling technique and the resolution of the analytical method used for biochemical characterization. We designed a new affinity-labelling probe in order to minimize structural modifications occurring within the affinity-labelled cholecystokinin (CCK) receptor protein. The probe was '251-labelled 2-(p-azidosalicylamido)-l,3-dithiopropionate-[Thr28,Ah~31]CCK-25-33, ('2sI-ASD-[Thr28,Ahx3']CCK-25-33), the peptide moiety of which was released from its binding site by reduction. It was obtained by coupling a photoactivable chemical to [Thr28,Ahx31]CCK-25-33 via its N-terminus. The resulting peptide was HPLC purified and radioiodinated in the presence of chloramine T. Binding of '251-ASD- [Thr", Ahx31]CCK-25-33was time- and temperature-dependent and reversible. At 25 "C, a steady-state level was reached after 60 min and half-maximal dissociation after 38 min. Binding was inhibited by [Thrz8,Ahx3']CCK- 25-33 and L-364-718 antagonist with IC500.4 nM and 0.9 nM, respectively. Photoaffinity labelling of pancreatic plasma membranes by '251-ASD-[Thr28,Ahx3']CCK-25-33 identified a glycoprotein of M, 85000 - 100000 which was retained on immobilized wheat germ agglutinin. Enzyme cleavage by endoproteinase Clu-C generated a main fragment of M, 30000 - 34000. The same glycoprotein was photoaffinity labelled with '251-~Tyr-Cly-[Ahx28z3 ', ~NO~Phe~~lCCK-26-33 (Ahx, 2-aminohexanoic acid; pN02Phe, p-nitrophenylalanine) an intrinsic probe having its photolabile group sited in the binding domain of cholecystokinin. '251-ASD-[Thr2s,Ahx3']CCK-25-33 is a potentially powerful tool for biologically and biochemically studying cholecystokinin receptors. Cholecystokinin is a major regulatory peptide, initially isolated as a 33-amino-acid form [l]. Cholecystokinin peptides induce multiple biological effects within the gut such as stimu- lation of gall bladder contraction and pancreatic enzyme se- cretion [2]. Receptors for cholecystokinin are pharmacologically characterized as several major subtypes, one of them, an A subtype being the most abundant on pancreatic acinar cells from rodents [3 - 71. Pancreatic A-subtype cholecystokinin receptor is a highly glycosylated membrane protein migrating to approximately M, 90000 on SDSjPACE [8]. We are attempting to purify a fragment of the chole- cystokinin receptor including the binding domain of cholecystokinin. Our strategy involves the use of complemen- tary techniques such as affinity chromatography on immobilized lectins, isoelectrofocusing, SDSjPACE and HPLC. To trace the purification, a fraction of membrane proteins is photoaffinity labelled with a cholecystokinin pro- be. However, photoaffinity labelling with available cholecystokinin probes resulted in an irreversible attachment Correspondence to D. Fourmy, INSERM U 151, Bat. L3, 1, av- enue Jean Poulhis, CHU Rangueil, F-31054 Toulouse Cedex, France Abbreviations. CCK, cholecystokinin; SASD, sulfosuccinimidyl 2-(p-azidosalicylamido)-1,3'-dithioproprionate; ASD, 2-(p-azido- salicylamido)-l,3'-dithiopropionate; TC5,,, median inhibitory concen- tration; Ahx, aminohexanoic acid. of radioiodinated cholecystokinin to the receptor molecule. This led to a modification of receptor characteristics such as mass, charge and hydrophobicity. Since our strategy for cholecystokinin receptor purifi- cation implies copurification of the labelled and the unlabelled receptor proteins, we were led to design a cholecystokinin probe which could covalently radiolabel the cholecystokinin receptor with minimal modification of its structure. This pro- be was a cleavable cholecystokinin ligand ('251-ASD-[Thr28, Ahx3']CCK-25-33), the peptide moiety of which was released from the photoaffinity-labelled receptor by reduction. This new probe bound with a high affinity to the pancreatic cholecystokinin receptor and photoaffinity labelled the same protein as ~Tyr-Gly-[Ahx~**~ ', pN02Phe33]CCK-26-33], an intrinsic cholecystokinin ligand having its photoactivable moiety in the binding domain of cholecystokinin [8]. EXPERIMENTAL PROCEDURES Muteriuls The C-terminal nonapeptide of cholecystokinin, in which the methionines have been substituted by threonine and 2- aminohexanoic acid, residues was a gift of Dr L. Moroder (Max-Planck-Institut fur Biochemie, Munchen, FRG). This peptide is abbreviated to [Thr28,Ahx31]CCK-25-33. The in-

Upload: daniel-fourmy

Post on 02-Oct-2016

213 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: A new probe for affinity labelling pancreatic cholecystokinin receptor with minor modification of its structure

Eur. J. Biochem. 185, 397-403 (1989) 0 FEBS 1989

A new probe for affinity labelling pancreatic cholecystokinin receptor with minor modification of its structure Daniel FOURMY Stephen P. POWERS3 and Nicole VAYSSE' ' Groupe de Recherche de Biologie et Pathologie Digestine, INSERM U 151, CHU Rangueil, Toulouse

Patrick LOPEZ', Sandrine POTROT ', Juan JIMENEZ', Marline DUFRESNE', Luis MORODER'.

