Photoaffinity Labeling the ,&Adrenergic Receptor ... - Semantic Scholar

1 downloads 0 Views 3MB Size Report
Apr 20, 1988 - John F. Resek and Arnold E. RuohoS. From the Department of ...... Chorev, M., Feigenbaum, A., Keenan, A. K., Gilon, C., and. 4. Rashidbaigi, A.
THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1988 by The American Society for Biochemistry and Molecular Biology, Inc

Vol. 263, No. 28, Issue of October 5, pp. 14410-14416, 1988 Printed in U.S.A.

Photoaffinity Labeling the ,&Adrenergic Receptorwith an Iodoazido Derivative of Norepinephrine* (Received for publication, April 20, 1988)

John F. Resek and Arnold E.RuohoS From the Department of Pharmacology, University of Wisconsin Medical School,Madison, Wisconsin 53706

An iodinated photosensitive derivative of norepi- tor. ['251]NAIN has been prepared carrier free at a specific nephine, N-(p-azido-m-iodophenethylamidoisobuty1)-activity of 2200 Ci/mmol and used to photolabel the @-adrenorepinephrine (NAIN), has been synthesized and nergic receptor-binding site. Since NAIN is an agonist, it can characterized. NAIN stimulated adenylate cyclase ac- be used to directly study receptor interactions with the guanyl tivity in guinea pig lung membranes in a manner sim- nucleotide-binding protein, G,, desensitization, and downilar to (-)-isoproterenol and was inhibited by (-)-al- regulation. prenolol. NAIN was shown to compete with ['2sI]iodocyanobenzylpindolol for the &adrenergic receptor in MATERIALS ANDMETHODS guinea pig lung membranes with an affinity which was All solvents were reagent grade, except where indicated. Thallium dependent on the presence of guanyl nucleotides. Car- trichloride was purchased from Alpha. Nalz5I (2200 Ci/mmol) was rier-free radioiodinated NAIN (['2sI]NAIN) was used purchased from Du Pont-New England Nuclear. All other reagents at 2 nM to photoaffinity label the &adrenergic receptor were obtained from Sigma or Aldrich. IR spectra were recorded on a in guinea pig lung membranes. Sodiumdodecyl sulfate- Perkin-Elmer 180 spectrometer calibrated via the 1601 cm-' peak of polyacrylamide gel electrophoresis analysis of (-)-al- polystyrene. Proton nuclear magnetic resonance spectra were obprenolol(1 p ~ protectable ) ['2sI]NAIN labeling showed tained on a Bruker 270 MHz spectrometer. All chemical shifts (6) the same molecular mass polypeptide (65 kDa) that were reported in parts/million relative to internal tetramethylsilane was specifically derivatized with the antagonist pho- standard. Mass spectrometric determinations were performed by the tolabel ['261]iodoazidobenzylpindolol. Specific labeling Midwest Center for Mass Spectrometry, a National Science FounRegional Instrumentation Facility (Grant CHE-8620177).Preof the &adrenergic receptor with ['2sI]NAIN was de- dation parative thin layer and analytical thin layer chromatography were pendent on the presence of MgClz and the absence of performed on precoated Merck Silica Gel 60F-254 glass-backedplates, guanyl nucleotide. Guanosine-5'-0-(3-thiotriphos- except where indicated. The synthesis of ['251]ICYPand ['261]IABP phate (100 p ~ abolished ) specific labeling by ['2sI] were performed as previously described (4, 11, 12). NAIN. N-Ethylmaleimide (2 mM) in the presence of Synthesis of 0-(p-Amino-m-iodophenyl)ethylamine (II)-Sodium ['2sI]NAIN protected against the magnesium and iodide (2100 mg,14.0 mmol) and 0-(paminopheny1)ethylamine (I) guanyl nucleotide effect. These data show that NAIN (1906 mg, 13.8 mmol) were dissolved in 400 ml of sodium acetate is an agonist photolabel for the &adrenergic receptor. buffer (0.1 M, pH 4.1), followed by the addition of thallium trichloride

