adipocytes and lung, .... without collagenase, and the adipocytes allowed to float to the surface. ..... order to test for this artifact, we replaced SANPAH in brain.
THEJOURNALOF BIOLOGICAL CHEMISTRY 0 1985 by The American Society of Biological Chemists, Inc.
Vol. 260,No. 19, Issue of September 5 , pp. 10806-10811,1985 Printed in U.S.A.
The AIAdenosine Receptor IDENTIFICATIONOF
THE BINDINGSUBUNIT BY PHOTOAFFINITY CROSS-LINKING* (Received for publication, March 29,
1985)
Gary L. Stiles$ From the Departmentof Medicine, Duke UniversityMedical Center, Durham, NorthCarolina 27710
Daniel T. Daly and Ray A. Olsson From the Departmentof Medicine, University of South Florida, Tampa, Florida 33612
Adenosine modifies the catalytic activity of adenyl- proaches to the elucidation of intrinsic membrane protein ate cyclase through both inhibitory (A1 or Ri) as well structure have evolved, namely, biochemical purification and as stimulatory (A, or R,) cell surface receptors. We photoaffinity labeling. Photoaffinitytechniques using [3H] developed 12”I-labeledNs-2-(4-aminophenyl)ethyla- nitrobenzylthioinosine have demonstrated that the nucleoside denosine as a selective ligand to probe the structure of (adenosine) transporter apparatusincludes a protein of M , E Al receptors. The binding of this radioligand to rat 47,000-66,000 intissuessuchasrat adipocytes and lung, cerebral cortex or adipocyte membranes is saturable, human erythrocytes, andguinea pig brain (19-22).This sugreversible, and of high affinity (KO 2 nM). A1receptor gests that the transporter is similar across bothdifferent agonists antagonize binding stereoselectivity and with tissuesand species. Thusfar,neithertechniquehas been a potency order appropriate for A1 receptors. The hetapplied to adenosine receptors. erobifunctional cross-linking reagent N-succinimidylTo probe the structure of the A, adenosine receptor, we 6-(4-azido-2-nitrophenylamino)hexanoatecovalently have developed the A, subtype selective probe ‘251-labeledN 6 couples the radioligand to a protein of M, = 38,000 in both tissues as demonstrated by sodium dodecyl sul- 2-(4-aminophenyl)ethyladenosine(lZ5I-APNEA’).This radiofate-polyacrylamide gel electrophoresis and autora- labeled ligand demonstrates reversible, saturable, stereospediography. Inhibition of covalent labeling by adenosine cific, and selective binding to the AI receptor in a variety of as analogs exhibited the stereoselectivity and potency or- tissues. Using heterobifunctional cross-linking agents such der typical of Al receptor ligands. Guanine nucleotides N-succinimidyl-6-(4’-azido-2’-nitrophenylamino)hexanoate, reduced both specific binding and covalent incorpora- we have covalently labeled the A, receptor binding subunit in adipocyte membranes and detected an tion of the radioligand, evidence that the radioligand rat cerebral cortex and is an A1 receptor agonist. These results suggest that M,’ = 38,000 protein in both tissues by SDS-PAGE/autorathe A1 receptor binding subunit of both brain and adi- diography. Incorporation of the radioligand could be blocked pocytes resides on a protein of M, = 38,000. The new by adenosine analogs. The inhibition of lZ5I-APNEAbinding radioligand should prove useful in studying the struc- by such analogs exhibiteda potency order appropriatefor an ture and regulation of A1 receptors. A, receptor. This technique provides new insights into the structure of adenosine receptors.
-
EXPERIMENTALPROCEDURES
Adenosine acting through specific membrane-bound recepor inhibittors is capableof either stimulating (Ap receptors) ing (A, receptors) the enzyme adenylate cyclase in a wide variety of cells (1-9).The interactionof adenosine with these receptors inducesa wide range of physiologic effects including vasodilation, suppression of ionotropy and chronotropy, inhibition of lipolysis, and sedative effects in the brain (5-11). The molecular mechanisms responsible for these effects are just now beginning to be elucidated but no insight has thus far been gained into the glycoprotein nature of either adenosine receptor subtype (12-15). The elucidation of the structure of hormone receptors has become an important tool in unraveling the mechanisms of drug action at the molecular level (16-18).Two major ap-
* This work was supported by a Clinical Investigator Award National Institutes of Health Grants HL-01027 (to G. L. S.) and HL30391 (to R. A. 0.).The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate thisfact. $ T o whom reprintrequestsshould be addressedat: Box 3444, Department of Medicine, Duke University Medical Center, Durham, NC 27710.