Max-Planck-Institut fur Biochemie, Munchen Mayo Clinic, Rochester

(Received April 17/July 11, 1989) - EJB 89 0479

Biochemical studies on receptors for peptides are most often carried out on affinity-labelled (peptide-receptor) complexes. Necessarily, the assumption is made that a covalent (peptide-receptor) complex behaves as the native receptor. The validity of this assumption is dependent on both the affinity-labelling technique and the resolution of the analytical method used for biochemical characterization. We designed a new affinity-labelling probe in order to minimize structural modifications occurring within the affinity-labelled cholecystokinin (CCK) receptor protein. The probe was '251-labelled 2-(p-azidosalicylamido)-l,3-dithiopropionate-[Thr28,Ah~31]CCK-25-33, ('2sI-ASD-[Thr28, Ahx3']CCK-25-33), the peptide moiety of which was released from its binding site by reduction. It was obtained by coupling a photoactivable chemical to [Thr28,Ahx31]CCK-25-33 via its N-terminus. The resulting peptide was HPLC purified and radioiodinated in the presence of chloramine T. Binding of '251-ASD- [Thr", Ahx31]CCK-25-33 was time- and temperature-dependent and reversible. At 25 "C, a steady-state level was reached after 60 min and half-maximal dissociation after 38 min. Binding was inhibited by [Thrz8,Ahx3']CCK- 25-33 and L-364-718 antagonist with IC50 0.4 nM and 0.9 nM, respectively. Photoaffinity labelling of pancreatic plasma membranes by '251-ASD-[Thr28,Ahx3']CCK-25-33 identified a glycoprotein of M, 85000 - 100000 which was retained on immobilized wheat germ agglutinin. Enzyme cleavage by endoproteinase Clu-C generated a main fragment of M , 30000 - 34000. The same glycoprotein was photoaffinity labelled with '251-~Tyr-Cly-[Ahx28z3 ', ~ N O ~ P h e ~ ~ l C C K - 2 6 - 3 3 (Ahx, 2-aminohexanoic acid; pN02Phe, p-nitrophenylalanine) an intrinsic probe having its photolabile group sited in the binding domain of cholecystokinin. '251-ASD-[Thr2s,Ahx3']CCK-25-33 is a potentially powerful tool for biologically and biochemically studying cholecystokinin receptors.

Cholecystokinin is a major regulatory peptide, initially isolated as a 33-amino-acid form [l]. Cholecystokinin peptides induce multiple biological effects within the gut such as stimu- lation of gall bladder contraction and pancreatic enzyme se- cretion [2].

Receptors for cholecystokinin are pharmacologically characterized as several major subtypes, one of them, an A subtype being the most abundant on pancreatic acinar cells from rodents [3 - 71. Pancreatic A-subtype cholecystokinin receptor is a highly glycosylated membrane protein migrating to approximately M , 90000 on SDSjPACE [8].

We are attempting to purify a fragment of the chole- cystokinin receptor including the binding domain of cholecystokinin. Our strategy involves the use of complemen- tary techniques such as affinity chromatography on immobilized lectins, isoelectrofocusing, SDSjPACE and HPLC. To trace the purification, a fraction of membrane proteins is photoaffinity labelled with a cholecystokinin pro- be. However, photoaffinity labelling with available cholecystokinin probes resulted in an irreversible attachment

Correspondence to D. Fourmy, INSERM U 151, Bat. L3, 1, av- enue Jean Poulhis, CHU Rangueil, F-31054 Toulouse Cedex, France

Abbreviations. CCK, cholecystokinin; SASD, sulfosuccinimidyl 2-(p-azidosalicylamido)-1,3'-dithioproprionate; ASD, 2-(p-azido- salicylamido)-l,3'-dithiopropionate; TC5,,, median inhibitory concen- tration; Ahx, aminohexanoic acid.

of radioiodinated cholecystokinin to the receptor molecule. This led to a modification of receptor characteristics such as mass, charge and hydrophobicity.

Since our strategy for cholecystokinin receptor purifi- cation implies copurification of the labelled and the unlabelled receptor proteins, we were led to design a cholecystokinin probe which could covalently radiolabel the cholecystokinin receptor with minimal modification of its structure. This pro- be was a cleavable cholecystokinin ligand ('251-ASD-[Thr28, Ahx3']CCK-25-33), the peptide moiety of which was released from the photoaffinity-labelled receptor by reduction. This new probe bound with a high affinity to the pancreatic cholecystokinin receptor and photoaffinity labelled the same protein as ~Tyr-Gly-[Ahx~**~ ', pN02Phe33]CCK-26-33], an intrinsic cholecystokinin ligand having its photoactivable moiety in the binding domain of cholecystokinin [8].