(5200 mg, 16.7 mmol) in 100 ml of water over a 30-min period. The solution was heated on a steam bath for 1 h under nitrogen, at which time the reaction was stopped by the addition of sodium sulfite (1760 mg, 13.9 mmol) in 40 ml of water. After the reaction had cooled to Several radiolabeled ligands which covalently react with room temperature, the solution was alkalinized, pH 9.0, with sodium the @-adrenergicreceptor are currently available. These in- carbonate and extracted with 50 ml of chloroform three times. The clude antagonist affinity labels, whichhave been used to chloroform fractions were combined, and the solution was dried with sulfate, filtered, and the solvent removed by flash evapderivatize the purified receptor (1) and an associated lipid magnesium oration. An oily brown residue remained in the round bottom flask. component (2, 3), as well as antagonist photoaffinity labels, Thin layer chromatographic analysis indicated that the reaction had whichwere first used to identify the membrane-bound @- gone to completion. Visualization with a mineral lamp showed that adrenergic receptor on sodium dodecyl sulfate-polyacrylamide the product was one spot that migrated separately from the starting gel electrophoresis (4-8). Agonist affinity labels have been materialintoluene/isopropylalcohol/ammoniumhydroxide synthesized and reported to covalently react with the receptor (85:30:5), ( R F = 0.5) and chloroform/methanol/ammonium hydroxide (9, lo), but the @-adrenergicreceptor has not been covalently (50:50:5), (RF= 0.3). The product was used without further purification. The yield was 48% (1770 mg, 6.7mmol). radiolabeled with an agonist photoaffinity label. In thispaper, Synthesis of 0-(p-Azido-m-iod0phenyl)ethylamine (III)-The prodwe report the synthesis and characterization of N-(p-azido- uct, P-(p-amino-m-iodopheny1)ethylamine(11) (1000 mg, 3.8 mmol), rn-iodophenethylamidoisobuty1)norepinephrine (NAIN),' an was dissolved in 58 mlof water containing 1.3 mlof 96% sulfuric agonist photolabel which derivatizes the @-adrenergicrecep- acid. The solution was cooledon ice, and sodium nitrite (420 mg, 6.08 mmol) in 32 ml of water was added dropwise to the stirring solution. * This work was supported by National Institutes of Health Grant When the addition was completed, the reaction was allowed to proGM 33138. The costs of publication of this article were defrayed in ceed for 15 min, followed by the addition of sodium azide (434 mg, part by the payment of page charges. This article must therefore he 6.7 mmol) in 4 ml of water. After 1 h, the reaction was extracted with hereby marked "aduertisement" in accordance with 18U.S.C. Section 30 ml of chloroform, and thechloroform was washed once with 20 ml of 2 N HCl. The reaction mixture and acid wash were combined, and 1734 solelyto indicate this fact. the aqueous mixture was alkalinized with sodium carbonate and $ To whom correspondence should he addressed. times with 30ml of chloroform. The chloroform I The abbreviations used are: NAIN, N-(p-azido-m-iodophenethyl- extractedthree amidoisohuty1)norepinephrine; ICYP, iodocyanobenzylpindolol; fractions were combined, andthe solvent wasremoved by flash IABP, iodoazidohenzylpindolol; NEM, N-ethylmaleimide; HPLC, evaporation. Thin layer chromatographic analysis of the oily residue high pressure liquid chromatography; EGTA, [ethylene- indicated that the reaction had gone to completion. The product was his(oxyethylenenitrilo)]tetraacetic acid; GTPyS, guanosine-5'-0-(3- a single spot on silica gel, migrating above the starting material in both toluene/isopropyl alcohol/ammonium hydroxide (85:305) (RF= thiotriphosphate); G,, guanyl nucleotide-binding protein;

14410

A Catecholamine PhotoaffinityLabel 0.6) and chloroform/methanol/ammonium hydroxide (50505) ( R F= 0.4). The IR spectrum showed a sharp stretch at 2100 cm", which confirmed the presence of an azide, and a broad stretch at2950 cm", which was consistent with a primary amine. The NMR (deutrochloroform) was consistent with a molecule having three phenyl protons, two methylene groups, and a primary amine: 6 7.60 (s, 1 H, aromatic H ortho to I),6 7.4 (m, 1 H, aromatic), 6 7.2 (m, 1H, aromatic), 6 3.1 (t, 2 H), 6 2.8 (t, 2 H), 6 1.0 (s, 2H). The product was obtained in 45% yield (580 mg, 1.7 mmol) and used without further purification. Synthesis of (p-azido-m-iodophenyl)ethylamidobutylketone(V)mmol) and 0-(p-azido-m-iodoLevulinic acid (IV) (362mg,3.12 pheny1)ethylamine (111) (300 mg, 1.05 mmol) were dissolved in 2.5 ml of dimethylformamide, followed by the addition of l-cyclohexyl3-(2-morpholinoethyl) carbodiimidemetho-p-toluenesulfonate(1587 mg,3.75 mmol) in 3.75 mlof dimethylformamide. The reaction proceeded at 4 ' C for 16 h, and then the contents of the reaction vesselwere added to 150ml of rapidly stirring chloroform. The chloroform was extracted two times with 50 ml of 0.1 M Na2P04, pH 7.0, and five times with 50 ml of water. The chloroform was removed by flash evaporation, yielding 300 mg of product. The residue, which was analyzed by silica gel thin layer chromatography in a toluene/ acetonitrile/acetic acid (20:12:0.2) solvent system, showed a major (RF= 0.5) and a minor (RF= 0.45) component. Further analysis by thin layer chromatography showed that themajor component reacted to give NAIN, suggesting that it was authentic(p-azido-m-iodopheny1)ethylamidobutylketone(V). The minor component was found to be unreactive and could be separated from NAIN by thin layer chromatography. Therefore, compound V was used without further purification. Three-hundred of mg crude (p-azido-m-iodopheny1)ethylamidobutylketone(V)were obtained/reaction. Synthesis of N-(p-azido-m-iodophenethylamidoisobutyl)norepinephrine (VI1)-(-)-Norepinephrine was prepared as the acetate salt by dissolving the (-)-norepinephrine-free base (220 mg, 1.3 mmol) in 2 ml of methanol and 0.2 ml of acetic acid and removing the solvent by vacuum. Prior to the reaction, the (-)-norephinephrine acetate salt and NaCNBH3 were dried in uacuo over PzOS.Crude (p-azido-miodopheny1)ethylamidobutylketone(V)(100 mg) and NaCNBH3 (16 mg, 0.26 mmol) were dissolved in 2.5 ml of HPLC grade methanol which had been stored over molecular sieves to remove water, and the solution was added to a test tube containing the (-)-norepinephrine acetate salt. The reaction was allowed to proceed for 4 h unde; argon in the presence of activated molecular sieves (1/16 inch, 4 A pore diameter) (13). The reaction solution was added dropwise to a rapidly stirring solution of 0.1 N HCl (130 ml). A white precipitate formed during the addition. The aqueous solution was allowedto stir for 15 min and thenextracted three times with 50 ml of ethyl acetate. The aqueous layer was discarded and theorganic solvent removed by flash evaporation. The residue was dissolved in methanol and approximately 20 mg spotted on a Whatman PLK5F preparative silica gel plate which was developed in chloroform/methanol/acetic acid (24060:12). The product was visualized with a mineral lamp ( R F= 0.5), and thecompound was extracted from the silica gel with ethanol. The ethanol was removed by flash evaporation, and the product formed crystal plates (melting point 89-90). The crystals were analyzed for purity by thin layer chromatography on silica gel in chloroform/methanol/acetic acid (4010:2) (RF= 0.5) and on Baker SiC18F reverse-phase silica plates in ethanol, 0.1 N HC1 (5050) (RF= 0.5). In both systems NAIN showed a single UV positive spot which separated from (-)-norepinephrine and the (p-azido-m-iodopheny1)ethylamidobutylketone (V). Analysis of NAIN by fast atom bombardment mass spectroscopy revealed an ion for the expected formula of 540.1120 (theoretical mass 540.1108, error 2.1 ppm). High resolution mass spectrometry revealed the composition of the molecular ion to be C21HZ,N50J. The NMR, IR, and UV spectral data for NAIN is as follows: IR spectra (potassium bromide): 3450cm-' (hydroxyl), 2140cm" (azide), 1650 cm" (amide), 1590 cm", 1430 cm" (aromatic); NMR (deuteromethanol): 6 7.6 (s, 1 H, aromatic H ortho to I),6 7.2 (d, 1 H, aromatic Hpara to I), 6 7.1 (d, 1H, aromatic H meta to I), 6 6.8 (s, 1 H, aromatic H), 6 6.7 (m, 2 H, aromatic H), 6 4.7 (t, 1 H, H on C a to catechol), 6 3.5 (m, 1 H), 6 3.3 (m, 2 H) , 6 2.9 (m, 4 H), 6 2.7 (t, 2 H, H on C a to iodo azido phenyl ring), 6 2.2 (m, 3 H), 6 1.6 (m, 2 H), 6 1.3 (s, 1 H), 6 1.1 (m, 3 H); UV (methanol): major peak a t 261 nm, ~ 2 = 6 ~1010 M" cm". Sixty-five mg (0.12 mmol) of pure NAIN could be obtained from 180 mg of starting crude (p-azido-m-iodopheny1)ethylamidobutylketone (V). Synthesis of f25Z]~-(p-amino-m-iodophenyl)ethylamine (IZ)-To 4 mCi of NalZ5I(2200 Ci/mmol) in 10 pl of 0.1 N NaOH was added 10 p1 HC1 )0.1 N), 25 pl of sodium acetate buffer (0.1M, pH 4.2), 5 pl of