Materials
(-)-N6-(3-[’2SI]Iodo-4-hydroxyphenylisopropyl)adenosine([”51] HPIA) specific activity 2000 Ci/mmol and NalZ5I(carrier-free) were from Amersham Corp. N6-R-phenyl-2-propyladenosine(R-PIA), N6S-1-phenyl-2-propyladenosine (S-PIA), N6-cyclohexyladenosine (N6CHA), adenosine deaminase, and Gpp(NH)p were from Boehringer Mannheim. 3-Isobutyl-1-methylxanthinewas from CalbiochemBehring. N-Succinimidyl 6-(4’-azido-2’-nitrophenylamino)hexanoate (SANPAH) and N-5-azido-2-nitrobenzoyloxysuccinimidewere from Pierce Chemical Co. Phenylmethanesulfonyl fluoride (PMSF), leupeptin, soybean trypsin inhibitor, pepstatin, and bovine serum albumin were from Sigma. Electrophoresis reagents were from Bio-Rad. X-ray film (XAR-5)and developing solutions were from Kodak. Intensifyingscreens(CronexLightingPlus) were from DuPont. Crude collagenase was fromWorthington. Male Sprague-Dawley rats (150-250 g) were from Charles River. One of us (R. A. 0.)synthesized
’ Nonstandard abbreviations and trivial namesused are: APNEA, N6-2-(4-aminophenyl)ethyladenosine; Gpp(NH)p, guanyl-5-yl (/3,rimido)diphosphate; R-PIA, N6-R-1-phenyl-2-propyladenosine; SPIA, N6-S-l-phenyl-2-propyladenosine; N6-CHA, N6-cyclohexyladenosine; NECA, N-ethyladenosine 5”uronamide; SANPAH, N-succinimidyl6-(4’-azido-2’-nitrophenylamino)hexanoate; PMSF, phenylmethanesulfonyl fluoride; SDS,sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis.
10806
Adenosine AI Receptor NECA having chemical characteristics identical to those reported (23). Membrane Preparation Rat Cerebral Cortex-Fresh cerebral cortex was dissected free of remaining brain tissue and placed in ice-cold 50 mM Tris-HC1, pH 7.4, 25 "C, 5 mM EDTA with PMSF M), soybean trypsin inhibitor (5 pg/ml), leupeptin (5 pg/ml), and pepstatin (2 pglml). The cortex was then minced and homogenized with 10 strokes with a motor-driven glass-Teflon homogenizer onice. The homogenate was then passed through 4 layers of cheesecloth and subjected to centrifugation at 600 X g for 10 min. The supernatant was retained and centrifuged at 40,000 X g for 10 min. The resultant pellet was washed once in the above buffer and recentrifuged at 40,000 X g for 10 min. The pellet was then suspended in 4 mlof 50 mM Tris-HC1, 10 mM MgCIZ,1mM EDTA to which wasadded adenosine deaminase, 0.3 unit/ml, and the mixture incubated at 30 "C for 15 min. The membranes were either used immediately or frozen in liquid N, and stored at -80 'C for use within 2 weeks. A, receptor binding decreased after 2 weeks of storage. AdipocyteMembranes-Rats were killed, the epididymidal fat pads removed and placed in Krebs phosphate buffer containing 128 mM NaCI, 1.4mM MgCI,, 5.2 mM KCl, 10 mM Na2HP04,pH 7.4, to which was added 3% (w/v) bovine albumin, 1 mg/ml crude collagenase, leupeptin (5 pglml), and soybean trypsin inhibitor (5 pg/pl). The fat was minced with scissors and incubated at 37 "C for 45 min with constant agitation. The disaggregated adipocytes were passed through 4 layers of cheesecloth. The filtrate was diluted with the above buffer without collagenase, and theadipocytes allowed to float to thesurface. The cells were aspiratedand washed once with buffer at room temperature and the cells floating on the surface were harvested in 10volumes of ice-cold 10 mM NaZHPO,, 5 mM EDTA, pH 7.4, to which wasadded 5 pg/ml leupeptin, 5 pg/ml soybean trypsin inhibitor, and lo-' M PMSF. The cells were then disrupted with 10 strokes of a glass-Teflon homogenizer. The homogenate was centrifuged at 15,000 X g for 15 min a t 4 "C. The fat adherent to the sides of the tube was carefully removed and the membranes resuspended in the above buffer and resedimented. The pellet was then suspended in 50 mM Tris-HC1, 10 mM MgCI,, 1 mM EDTA, pH 7.4, and treated with adenosine deaminase at 0.5 unit/ml for 10 min at 30 "C. The membranes were then used immediately. Photoaffinity Cross-linking-Membranes were suspended in 50 mM Tris-HC1, 10 mMMgC12, 1 mM EDTA, pH 8.26 at 7 "C to a receptor concentration of-200 pM in a 5-ml volume. To separate aliquots were added either H,O or competing ligand at the indicated concentration. '"I-APNEA was then added to all aliquots at a concentration of 1.5 nM. The aliquots were then incubated at 37 "C for 45 min, the membranes diluted