EXPERIMENTAL PROCEDURES

Muteriuls

The C-terminal nonapeptide of cholecystokinin, in which the methionines have been substituted by threonine and 2- aminohexanoic acid, residues was a gift of Dr L. Moroder (Max-Planck-Institut fur Biochemie, Munchen, FRG). This peptide is abbreviated to [Thr28,Ahx31]CCK-25-33. The in-

Page 2: A new probe for affinity labelling pancreatic cholecystokinin receptor with minor modification of its structure

398

trinsic photoactivable cholecystokinin peptide DTyr-Gly- [A~x,* ,~’ , pN02Phe33]CCK-25-33 was synthesized by Dr S.P. Powers and L.J. Miller, Mayo Clinic, Rochester, MN (USA). This peptide is composed of the C-terminal octa- peptide of cholecystokinin in which the methionines have been replaced by Ahx residues and a nitro group incorporated at the para position on the phenylalanine. Moreover, this peptide was extended at the N-terminus by glycine and D-tyrOSine residues. Radioiodination of the peptide occurred on the D- tyrosine [8]. L-364-718, a specific antagonist of A-subtype cholecystokinin receptor was given by Merck Sharp and Dohme (Rahway, USA). Sulfosuccinimidyl 2-k-azido- salicylamido)-l,3’-dithiopropionate (SASD) and activated agarose were purchased from Pierce, Holland. Other com- pounds were from the following sources: Na12’I from Amer- sham International (Les Ulis, France); wheat germ agglutinin from IBF (Villeneuve La Garenne, France); Nonidet P40 from Sigma Chemical Compagny (St Louis, MO, USA); di- thiothreitol and chemicals used in SDSjPAGE from Bio-Rad (Richmond, USA) ; endoproteinase Glu-C from Boehringer Mannheim Biochemicals (Meylan, France). All HPLC equip- ment was from Millipore, SA (USA).

Synthesis and purification of ASD-( Thr2’, Ahx31/CCK-25-33 SASD (4.3 mg) was dissolved in 500 pl dimethyl form-

amide and reacted with [ThrZ8,Ahx3’]CCK-25-33 (0.5 mg) in 500 p1 50 mM sodium tetraborate, pH 8.5. The reaction was carried out under agitation, in the dark at 2 0 T , for 15 h. The medium was chromatographied on a C18 p-Bondapak HPLC column with 0.125 M triethylammonium phosphate, pH 3.5, plus acetonitrile as eluant. The analytical run was monitored by recording absorbance at 280 nm and the fractions corre- sponding to the absorbance peaks were tested for their ability to displace [Thr”, Ahx3’]CCK-25-33, covalently linked to ‘251-labelled Bolton-Hunter reagent, from pancreatic plasma membranes, as previously described [7].

In the preparative run, the ultraviolet detector was switched off while ASD-[Thr28,Ahx31]CCK-25-33 was being eluted. Finally, A S D - [ T ~ ~ ” , A ~ X ~ ~ ] C C K - ~ ~ - ~ ~ was re- chromatographed on the same column with 1% formic acid plus acetonitrile as eluant and lyophilized for storage.

Radioiodination and purification o f 1 2’I-ASD-(Thr28, Ahx3‘]CCK-25-33

The radioiodination was performed under non-actinic light at 22°C in 0.2 M sodium phosphate buffer, pH 7.5. To 10 pg ASD-[Thr2*, Ahx3’]CCK-25-33, dissolved in 50 p1 buffer, was added 10 p1 of a NalZ5I solution (1 mCi) followed by four additions of 5 p1 of a chloramine-T solution (0.5 pl/ pl) at 10-s intervals. The reaction was quenched by addition of an excess of a NaI/tyrosine mixture. Radioiodinated ASD- [Thr28,Ahx31]CCK-25-33 was purified on a C18 column with 0.125 M triethylammonium phosphate, pH 3.5, containing 37.5% acetonitrile as mobile phase. Elution of radioactivity was recorded by using a radioisotope detector (model 170, Beckman) and fractions were tested for cholecystokinin bind- ing on pancreatic membranes. The same procedure was used to radioiodinate ~Tyr-Gly-[Thr’’,~ ’, ~ N O ~ P h e ~ ~ l C C K - 2 6 - 3 3 yielding a probe with the same properties as in [8].

BINDING STUDIES Enriched pancreatic plasma membranes were prepared

from male rats (1 50 - 200 g) as described essentially in [8]. All

binding studies were performed under non-actinic light. In competition assays, membranes ( 5 - 10 pg protein) and 1251- ASD-[ThrZ8,Ahx3’]CCK-25-33 (100 pM) were incubated at 25°C in 0.5 ml 50 mM Hepes buffer, pH 7.0, containing 11 5 mM NaCl, 5 mM MgCl,, 0.01 % soybean trypsin inhibi- tor, 0.1% bacitracin, 1 mM EGTA, 0.1 mM phenylmethyl- sulfonyl fluoride and 0.2% bovine serum albumin. The incu- bation was stopped by adding 0.5ml iced buffer to the microfuge tubes ; afterwards, bound and free ligand were sep- arated by centrifugation for 3 min at 12000 x g. The radioac- tivity associated with membranes was determined, the specific binding being defined as the total binding minus that found in assays containing 1 pM [Thr28, Ahx31]CCK-25-33.

In association and dissociation experiments, 8-ml incu- bations were used. For association, aliquots were removed in duplicate at specified times, diluted with 0.5 ml iced buffer and centrifuged to determine bound radioligand. In dissocation assays, initial binding was performed for 60 min at 25 ‘C then 1 pM [Thr28,Ah~31]CCK-25-33 was added. At specified times, 0.5-ml aliquots were removed, diluted with 0.5 ml cold buffer and centrifuged to determine the remaining bound radioligand. Non-specific binding was determined in parallel incubations in presence of 3 pM [Thr”, Ahx3’]CCK-25-33.