14411

P-(p-aminopheny1)ethylamine(13 mg/ml in sodium acetate buffer), and 10 p1 of thallium trichloride (50 mg/ml in sodium acetate buffer). The vial was capped and heated for 1 h on a steambath, followed by the addition of 300 p1 of sodium sulfite (120 mg/ml). The solution was transferred to a thick-walled conical tube and extracted three times with 300 p1 of ammonium hydroxide-saturated chloroform. The combined chloroform washes were evaporated under a nitrogen stream. The remaining residue was dissolved in 100 p1 of chloroform. The product was completely separated from starting material by silica gel thin layer chromatography in toluene/isopropyl alcohol/ammonium hydroxide (85:305). The product was visualized by autoradiography and extracted from the corresponding silica gel with ammonium hydroxide-saturated chloroform. Since the product was completely separated from the starting material, it was assumed to be carrier free. This material showed a single spot which co-migrated on silica gel with authentic P-(p-amino-m-iodopheny1)ethylamine thin layer chromatography in toluene/isopropyl alcohol/ammonium hydroxide (85:30:5) and chloroform/methanol/ammonium hydroxide (11) (50505). Theyield of [1251]~-(p-amino-m-iodophenyl)ethylamine from NalZ5Iwas 95% (3.8 mCi). Synthesis of f25Z]~-(p-azido-m-iodophenyl)ethylamine(III)-Chloroform containing the ['251]/3-(p-amino-m-iodophenyl)ethylamine (11) was evaporated with a nitrogen stream. To assure complete removal of ammonium hydroxide, the residue was twice redissolved in 1 ml of chloroform and taken toa 50-pl volume with nitrogen. The chloroform was then removed by evaporation, and 100 pl of 3% H&O, was quickly added to thereaction vessel. The reaction was cooled on ice for 10 min. Sodium nitrite (10 pl, 69 mg/ml) was added, and the reaction was cooled for an additional 20 min. Sodium azide (50 pl, 65 mg/ml) was added, and the reaction was allowed to proceed on ice for 30 min. The reaction was stopped by adding 1 ml of 2% Na2C03 and extracting the radioactivity into ethyl acetate. The ethyl acetate was washed twice with water, and the aqueous layer was discarded. The radioproduct showed a single spot which comigrated with P - ( p azido-m-iodopheny1)ethylamine(111) on a silica gel plate in chloroform/methanol/ammonium hydroxide (50:5050) and toluene/isopropyl alcohol/ammonium hydroxide (85:305). The yield of [Iz5I]/3-(pazido-m-iodopheny1)ethylamine(111) was 80% (3.0 mCi). Synthesis of f2SZ](p-azido-m-iodophenyljethylamido-butylketone (V)-Ethyl acetate containing the [1251](l-(p-azido-m-iodopheny1)ethylamine (111) was evaporated with a nitrogen stream. Dimethylformamide (50 pl) containing l-cyclohexyl-3-(2-morpholinoethyl) (10 mg) carbodiimidemetho-p-toluenesulfonate and levulinic acid (5 mg) was added to the residue. The reaction was allowed to proceed on ice for 17 h. The radioproduct was purified by silica gel thin layer chromatography in toluene/acetonitrile/acetic acid (20:12:0.2) (RF= 0.5). The product was identified by autoradiography and extracted from the silica gel with 300 p1 of ethyl acetate three times. The radioactive product showed one spot by thin layer chromatography analysis in toluene/acetonitrile/acetic acid (2012:0.2) and comigrated with authentic compound V. One mCi of pure [1251] (p-azido-m-iodopheny1)ethylamidobutylketone (V)could be obtained (111) (33% from 3 mCi of [1251]~-(p-azido-m-iodophenyl)ethylamine yield). Because of the rapid decomposition of the radioiodinated intermediates (11) and (111), the first three reactions were performed in a single day. Synthesis of f25I]N-(p-azido-m-wdophenethylamidoisobutyl)-norepinephrine (VZZ)-Ethyl acetatecontaining the [1251](p-azido-miodopheny1)ethylamidobutylketone(V)was evaporated using a nitrogen stream. HPLC grade methanol (50 pl) containing (-)-norepinephrine as the acetate salt (36 mg/ml) and NaCNBH3 (12 mg/ml) were added to the residue. The reaction proceeded for 4 h over a single activated molecular sieve bead (1/16 inch, 4 A pore diameter) (13) and under argon. Purification of the product from (-)-norepinephrine and theradioactive starting material was performed by TLC in chloroform/methanol/acetic acid (40:10:2). The radioproduct was identified by autoradiography (RF= 0.5) and extracted from silica by suspending the gel in 300 pl of 0.1 N HCl and washing the slurry four times with 700pl of ethyl acetate. The radioactivity entered the organic phase. Extreme care was taken to remove all silica gel from the preparation by washing the organic phase twice with 300 p1 of water. The ethyl acetate was then transferred to a conical tube and taken to a volume of approximately 20 p1 with an argon stream. The radioproduct was then transferred to a microcentrifuge tube. The volume was increased with methanol so that 4 pl of [1251]NAINcould be added to a 200-pl suspension of membranes to give a final NAIN concentration of 2 nM. The [lZ5I]NAINwas a single spot which comigrated with authentic NAIN on silica gel thin layer chromatog-