to 40 ml with ice-cold 50 mM Na,HP04, pH 7.4, and centrifuged at 45,000 X g for 5 min. Thesupernatant was discarded, and themembranes were washed once with the same buffer min. The pellet was then and recentrifuged at 45,000 X gfor5 suspended in 5 ml of the same phosphatebuffer and thiswas brought t o 25 "C. To this was added 30 pl of 5 mM SANPAH dissolved in dimethyl sulfoxide (24). After 5 min, 100 p1 of 1 M glycine wasadded. The suspension was placed in a Petri dish and exposed to UV light from a Hanovia 450 W medium pressure mercury arc lamp for 75 s 12 cm from the lamp, as previously described (25). The membranes were then sedimented by centrifugation and washed twice with 50 mM Tris-HC1, pH 7.4 at 25 "C, 5 mM EDTA, and then solubilized and prepared for SDS-PAGE as described below. Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis-The electrophoresis was performed according to the method of Laemmli (26) using homogeneous slab gels, with the exception that samples were solubilized in 10% SDS, 10% glycerol, 25 mM Tris-HC1, pH 6.8, with 5% P-mercaptoethanol unless otherwise indicated. Samples were denatured for 45 min at room temperature. Following electrophoresis, the gels were dried and exposed to Kodak XAR-5 x-ray film with intensifying screens. Radioligand Binding Data Analyses-Competition curves involving antagonist and agonists were analyzed according to a four-parameter logistic equat.ion to determine EC, values as previously described (27). Saturationcurves were analyzed using a nonlinear least squares curve fitting techniquewith statistical analyses as recently published and validated (28). Protein Determinations-Proteins were determined either by the method of Lowry et al. (29) or by the Amido Schwartz method (30) using bovine serum albumin as standard.
10807
Synthesis and Radioiodination of N6-2-(4-Aminophenyljethyladenosine-The preparation of APNEA followed a general method for the synthesis of N'-substituted adenosines (31). Refluxing a magnetically stirred suspension of 4 g of (14 mmol) 6-chloropurine riboside, 2.72 g of (20 mmol) 2-(4-aminophenyl)ethylamine,5 g of (50 mmol) triethylamine, and 50 ml absolute ethanol for 48 h with exclusion of moisture gave a clear solution. The residue after vacuum evaporation was dissolved in 1 liter of boiling water and filtered. Product crystallizing on cooling wasfiltered, washed with cold water and ether, then dried over P205 at 85 "C. Yield was 4.4 g (82%) of colorless needles, ( 6 ) 270 (18,000). m.p. 174-175 "C. UV, A,, CISHZZN~O~ (386.42) Calculated C 55.94 H 5.74 N 21.75 Found C 55.90 H 5.68 N 21.87 Radioiodination followed the method of Schwabe et al. (32) with modifications. A 10-pl aliquot of a 2.5 mM solution of APNEA in methanol was placed in a 1.5-ml Eppendorf centrifuge tube and dried in a stream of NZ.The nucleoside was dissolved in 40 MI of 0.3 M Na2HP0,, pH 7.55,by repeated pipetting and heating in a 40 "C water bath. The sequential addition of 10-20 pl of carrier-free Na'T (1-2 mCi) and 10 p1 of chloramine-T (1 mg/ml) initiated iodination. After 1 min of continuous vortex mixing the reaction was stopped by the addition of 25 pl of Na2S205(2 mg/ml), and the mixture was extracted with four 200-p1 aliquots of ethyl acetate. The pooled extracts were dried under a streamof N, and theproduct was purified by high performance liquid chromatography (Fig. 1). The '251-APNEApeak, which emerged at 5 min, was completely separated from the more polar starting material, which eluted at 3 min. Product contained 90% of the radioactivity contained in the ethyl acetate extractof the radioiodination mixture. Thin layer chromatography on 5 X 20-cm silica gel plates (Scientific Products, Silica Gel 60 F2s4)developed with chloroform/methanol (3:2, v/v), showed product (RF0.48) uncontaminated by APNEA (RF 0.43). Development with methanol, 50 mM ammonium formate, pH 8 (l:l, v/v), likewise showed iodonucleoside (RF0.62) free of starting material (RF0.68). The product can be stored in the same methanol/ammonium formate buffer and used for binding or photoaffinity cross-linking as described. Since the product is completely separated from starting material and is a single compound on TLC, a specific activity of 2175 Ci/mmol was assumed.