Photoaffinity labelling

For affinity labelling, binding on membranes was per- formed as described above. Pellets of labelled membranes were then resuspended in iced buffer without bovine serum albumin and transferred into Pyrex tubes. Photolysis was performed at 4°C by using different ultraviolet sources: (a) a 125-W mercury lamp; (b) two 20-W mercury lamps; (c) sun- light. After photolysis, membranes were collected by cen- trifugation for subsequent analysis.

Yield of covalent attachment of 12’I-ASD-[Thr28, Ahx3’]CCK-25-33 was estimated by counting the radioac- tivity which migrated in the specific band at M , 85000- 100000 on SDS/PAGE. It was expressed as a percentage of specific cholecystokinin binding on membranes before pho- tolysis.

Wheat-germ-agglut inin - agarose chromatography

Affinity-labelled membranes were solubilized in binding buffer (0.5 mg proteins/ml) containing 5% Nonidet P-40, by agitation at 4°C for 1 h. The soluble fraction obtained after centrifugation at 100000 x g for 30 min was diluted 10-fold and incubated with wheat germ agglutinin immobilized on agarose (10 ml supernatant/0.7 ml gel) for 15 h at 4°C. The gel was decanted (500 x g , 3 min) and packed into a glass column. After sequential washings with solutions of 0.1 M NaCl, 0.5 M NaCl in 25 ml Hepes, pH 7.5, containing 0.5% Nonidet P40, adsorbed proteins were eluted with 0.5 M N - acetyl-~-D-glucosamine in the same buffer. Fractions were collected and those corresponding to the radioactive peak were pooled and concentrated by ultrafiltration (Centricon 10, Amicon).

Enzyme treatment of aj$nity-labelled cholecystokinin receptor

Concentrated fractions eluted from immobilized wheat germ agglutinin column were precipitated by the method de- scribed in [9] with chloroform/methanol in order to eliminate excess of both detergents and free 1251-ASD-[Thr28, Ahx3’]CCK-25-33. Precipitated proteins (25 pg) were re-

Page 3: A new probe for affinity labelling pancreatic cholecystokinin receptor with minor modification of its structure

399

0 0: m N

Q

0

0.05

0 N

Q

k 0

0.05 0

4

0

A. v

0 20 LO 60 5-1 5 10 15 0 A Time (min) B Time (min) C

5 10 Time ( m i d

Fig. 1. Reverse-phase HPLC of ASD-[ Thr", Ahx3' JCCK-25-33. (a) Analytical chromatography of coupling medium. Peptide (0.5 mg) was coupled to SASD (4.3 mg) as described in Experimental Procedures. After 15 h, an aliquot of medium was injected on a p-Bondapak C t 8 column (0.39 cm x 30 cm, waters). The mobile phase consisted of 0.125 M triethylammonium phosphate, pH 3.5, and acetonitrile, using a gradient over the range 25 - 50%. Fractions corresponding to ultraviolet peaks were tested for their ability to inhibit cholecystokinin binding to pancreatic membranes. The ultraviolet peak at 53 rnin possessed cholecystokinin-binding activity. (b) Analysis of an aliquot (2 pg) of ASD- [Thr", Ahx3']CCK-25-33. After purification as described in (a) and chromatography with water/formic acid (99 : 1) and acetonitrile as eluant, homogeneity of ASD-[Thr28,Ahx31]CCK-25-33 was controlled by HPLC analysis using 0.125 M triethylammonium phosphate, pH 3.5 and acetonitrile (35 -60% gradient in 3.5 min). Detection was performed by recording absorbance at 210 nm. (c) Analysis of dithiothreitol-reduced ASD-[ThrZ8,Ahx3']CCK-25-33. Peptide (10 nmol) was reduced by dithiothreitol (1 pmol) in 100 pl 0.2 M phosphate buffer, pH 7.5. After reaction for 1 h at 25"C, an aliquot was chromatographed on a p-Bondapak CIS column as in (b). The arrow shows the initial retention time of ASD-[Thr28,Ahx31]CCK-25-33

solubilized in 20 p1 0.1 M sodium phosphate buffer, pH 6.1, containing 50 mM EDTA, 1% Nonidet P40, 0.1% SDS and 1% 2-mercaptoethanol and incubated at 37°C for 3 h with 2.5 pg endoproteinase Glu-C. After digestion, the samples were subjected lo SDSjPAGE according to Laemmli [lo] and as previously described in detail [7].

Other methods and chemicals SDSjPAGE and autoradiography of gels were as in 181.

Apparent relative molecular masses of labelled components were calculated using the standard prestained proteins (BRL, Bethesda, USA): myosin, 200000; phosphorylase b, 97 000; bovine serum albumin, 68000; ovalbumin, 43 000; carbonic anhydrase, 29000; j-lactoglobulin, 18 400; lysozyme, 14300.

All other chemicals were of the highest analytical grade available.