14412

A Photoaffinity Catecholamine

raphy (RF= 0.5) in chloroform/methanol/acetic acid (40:10:2) and on C-18 reverse-phase plates (RF= 0.5) in ethanol, 0.1 N HCl (50:50). Thirty pCi of ['251]NAINat a specific activity of 2200 Ci/mmol could be obtained from 250 pCi of starting ['251](p-azido-m-iodopheny1)ethylamidobutylketone (V) (12% yield). Membrane Preparation-Guinea pigs(300-500g)were executed by decapitation and the lungs removed and stored in liquid nitrogen. To prepare membranes, the frozen lungs were minced in the presence of a buffer containing 50 mM Tris-HC1, pH 7.4,5 mM EGTA, 100 p M phenylmethylsulfonyl fluoride, 100 p~ benzamidine, 5 pg/ml of soybean trypsininhibitor, and 20 pg/ml of leupeptin. One g of the preparation was then suspended in 20 ml of buffer and homogenized with a polytron homogenizer. The mixture was centrifuged at 2000 rpm for 10 min. The supernatant was removed and centrifuged at 15,000rpm for 45 min. The pink fraction of the pellet was resuspended in 10 ml of buffer and stored in aliquots a t -79°C until needed. Binding Assay-Membranes (25 pg of protein) were incubated with ['251]ICYP(80 PM) and competing drug in a total volume of 0.25 ml 10 mM Tris,pH 7.5, 4 mM MgC12, lo" M catechol, and lo" M ascorbate. Catechol and ascorbate were included in all photolabeling and binding experiments to prevent oxidation of NAIN. Samples were incubated for 60 min at 30°C. Binding was terminated by diluting 50-p1 aliquots in duplicate into 10 ml of ice-cold 10 mM Tris, pH 7.5, and 4 mM MgClz at 4°C and rapidly filtering through a glass fiber filter (Whatman G/FA) under vacuum. Each filter was washed with an additional 10 ml of buffer. The radioactivity content of the filter paper was determined using aPackardauto gammamodel800C gamma counter with 65% efficiency. Nonspecific binding, defined as the amount of '"I bound in the presence of 10 p~ (-)-alprenolol, was usually less than 5% of the total binding. All values represent the mean & S.E. from a t least three experiments. The specific activity of ['251]ICYPwas assumed to be 2200 Ci/mmol (14). Adenylate CyclaseAssay-Enzyme activity was determined by measuring the amount of 32P-cyclicAMP formed from [cx-~'P]ATP according to the method of Salomon et al. (15). The reaction was carried out in a final volume of 0.25 ml containing 50 mM Tris-HCI, pH 7.5, 1 mMMgC12, 1 mM isobutyl methylxanthine, 0.1 mM ATP, 10 p~ GTP, 5 units of creatine phosphokinase, 2 mM creatine phosphate, 2.2 X 10' cpm of [cx-~'P]ATP,and agonist. Reactions were carried out for 10 min at 30°C and terminated by the addition of 0.5 ml of 10% trichloroacetic acid. Approximately 10,000 cpm of 3Hcyclic AMP in 0.25 ml of 0.1 mM cyclic AMP was added as a tracer to each sample to determine recovery. Samples were centrifuged for 30 min at 1500 rpm at 4°C in an international portable refrigerated centrifuge. 32P-CyclicAMP was isolated using consecutive columns containing Dowex AG 50 W-X4 resin and 200-400 mesh neutral alumina, as described by Salomon et al. (15). (-)-Isoproterenol and NAIN-stimulated adenylate cyclase activity was linear with time for 15 min. Photolabeling-Photolabeling of guinea pig lung membranes was performed by incubating membranes at 22°C for 1h in the dark with radioligand in the presence or absence of protector. The membranes were at a protein concentration of 1 mg/ml and were suspended in 200 pl of degassed buffer containing 50 mM Tris, 4 mM MgClz, pH 7.4, catechol M), ascorbate M), and 1 mM glutathione. In conditions where MgCl, was absent from the buffer, 4 mM EDTA was included. ['261]NAIN wasincubated at a concentration of 2 nM, while ['251]IABP was used a t a concentration of0.5nM. Following incubation, membranes were diluted in 5 mlof ice-cold degassed incubation buffer. Immediately following dilution, the membranes were photolyzed at 4" C through a 2-mm thick Pyrex tube for 4 s at a distance of 10 cm from a 1 kilowatt mercury lamp (4, 16). The membranes were then pelleted by ultracentrifugation a t 300,000 X g in a Beckman type 65 rotor. The membranes were solubilized in a buffer containing 10 mM Tris, pH 6.8, 2% sodium dodecyl sulfate, 10% glycerine, 5% 2-mercaptoethanol, and 0.05% bromphenol blue dye and subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis on a 9%Laemmli gel (17). Resultswere usually obtained with a 7-h autoradiogram using a Quanta I11 intensifier. The molecular weight of the receptor was determined using the following molecular weight standards: bovine erythrocyte carbonic anhydrase, 29 kDa; egg albumin, 45 kDa; bovine plasma albumin, 66 kDa; rabbit muscle phosphorylase b, 97 kDa; Escherichia coli p-galactosidase, 116 kDa; rabbit muscle myosin, 205 kDa.

Label

+

N

A

I 0

no. W

\N /

Cnrbodiimidc

(V)

1

OH H

N

+

N

A

N

3

(VW

FIG. 1. Synthesis of NAIN. Iodination on P-(p-aminopheny1)ethylamine (I) was performed with thallium trichloride and the product (11) converted to thecorresponding azide (111). Levulinic acid (IV) was coupled with the azidoiodophenylamine using a carbodiimide to produce (V). (R)-Norepinephrine (VI) was reacted with the resulting amide in the presence of NaCNBH3 to produce (R, RS)-(p-azido-m-iodophenethylamidoisobutyl)norepinephrine, NAIN (VII). RESULTS AND DISCUSSION

Synthesis of NAIN-The synthetic scheme and the structure of NAIN is presented in Fig. 1. The synthetic protocol for NAIN was similar to that described by Jacobson et al. (18).The product was prepared in nonradioactive form and carrier-free form with lZ5I.The nonradioactive compound was used for chemical characterization, competitive displacement experiments, and adenylate cyclase assays, whereas the radioiodinated compound was used for photolabeling the &adrenergic receptor. The ['251]@-(p-amino-rn-iodopheny1)ethylamine (11) and the ['251]~-(p-azido-rn-iodopheny1)ethylamine (111) were found to be labile and required immediate use in the next reaction. Both compounds decomposed at -20°C overnight. The intermediate compound, [1261] (p-azido-rn-iodopheny1)ethylamidobutylketone (V) was stable at -20°C for several weeks. During the final synthesis of ['251]NAIN, it was found that removal of all silica from the ethylacetate wasvery important. The presence of silica resulted in adecreased photolabeling efficiency which may be a result of surface catalyzed oxidation of ['251]NAIN.Carrierfree ['251]NAIN wasstable in methanol at -79°C for at least 2 weeks. The synthetic protocol that has been outlined has several novel features. Among these is the use of NaCNBH3to couple the (p-azido-rn-iodopheny1)ethylamidobutylketone(V) to

3

A Catecholamine PhotoaffinityLabel (-)-norepinephrine by reduction of the Schiffs base intermediate. Cyanoborohydride is a mild reducing agent which stabilizes the catecholamine against oxidation while sparing the easily reduced phenyl azide moiety. Additionally, the two (111) and intermediates, @-(p-azido-rn-iodopheny1)ethylamine (p-azido-rn-iodopheny1)ethylamidobutylketone (V), have general applications to the development of photoaffinity labels. The @-(p-azido-rn-iodopheny1)ethylamine(111), which can generate photoaffinity labels from compounds possessing a carboxylic acid, has recently been used to synthesize forskolin (19) andcymarin (20) azidoiodophenyl derivatives. Moreover, the (p-azido-rn-iodopheny1)ethylamidobutylketone(V) can reductively alkylate aminesandmaintaina full positive charge. Agonist Properties of NAIN-The agonist properties of NAIN were determined by stimulation of adenylate cyclase and by competitive displacement of ['251]ICYPin the presence and absence of guanyl nucleotide. The adenylate cyclase dose response curves for NAIN and (-)-isoproterenol in guinea pig lung membranes are shown in Fig. 2. The Kactvalues, whichwere determined using the computer program ROSFIT, were found to be 70 nM for NAIN and 63 nM for (-)-isoproterenol. The same maximal stimulation of adenylate cyclasewas seen for (-)-isoproterenol M, 42.6 f 1.63 pmol/min/mg of protein, n = 3) and NAIN M, 41.7 f 2.54 pmol/min/mg of protein,n = 3). M ) and NAIN M) were When (-)-isoproterenol incubated together, therewas no increase in adenylate cyclase activity over the maximal stimulatedvalues (41.7 f 1.76 pmol/ min/mg of protein, n = 3). Fig. 3 shows that the@-adrenergic antagonist, (-)-alprenolol, competes with (-)-isoproterenol M ) and NAIN M ) for stimulation of adenylate cyclase activity. These data illustrate that NAIN is a full agonist for the @-adrenergicreceptor. Competitive binding assays using nonradioactive agonists to displace radiolabeled antagonists have shown biphasic displacement curves in the absence of guanyl nucleotide (21). Computer analysis of these results has shown that agonists bind to the receptor in high affinity and low affinity binding states (22). The addition of guanyl nucleotide results in a decrease in thenumber of high affinity binding sites, resulting in a monophasic low affinity binding curve. Incontrast, antagonists are unaffected by guanyl nucleotide, and analysis of the binding data shows that thecurve fits a one-site binding 120

,

-10 -8-9

-7 -6 log[Agonist] (Molar)

-5

-4

FIG. 2. Adenylate cyclase stimulation by NAIN and (-)isoproterenol. NAIN (W) and (-)-isoproterenol (0)were shown to stimulate adenylatecyclase activity in a dose-dependent manner. (-)Isoproterenol (lo-' M) and NAIN (lo-' M) gave the same maximal stimulation of adenylate cyclase. Adenylate cyclase activity was measured by a modified method of Salomon et a1 (15). Basal levels were 69 pmol/min/mg of protein. One-hundred% stimulation represents an increase of 42 pmol/min/mg of protein. All values are themean f S.E. for at least three independent observations.