RESULTS

Synthesis andpurijication of ASD-[ Thr28, Ahx3 'ICCK-25-33

After 15 h of reaction between SASD and [Thr28, Ahx3 'ICCK-25-33, HPLC analysis of the coupling medium revealed that 75% of peptide was coupled as calculated from the intensity of the peptide peak eluted at 12.5 min (Fig. 1 a). Several peaks were observed at 31 min, 48 rnin and 53 min. All fractions corresponding to these peaks were tested for their ability to inhibit binding of [Thr28, Ahx3']CCK-25-33, covalently linked to '251-labelled Bolton-Hunter reagent, to pancreatic membranes. Among these, only the fraction eluted at 53 min inhibited cholecystokinin binding in a concen-

tration-dependent manner (not illustrated). This fraction was saved in a preparative HPLC run and desalted on a cl8

column with formic acid/acetonitrile as eluant. Presence of ASD at the amino-terminus of [Thr28,Ahx31]CCK-25-33 and its cleavability were established by reducing an aliquot of the new peptide by dithiothreitol and analyzing the product by HPLC. As illustrated in Fig. I b, the newly prepared peptide ASD-[Thr2*,Ahx3']CCK-25-33 was eluted as a symmetrical peak at 11 min, whereas reduced products were eluted as more polar components, at 3.4 rnin and 3.8 rnin (Fig. 1 c).

Radioiodination and purifi:cation of ' 251-ASD-[ Thr28, Ahx3' JCCK-25-33

HPLC separation of an iodination medium is illustrated on Fig. 2. At the void volume, free '''1 and '"I-tyrosine were eluted, followed much later by two close radioactive peaks (retention times, 29 min and 32.5 min). Assays for binding to plasma membranes indicated that each fraction contained molecules possessing cholecystokinin-binding activity. There- fore, the production of two radioiodinated derivatives of ASD-[Thr28,Ahx3']CCK-25-33 may reflect radioiodination at two different positions on the aromatic ring of ASD. How- ever, the latter fraction (retention time 32.5 min) showed a 3 - 4-fold-higher binding ability than the former. This latter fraction was used for the present study.

Binding studies Binding of '251-ASD-[Thr28, Ahx3']CCK-25-33 was car-

ried out on pancreatic membranes under non-actinic light. Binding characteristics were similar to that of usual chole-

Page 4: A new probe for affinity labelling pancreatic cholecystokinin receptor with minor modification of its structure

400

V

._ 7J 5 - K

non-specific

Time (min)

Fig. 2. Reverse-phase HPLC purification of '251-ASD-[ ThrZ8, Ahx3'/CCK-25-33. Peptide (1 0 pg) was radioiodinated in presence of chloramine T as described in Experimental Procedures. After reaction, the medium was diluted with the same volume of eluant containing 50% acetonitrile and injected on a p-Bondapak CIS column. Chromatography was performed in 0.125 M triethylammonium phos- phate, pH 3.5, containing 37.5% acetonitrile at a flow-rate of 1.5 ml/ min. Eluted radioactivity was detected with a radioisotope detector. Fractions were collected and tested for cholecystokinin binding an ability to photoaffinity label the cholecystokinin receptor on pancre- atic membranes. The fraction corresponding to the last peak was extensively used in this study

cystokinin ligands. First, kinetics of binding were dependent on temperature (not illustrated). At 25 "C, (Fig. 3 a) steady state was reached after 60 min of incubation and binding remained stable over a period of at least 3 h. Non-specific binding represented 5 - 10% of total binding. At 2 5 T , half- maximal dissociation in the presence of 1 FM [Thr", Ahx31]CCK-25-33 was observed at 38 min (Fig. 3 b). Second, the probe bound specifically and with high affinity to pancre- atic membranes. Specific binding was inhibited in a concen- tration-dependent manner by either [Thr", Ahx3']CCK-25- 33, ASD-[Thr28,Ahx3']CCK-25-33 or compound L-364-718, a specific antagonist of the A-subtype cholecystokinin recep- tor [6]. Concentrations of these compounds, which inhibited 50% of the binding were 0.4 nM, 2.0 nM and 0.9 nM, respec- tively (Fig. 3c).

Photoaffinity labelling

After binding of 1251-ASD-[Thr28,Ahx3 'ICCK-25-33 to pancreatic membranes, the receptor-ligand complex was ir- radiated under ultraviolet light and analyzed by SDS/PAGE. Fig. 4A shows an autoradiography of a typical gel. Photo- affinity labelling occured on a component migrating as a broad band at Mr 85000- 100000. A non-specific band, prob- ably bovine serum albumin, was detected at M , 67000. In the absence of exposure to light, the component of M , 85000- 100000 was not seen (Fig. 4A, lane 6). Using the 125-W lamp, the intensity of the band increased with photolysis time until 5 min then decreased. This may indicate that a split of the