'*O

14413

I 1

i3

I

\\

5

3 .-

20 400 1-"J -10

-9

-8 -7 -6 -5 log [Alprenolol] (Molar)

-4

-3

FIG. 3. (-)-Alprenolo1 inhibition of NAIN and (-)-isoproterenol-stimulated adenylate cyclase activity. NAIN (lo-' M ) (W) and (-)-isoproterenol (lo-' M) (0) were incubated in the presence of increasing concentration of (-)-alprenolol. All values are the mean f S.E. for at least three independent observations.

model. It has been shown that the fraction of high affinity agonist sites represents receptors which are coupled to the guanyl nucleotide-binding protein (G,) (23-25). When guanyl nucleotide binds to G,, the ternary complex dissociates, resulting in a loss of the high affinity state. The high affinity ternary complex in which the agonist, receptor, and G, are associated is likely to be an intermediate state in agonist stimulation of adenylate cyclase (26, 27). Fig. 4 shows that NAIN competes for the binding sites occupiedby the @-adrenergicantagonist ['251]ICYP.Inhibition of ['251]ICYPbinding by both NAIN (A) and (-)-isoproterenol (B) was biphasic in theabsence of GTPyS. When assays were carried out in the presence of GTPyS M), the apparent affinity for both agonists decreased, and the competition curves became monophasic. The results were analyzed by the computer program LIGAND (28), assuming the presence of either one class of receptors or two classes of receptors. In theabsence of GTPyS, thecompetitive binding curves for (-)-isoproterenol and NAIN fit significantly better ( p < 0.001) to a model, assuming two independent classes of receptors. The percentage of high affinity and low affinity sites for (-)-isoproterenol was 39.6 f 4 and 59.4 f 4% ( n = 3), respectively, while for NAIN the proportion was 50.6 f 9 and 49.4 f 9% ( n = 3), respectively. With the addition of GTPyS, no high affinity sites were observed for either agonists. The number of binding sites was determined by [lZ5I] ICYP to be 339 f 26 fmol/mg of protein using ['251]ICYPbinding isotherms. These competitive binding assays show that NAIN binds to the receptor while it is in the ternary complex with G,. The competitive binding curves were used to determine the equilibrium dissociation constants for NAIN and (-)-isoproterenol. The K d for ['251]ICYP wasdetermined by Scatchard analysis to be14.37 f 0.34 PM ( n = 3). The equilibrium dissociation constants in the presence and absence of GTPyS were determined by LIGAND and are shown in Table I (28, 29). NAIN showed high affinity binding to the @-adrenergic receptor whichwas similar to (-)-isoproterenol (Kd high). However, under conditions which did not favor interaction with G, ( K d low and K d GTPyS), affinity to the @-adrenergic receptor was higher for NAIN than for (-)-isoproterenol. Photolabeling the @-AdrenergicReceptor with pz5I]NAIN and Pz5IJIABP-Fig. 5 shows a 7-h autoradiogram illustrating specific photolabeling of the @-adrenergicreceptor in guinea pig lung membranes by ['251]NAIN. The 65-kDa protein photolabeled by ['251]NAIN was protected by the @-adrenergic ligands, (-)-alprenolo1 M), (f)-propanol01 M), and

A Catecholamine Photoaffinity Label

14414

A 100

W

a (8

s?

D

E

.."

.

"

116 b 97 b

2,

o c

C ",

205,

: 80

B

60

E

67 b

40

20

45 b

0 -10

-7

-8 -9

-6

-5

-4

-3

29 b

log [NAIN] (Molar)

FIG.5. Photolabeling of the guinea pig lung B receptor with ['261]NAIN.['2sII]NAIN (2 nM) was incubated in the absence ( A ) or presence (&E) of adrenergic ligands in degassed buffer containing 50 mM Tris, 4 mMMgC12, ascorbate (10" M), catechol (10" M), and 1 mM reduced glutathione. The addition of the &adrenergic ligands (-)-alprenoIoI M ) ( B ) , (+)-propranolol (IO6 mM) ( C ) , or (-1pindolol M ) (D) resulted in the loss of specific labeling. Incubation of membranes with the al-adrenergicligand (-)-prazosin (10" M ) ( E ) did not affectphotolabeling. Prior to sodium dodecyl sulfatepolyacrylamide gel electrophoresis, ['2sII]NAIN bound to membrane pellets was determined in the absence (210,034 4,270 cpm, n = 5) or presence (136,097 f 5,000 cpm, n = 4) of &adrenergic protectors. The average specific binding was tinerefore 74,000 cpm. The protectable photolabeled polypeptide contained 2500 cpm, indicating that 3% of the specifically bound ligand had inserted into the receptor. Most of the ligand that wasnonspecifically bound did not react covalently with membrane proteins. In this preparation, there may have been a small amount of proteolysis, resulting in B-adrenergic polypeptides of smaller molecular weight.