non-specific

+-+-+/-+-+-+ -t 1 , ,

30 60 90 Time (rnin) A

"0 30 60 90 B Time ( m i d

I Z 0 r . SD -IT hr 28 A hx3'lCCK-25 -33

I

m ^ ^ I \ \ \

._

CCK-25-33

- L 36L 718 \

-12 -10 -9 -a -7 -6 C log [Compoundl/M

Fig. 3. Binding of '251-ASD-(Thr28, Ahx3']CCK-25-33 to pancreatic plasma membranes. (a, b) Time course of association and dissociation. For association, membranes (20 pg/ml) were incubated at 25'C with the radioligand alone (100 pM) (M) or in presence of 1 pM [Thr", Ahx3']CCK-25-33 (+). At a specified time, aliquots were removed in duplicate, diluted with 0.5 ml iced buffer and centrifuged to determine bound ligand. For dissociation, initial binding was performed for 60 min at 25°C. Then, 1 pM [Thrzs,Ahx3']CCK-25-33 was added and at specified times 0.5-ml aliquots were removed and treated as described above. (c) Inhibition of binding at steady state. Membranes (5-10 pg) were incubated at 25°C with radioligand alone (100 pM) or in the presence of various concentrations of reagents: (*) [Thr", Ahx31]CCK-25-33; (+) ASD-[ThrZ8, Ahx3']CCK-25-33; (M) L-364- 718. After 60 min, the incubation was stopped by adding 0.5 ml iced buffer. Membrane-bound radioligand was determined after centrifugation at 12000 x g for 3 min. Results are the mean of five separated assays. All binding assays were performed under non-ac- tinic light

Page 5: A new probe for affinity labelling pancreatic cholecystokinin receptor with minor modification of its structure

40 1

A B LANE^ 2 3 4 5 6 1 2 3 4

D T T - + + + + c + + + + +

Mr .

x s 85.100 [

u v 2x SOURCED 1 2 5 W 1 2 5 ~ 2OW SUN 125W

T I M E 3 30 10 5 2 1 o 5 20 30 30 I rnl.>)

Fig. 4. SDSjPAGE analysis ofphotoaffinity-labelled pancreatic plasma membranes with "'I-ASD-[ Thrzs, Ahx3']CCK-25-33. Membranes (SO pg in A; 10 pg in B) were incubated with 100 pM IZ5I-ASD- [Thrz8,Ahx3']CCK-25-33 for 60 min at 25°C and photolyzed at 4°C for the specified times. (A) Using a 125-W mercury lamp; (B) using the different ultraviolet (UV) sources indicated. After photolysis, the membranes were collected by centrifugation and submitted to SDS/ PAGE in a 9% acrylamide gel. Lane B4 shows labelling pattern using '251-uTyr-Gly[Ahx28~31, P N O ~ P ~ ~ ~ ~ I C C K - 2 6 - 3 3 . The figure shows that cholecystokinin receptor was identified as a broad band at M , 85000 - 100000, whatever the ultraviolet source employed, providing a sufficient time of ultraviolet exposure was used

covalent bond in the photoaffinity-labelled receptor complex occurred if photolysis was prolonged. To prevent such risks of protein damage, we compared several ultraviolet sources for their ability to induce photoaffinity labelling. Fig. 4B shows an identical labelling pattern whatever the ultraviolet source used. However, maximum yield of covalent labelling was achieved after different photolysis times: 5 min for the 125-W lamp (Fig. 4B, lane l), 20 min for the 20-W lamps (Fig. 4B, lane 2) and 30 min for sunlight (Fig. 4B, lane 3). Under the two last ultraviolet sources, no decrease in the band intensity at MI 85000-100000 was seen if irradiation was prolonged (not shown).

Finally, photoaffinity labelling using '251-~Tyr-Gly- [ A ~ X ~ ~ . ~ ' , ~NO~Phe~~lCCK-26-33 irradiated for 30 min (op- timal time) with the 125-W lamp, resulted in identification of the same protein of M , 85000- 100000 (Fig. 4B, lane 4).

However, yield of covalent attachment of '251-~Tyr-Gly- [ A ~ x ~ ~ , ~ ' , ~NO~Phe~~lCCK-26-33 to cholecystokinin recep- tor was 3 -4-fold lower than that of '251-ASD-[Thr28, Ahx3 'ICCK-25-33 (respectively 1 % and 4%).

Furthermore, specificity and affinity of photoaffinity labelling was determined by incubating membranes and 1251- ASD-[ThrZ8, Ahx3']CCK-25-33 in the presence of increasing concentrations of unlabelled peptide or antagonist prior to photolysis. As shown in Fig. 5 , labelling intensity at M , 85000- 100000 was inhibited by [Thr28,Ahx3']CCK-25-33 and L-364-718 antagonist at concentrations closely compa- rable to that inhibiting specific binding on pancreatic plasma membranes (Fig. 3 C).

0 3.10-" 3 .lo-"

I T hr ", Ahx3'lCCK25-33

0 lo-'' lo-'' 3

L 361 718

Fig. 5. SDSjPAGE analysis ofphotoaffinity-labelledpancreatic plasma membranes with 1251-ASD-[ ThrZ8, Ahx3']CCK-25-33. The figure shows inhibition of covalent labelling of the cholecystokinin receptor at M , 85000- I00000 by increasing concentrations of [Thr2*, Ahx3']CCK-25-33 and L-364-718 antagonist. Half-maximal inhi- bition of labelling occurred at concentrations similar to that observed in binding experiments (Fig. 3c). When samples were not reduced by dithiothreitol, the same pattern of labelling was obtained except that a smear was observed along the top part of the gel. DTT, dithiothreitol

Wheat-germ-agglutinin - agarose chromatography and endoproteinase Glu-C cleavage

When membranes were affinity labelled with ' 251-~Tyr- G l y - [ A h ~ ~ ~ , ~ ' , pN02Phe33]CCK-26-33 or '251-ASD-[Thr28, Ahx3 'ICCK-25-33, solubilized by Nonidet-P40 and sub- sequently incubated with immobilized wheat germ agglutinin, a fraction (4%) of incubated radioactivity was adsorbed to the gel and was eluted in presence of 0.5 M N-acetyl-j-D- glucosamine (Fig. 6). SDSjPAGE analysis of the peak content indicated that radioactive molecules migrated at M , 85 000 - 100 000.