+

-10

-9

-8

log

-7

-6

-5

-4

-3

[Isoproterenol] (Molar)

FIG.4. Competitive displacement isotherms for NAIN and (-)-isoproterenol. A, in the absence of GTPyS, the competitive (A) was biphasic and spanned four displacement curve for NAIN orders of magnitude. The addition of GTPyS (A)to the incubation mixture resulted in a sharper monophasic displacement curve, without changing the total number of binding sites. B, shows the competitive displacement curve for (-)-isoproterenol in the absence (0)or presence (W) of GTPyS. Values represent the mean of three experiments carried out in duplicatef S.E.

A

116 b 97 b 67 b

Kdhigh) -GTPyS

&(low) +GTPySKd

E

F

G

H

-

+

I

29 b

NAIN

nM

6.12 f 1.98

3.09 f 0.77

/ \

D

45 b

~

nM

C

-

205 b

TABLEI Binding constantsfor isoproterenol and NAIN in guineapig lung membranes T o determine the binding affinity of both isoproterenol and NAIN, the competitive binding results were analyzed using the tomputer program LIGAND. In the absenceof GTPyS, a two-site model gave a significantly better fit to the data than a one-site model, whereas in the presence of GTP-yS a one-site model is preferred. All values remesent the mean h S.E. for three seDarate exDeriments. Isoproterenol

B

425 f 80 921 f 12

+

81 13.8 165 f 5.7

(-)-pindolo1 (1O"j M), but was not protected by the al-adrenergic ligand, (-)-prazosin (1O"j M). Specific photolabeling of the @-adrenergic receptor by theantagonist ['2sI]IABP is shown in Fig. 7, A and B. The results shown in Fig. 6 are consistent with ['251]NAIN being an agonist for the @-adrenergicreceptor. Incubation of membranes with ['2sI]NAIN andGTPyS M) resulted in a loss of photolabeling of the 65-kDa protein (Fig. 6C). In contrast, photolabeling by the antagonist['2sI]IABPwas not

Alprenolol GTPxS Mg+* NEM

+

+

+

-

+ + -

I

+ + +

-

-

"

-

+ +

+ I

'

FIG.6. The agonist properties of [1261]NAINin guinea pig lung membranes. [12sI]NAIN(2 nM) was incubated in the absence ( A ) or presence ( B ) of (-)-alprenolo1 M). The addition of M) ( C ) resultedinthe loss of the specific labeling. GTPyS Incubation of membranes with NEM (10" M) prior to the addition The removal of of GTPyS protected against theloss of labeling (D). MgC12 from the buffer and the additionof 4 mM EDTA ( E ) resulted in the loss of the specific labeling. Incubation of membranes with NEM (lo" M ) in the samebuffer as ( E )protected against theloss of labeling (F).The Coomassie stainingpattern is shown, corresponding to A and B (G,H).

A Catecholamine Photoaffinity Label

A

B

C

D

205 b 116 b 97 b 67 b 45 b

29 b

Alprenolol GTPXS Mg+* Flc. 7. Photolabeling of the guinea pig lung B receptor by theantagonist, [12'II]IABP. The &adrenergic antagonist, [12sI] IABP, (0.5nM) photolabeled the same molecular weight polypeptide M ) protectable manner as ['251]NAIN( A ) in an (-)-alprenolo1 ( B ) .The addition of GTP-ySdid not result inthe loss of photolabeling by ['*'I]IABP ( C ) . The removal ofMgC12 from the buffer and the addition of 4 mM EDTA did not result in the loss of photolabeling by ['2sI]IABP( D ) .

14415

(-)-isoproterenol. We have shown that NAIN photolabels the high affinity state of the P-adrenergic receptor. Competitive ['2sII]ICYP-binding assays and receptor photolabelingwith NAIN showed sensitivity to the presence of guanyl nucleotides andwas dependent onmagnesium, indicating thatNAIN binds to the receptorwhile it is coupled to G.. These results indicate that NAIN is an agonist photolabel for the P-adrenergic receptor. The photolabels currently available for studying the /3adrenergic receptor have been antagonists whichhave the limitation of allowing only indirect observation of agonistrelated events. The agonist photolabel, ['2'I]NAIN, can be used to derivatize the receptor in the ternarycomplex. Since ['2'I]NAIN is an iodinated, carrier-free agonist, it may also prove useful for receptor binding studies. These properties will make NAIN a valuable tool for studying receptor interaction with G. as well as agonist-mediated down-regulation. Acknowledgments-We wish to thank Marty Arbabian and Greg Hockerman for preparing guinea pig lung membranes and Brian LeBlanc for preparing NMR data. We wish to also thank Brian Wadzinski, Dr. N. Dhanasekaran, Joseph Lowndes, and Richard Vaillancourt for their assistance and valuable discussions.