In a second set of experiments, eluted proteins from immobilized wheat germ agglutinin were precipitated, enzy- matically digested by endoproteinase Glu-C and the products of digestion analyzed by SDSjPAGE in a 15% acrylamide gel. As shown in Fig. 7, the previous glycoprotein of MI 85000- 100000, which was labelled by either cleavable or intrinsic cholecystokinin probes, was converted to the same M , 30000 - 34000 component by endoproteinase Glu-C diges- tion. This gives additional proof for the structural identity of the photoaffinity-labelled protein.

DISCUSSION

In the current study, we coupled a photoactivable chemi- cal, SASD, to [Thr28,Ahx3']CCK-25-33 via its N-terminus, purified the coupling product and radioiodinated it in the presence of chloramine T. Binding characteristics of the purified radioiodinated peptide, '251-ASD-[Thr28, Ahx3']- CCK-25-33 on rat pancreatic plasma membranes were in good agreement with those previously obtained by others [3,4] and

Page 6: A new probe for affinity labelling pancreatic cholecystokinin receptor with minor modification of its structure

402

T 125000

Mr . lo-'

4 85-100

100000 --

- E

75000 -- x > u 0

"3 0 CK

c .- ._ c

.O 50000- -

25 000 _ _ glucosornine

I I I 1 I I I 5 10 15 20

0 0

Fraction number

Fig. 6. Wheut-germ-agglut~n in - ugurose chromatography of photoaffinity-labelled pancreatic cholecystokinin receptor with lZ5Z- ASD-[ Thr", Ahx31]CCK-25-33. A batch of pancreatic plasma membranes (10 mg protein) was photoaffinity labelled with '251-ASD-[Thr28, Ahx31]CCK- 25-33, solubilized in the presence of Nonidet P-40 and incubated with immobilized wheat germ agglutinin for 15 h at 4°C. After washing, retained components were eluted by 0.5 M N-acetylglucosamine at a flow rate of 1 mljmin. Fractions corresponding to the peak were pooled and ultraconcentrated for subsequent analysis. Insert: SDSjPAGE analysis of an aliquot of the peak content showing that the radioactivity adsorbed to wheat germ agglutinin corresponded to the labelled cholecystokinin receptor

Fig. 7. SDSjPAGE analysis of' endoproteinase-Clu- C-di~ested chole- cystokinin receptor. Membranes were photoaffinity labelled with "1- ASD-[ThrZ8,Ahx3']CCK-25-33 (lane a) or with lZ5I-~Tyr- G l y [ A h ~ ' ~ , ~ I , P N O ~ P ~ ~ ~ ~ I C C K - 2 6 - 3 3 , applied to a wheat-germ- agglutinin - agarose chromatography and digested by endoproteinase Glu-C for 3 h at 37°C. The digestion products were analyzed on a 15% acrylamide gel. The same fragment of M , 30000-34000 was obtained which gives additional proof of the identity of the glyco- protein photoaffinity labelled with both intrinsic and cleavable probes

by us [8]. This allowed us to use 1251-ASD-[Thr28, Ahx3']CCK-25-33 as a probe to photoaffinity label the cholecystokinin receptor in rat pancreas.

However, several studies showed that cholecystokinin pro- bes having different sizes can affinity label distinct membrane proteins. CCK-25-33-based probes have identified predomi- nant glycoprotein estimated at M , between 76000 and 95 000, whereas cholecystokinin-8- and cholecystokinin-10-based

probes have identified a glycoprotein of M , 85 000 - 95 000 which has a protein core different from the former [I1 -141. A multi-subunit structure or heterogeneity of cholecystokinin receptors were speculatively proposed to interpret these re- sults [8, 151. Certainly, the strongest evidence that the binding subunit of cholecystokinin receptor does correspond to the protein of M , 85000- 100000 was provided by synthesis of the new cholecystokinin probe, '251-~Tyr-Gly-[Ahx28331, pN02Phe33]CCK-26-33, which has its photolabile moiety, pNOzPhe sited in the binding region of cholecystokinin [8]. Therefore, we used this intrinsic cholecystokinin reference probe to study the molecular properties of the protein which was photoaffinity labelled by '251-ASD-[Thr28, Ahx3']CCK- 25-33. We showed that both probes photoaffinity labelled a protein migrating at M , 85000- 100000 in SDS/PAGE. This protein was glycosylated since it was retained on immobilized wheat germ agglutinin and subsequently eluted by N-acetyl- p-D-glucosamine. Moreover, endoproteinase-Gly-C cleavage of the labelled glycoprotein generated the same predominant component of M , 30000 - 34000. All these observations taken together demonstrated that the new cleavable probe, lZ5I- ASD-[ThrZ8, Ahx3']CCK-25-33, covalently labelled the bind- ing domain of cholecystokinin on rat pancreatic plasma mem- branes.

In spite of a relatively good knowledge of the pharma- cology of cholecystokinin, little is known about its fine struc- ture and, until now, attempts to purify and partially sequence it have failed. Difficulties are those encountered in the purifi- cation of an extremely minor membrane protein which, in addition, easily loses its ability to recognize cholecystokinin ligands after solubilization. To circumvent these difficulties, we wish to purify a denatured fragment of the cholecystokinin receptor protein which should include the binding site of cholecystokinin peptides and cholecystokinin antagonists. Photoaffinity labelling will be used for tracing the receptor protein fragment during the purification process.

The intrinsic probe, '251-~Tyr-Gly-[Ahx28931, pNOz- Phe33]CCK-26-33, accurately identified the cholecystokinin receptor. However, the photoaffinity reaction via the pNOz

Page 7: A new probe for affinity labelling pancreatic cholecystokinin receptor with minor modification of its structure

403

group on the phenylalanyl residue requires hard photolysis conditions that may structurally damage proteins [S]. More- over, after photolysis, the whole peptide molecule remains linked to the receptor protein, probably at or near the binding domain of cholecystokinin. Consequently, molecular charac- teristics such as mass, charge and hydrophobicity of the photoaffinity-labelled receptor are modified with respect to the unlabelled receptor. Finally, in the intrinsic probe, I2'I is incorporated as the N-terminus, whereas the peptide is covalently attached to the receptor via its C-terminus. If chemical or enzymatic cleavages were then used to generate receptor protein fragments, the peptide moiety bearing the radioactive label might be separated from the rest of the molecule leading to a loss of receptor radiolabelling.

We designed '251-ASD-[Thr2s,Ahx31]CCK-25-33 in order to minimize such disadvantages. This unique cleavable cholecystokinin probe accurately labelled the cholecystokinin reccptor too. The presence of an azidobenzyl moiety at the N-terminus of the peptide allowed us to use a relatively gentle ultraviolet source for photoactivation, and provided a much better yield of covalent attachment than did the p- nitrophenylalanyl cholecystokinin probe. Finally, the peptide moiety of '2SI-ASD-[Thr28,Ahx31]CCK-25-33 can easily be released from the photoaffinity-labelled receptor protein by reduction with dithiothreitol which opens intramolecular dis- ulfide bonds [16]. Thus, it may be predicted that structural modification of the receptor protein due to the remaining radioiodinated moiety is less important than that taking place with the intrinsic probe.

In conclusion, for the first time, a cleavable photoactivable cholecystokinin ligand bearing a "'I atom on its photolabile moiety was prepared. This new probe showed similar binding characteristics as cholecystokinin peptides and covalently identified cholecystokinin receptor in pancreatic plasma mem- branes from rat. This probe should be a useful tool for cholecystokinin receptor characterization and purification and for studying internalization and recycling of chole-

cystokinin receptors in acinar cells. This work is currently in progress in our laboratory.

This work was supported in part by a grant from the European Economic Community, no. STZJ 0001 3 F, and Rigion Midi-PyrPnkes, no. 87/11 1209.

REFERENCES

1. Mutt, V. & Jorpes, J. E. (1971) Biochem. J . 125, 57-58. 2. Jorpes, J. E. & Mutt, V. (1973) in Secretin, cholecystokinin-

pancreozymin andgastrin (Jorpes, J. E. & Mutt, V., eds) pp. 1 - 144, Springer-Verlag, Berlin.

3. Christophe, J., De Neef, P., Deschodt-Lanckman, M. & Robberecht, P. (1978) Eur. J. Biochem. 97, 31 -38.

4. Sankaran, H., Goldfine, I. D., Devenez, C. W., Wong, K . Y. & Williams, J. A. (1980) J . Biol. Chem. 255, 1849-1853.

5. Jensen, R. T., Lemp, G. F. & Gardner, J. D. (1980) Proc. Nut / Acad. Sci. USA 77,2079 - 2083.

6. Chung, R. S. L. & Lotti, V. J. (1986) Proc. Natl Acad. Sci. USA

7. Fourmy, D., Zahidi, A,, Fabre, R., Guidet, M., Pradayrol, L. &

8. Powers, S. P., Fourmy, D., Gaisano, H. & Miller, L. J. (1988) J .

9. Wessel, D. & Flugge, U. I. (1984) Anal. Biochem. 138. 141 -143. 10. Laemmli, U. K. (1970) Nature 227,680-685. 11. Svoboda, M., Lambert, M., Furnelle, J . & Christophe, J. A.

12. Rosenzweig, S. A., Miller, L. J. & Jamieson, J. (1983) J. Cell Bid.

83,4923 -4926.

Ribet, A. (1987) Eur. J . Biochem. 165, 683-692.

Biol. Chem. 263, 5295 - 5300.

(1982) Regul. Peptides 4 , 163-172.

96, 1288-1297. 13. Sakamoto, C., Goldfine, I. D. & Williams, J. A. (3983) J . B id .

Chem. 255, 12707-12711.

J . Biol. Chem. 262,13 850 ~ 13 856.

Biorhem. Biophys. Res. Commun. 143, 761 -767.

G. M. (1968) Biochemistry 7, 1959-1962.

14. Pearson, R. K., Miller, L. J., Hadac, E. M. & Powers, S. P. (1987)

15. Madison, L. D., Jamieson, J. D. & Rosenzweig, S. A. (1987)

16. Wdxdal, M. J., Konigsberg, W. H., Henley, W. L. & Edelman,