REFERENCES

1. Dickinson, K. E.J., Heald, S. L., Jeffs, P. W . , Lefkowitz, R. J., and Caron, M. G. (1985)Mol. Pharmmol. 27,499-506 2. Bar-Sinai, A., Aldouby, Y., Chorev, M., and Levitski, A. (1986) EMBO J. 5, 1175-1180 3. Chorev, M., Feigenbaum, A., Keenan, A. K., Gilon, C., and Levitski, A. (1985)Eur. J. Biochem. 146.9-14 4. Rashidbaigi, A., and Ruoho, A. E. (1981)Proc. Nutl. Acud. Sci. U. S. A. 78,1609-1613 affected by GTPyS M) (Fig. 7C). The loss of photo5. Rashidbaigi A., and Ruoho, A. E. (1982)Biochem. Biophys. Res. labeling by ['2'I]NAIN suggests a decrease in agonist affinity Commun. 106, 139-147 6. Rashidbaigi A., Ruoho, A. E., Green, D.A., and Clark, R.B. caused by guanyl nucleotide dissociating the ternary complex, (1983)Proc. Nutl. Acud. Sci. U.S. A . 80,2849-2853 analogous to the results obtained in the competitive binding 7. Lavin, T. N., Nambi, P., Heald, S. L., Jeffs, P. W . , Lefkowitz, R. assays. Additionally, it was found that['2'I]NAIN was unable J., and Caron, M. G. (1982)J. Biol. Chem. 257,12322-12340 to photolabel the receptorif magnesium was absent from the 8. Burgermeister, W . , Hekman, M., and Helmreich, E. J. M. (1982) incubation buffer (Fig. 6 E ) ,whereas labeling of the receptor J. Biol. Chem. 257:5306-5311 by ['251]IABPwas not affected by magnesium (Fig. 70). This 9. Baker, S. P., Liptak, A., and Pitha,J. (1985)J. Biol. Chem. 260, 15820-15828 is consistent with data from previous studies showing that formation of the high affinity agonist state is dependent on 10. Milecki, J., Baker, S. P., Standifer, K. M., Ishizu, T., Chida, Y., Kusiak, J. W., and Pitha, J. (1987)J. Med. Chem. 36, 1563magnesium (30,31). Theremoval of M$+ from the incubation 1566 buffer resulted in the labeling of a 45-kDa protein by [12'1] 11. Rashidbaigi A., and Ruoho, A. E. (1982)J. Pharmmol. Sci. 3, M) NAIN whichwas notprotectable by (-)-alprenolo1 305-307 (data not shown). This nonspecifically labeled protein was 12. Engel, G., Hoyer, D., Berthold, R., and Wagner, H.(1981)Nuunyn-Schmiedenberg's Arch. Pharmacol. 317,277 most likely the heavily stained Coomassie band which migrated at this position (Fig. 6, G and H ) . Finally, N-ethyl- 13. Borch, R. F., Bernstein, M. D., and Durst, H. D. (1971)J. Am. Chem. Soc. 93,2897-2903 maleimide (NEM) hasbeen shown to restore mostof the high 14. Doyle, V.M., Buhler, F. R., and Burgisser, E. (1984)Eur. J. affinity binding sites lost in the absence of magnesium and to Pharmmol. 99,353-356 prevent the shift in agonist affinity seen in the presence of 15. Salomon, Y.,Londos, C., and Rodbell, M. (1974)Anal. Biochern. 58,541-548 nonhydrolyzable nucleotides (31, 32). Thisobservationis 16. Ruoho, A. E., Rashidbaigi A., and Roeder, P. E. (1984)Memthought to be due to alkylation of sulfhydryl groups which branes, Detergents, and Receptor Solubilization 1, 119-169 become exposed as a result of ternary complex formation. 17. Laemmli, U. K. (1970)Nature 227,680-685 Once the sulfhydryls have been alkylated, the agonist locked is18. Jacobson, K. A., Marr-Leisy, D., Rosenkranz, R. P., Verlander, onto the receptor. Photolabeling of the receptorwas not lost M. S., Melmon, K. L., and Goodman, M. (1983)J. Med. Chem. 26,492-499 M) and when membranes were incubated with NEM ['*'I]NAIN for 30 min prior to the addition of GTP+ (lod4 19. Wadzinski, B. E., Shanahan, M. F., and Ruoho, A. E. (1987)J. Biol. Chem. 262, 17683-17689 M) (Fig. 6D).In a similarmanner,membranes could be 20. Lowndes, J. M., Millan, N. M., Ruoho, A. E., and Hokin-Neavphotolabeled by [12'I]NAIN in the absence of magnesium if erson (1987)A.S.B.C. 78,339 (Abstr.) M) wasincluded intheincubationmixture (Fig. 21. Lefkowitz, R. J., Mullikin, D., and Caron, M.G. (1976)J. Biol. NEM Chem. 251,4686-4692 6 F ) . These results illustrate that NAIN photolabels recepthe 22. Kent, R. S., De Lean, A., and Lefkowitz,R. J. (1980)Mol. tor while it iscoupled to G.. Pharmacol. 17, 14-23 Conclusion-In this paper, we have reported the synthesis 23. Limbird, L. E.,and Lefkowitz, R. J. (1978)Proc. Nutl. A c d . Sci. and characterization of NAIN, a catecholamine derivative U. S. A . 75,228-232 containing an azidoiodophenyl moiety. NAIN was found to 24. Limbird, L. E., and Lefkowitz, R. J. (1980)Proc. Nutl. Acud. Sci. U.S. A . 77,775-779 stimulate adenylate cyclase activity in a manner similar to

A Catecholamine Photoaffinity Label

14416

25. Brandt, D. R., Asano, T., Pederson, S. E., and Ross, E. M. (1983) Biochemistry 22,4357-4382 26. Stadel, J. M., De Lean, A., and Lefkowitz, R. J. (1980) J. Biol. Chem. 2 5 5 , 1436-1441 27. Cerione, R. A., Sibley, D. R., Codina, J., Benovic, J. L., Winslow, J., Neer, E. J., Birnbaumer, L., Caron, M. G., and Lefkowitz, R. J. (1984) J. Biol. Chem. 259, 9979-9982 28. Munson. P. J.. and Rodbard., D. (1980) Anal. Biochem. 107.220. 239 I

29. Cheng, Y., and Prusoff, W. H. (1973) Biochem. Phurmacol. 2 2 , 3099-3108 30. Bird, S. J., and Maguire, M. E.(1978) J. Biol. Chem. 253,88268834 31. Heidenreich, K. A., Weiland, G. A., and Molinoff, P. B. (1982) J . Biol. Chem. 257,804-810 32. Korner, M., Gilon, C., and Schramm, M. (1982) J. Biol. Chem. 257,3389-3396