- .08-
-
n
I
II
E .06-
II 11
-6
I I
z
a
U
m
-8
1
I
.04
1 1 W I I
-
f l
L
a1
- 4
I
I I
v)
I
I
a
I
I
m
-2 " "
1
4
I3
-
z -
2
.
1
.
1
.
1
.
4 6 8 1 T I M E (min)
1
0
FIG. 1. High performance liquid chromatographic separation of APNEA from '"I-APNEA. The radioiodination reaction products were taken up in 50% methanol, 50% 50 mM ammonium formate, pH 8.0, and separated on a CIS pBondapak reverse phase column a t 1.2 ml/min. The mobile phase was 50% methanol, 50% 50 mM ammonium formate, pH 8.0. Eluate was monitored by a Waters UV monitor at 254 nm. Samples (5 pl) from each 0.6-ml aliquot were counted in a Packard y-counter. The UV absorbance peak labeled APNEA coincides with the migration of starting material.
Adenosine AI Receptor
10808 RESULTS
Characterization of Binding of lZ5I-APNEAto A, Adenosine R e c e p t o r ~ ~ ' ~ ~ 1 - A P demonstrates NEA reversible high affinity binding to A, receptors in rat cerebral cortex membranes. The association of lZ5I-APNEAmeasured at a radioligand concentration of 1 nM reached steady state at approximately 40 min at 37 "C and remained stable for at least 90 min (Fig. 2 A ) . The ligand demonstrates rapid reversibility in the presence of IOW6 M R-PIA at 37 "C; by 20 min the extentof dissociation is almost 95% (Fig. 2B). Saturation binding isotherms dem-
TIME (min)
TIME (min)
FIG. 2. Association ( A ) and dissociation (B)kinetics of "'IAPNEA in rat cerebral cortex membranes. In A , rat cerebral cortex membranes were incubated with 1 nM lz5I-APNEAat 37 "C and aliquots were withdrawn at the times shown and filtered as described under "Experimental Procedures." Binding shown is speM R-PIA. In B, cerebral cortex cific binding as determined with membranes were incubated with 1 nM radioligand for 45 min at 37 "C M ) was added tothe incubation mixture. at which time R-PIA Aliquots were then withdrawn at theindicated time after the addition of R-PIA and filtered as described. Each data point is the mean of duplicate samples. This experiment is representative of two similar experiments.
04
5' n Z
-""
.os /
NONSPECIFIC BINDING L
2.0
['251]
3.0
4.0
5.0
APNEA ( n M )
FIG. 3. '"I-APNEA saturation curves in the presence and absence of Gpp(NH)p (lo-" M) in rat cerebral cortex membranes. Membranes were prepared as described under "Experimental Procedures." Both lZ5I-APNEAand [lZSI]HPIAradioligand binding were performed in a similar manner. Assayvolumewas 0.25 ml consisting of 150 pl of membranes (-100 pg/ml protein) in 50 mM Tris-HC1, pH 8.26, at 7 "C, 10 mMMgClZ, 1 mM EDTA, 50pl of radioligand, and 50 pl of water or competing ligand. The incubation was at 37 "C for 1 h for ['251]HPIA and 45 min for '=I-APNEA. Binding was terminated by vacuum filtration over Whatman G/FC glass fiber filters and with three washes of 5 ml each with the above buffer and lZ6Iactivity detected in a Packard y-counter at an efficiency of 75%. lZ5I-APNEAwas added at a final concentration shown on the abscissa. The concentration of '=I-APNEA bound is on the ordinate. Nonspecific binding was defined with M unlabeled RPIA. The data points are means of duplicate determinations. This experiment is representative of three similar experiments. The curves were drawn with the aid of a computer modeling program based on the law of mass action (27).
FIG. 4. Competition of adenosine analogs with '"I-APNEA in cerebral cortex membranes. Radioligand binding was performed as described in the legend to Fig. 3 with the indicated concentrations of competitors. The lines through the experimentally derived points were drawn with computer assisted analysis based on a four parameter logistic equation (26). Radioligand was present at a concentration of 0.4 nM and binding in the absence of competition was 40 PM. IBMX, isobutylmethylxanthine. Each curve was replicated 35 times and mean data points varied less than 10%.
onstrate saturable, high affinity binding (Fig. 3). The dissociation constant (Ku) averaged 2.0 f 0.3 nM and Bmax720 f 87 fmol/mg. Many N6 substituted analogs of adenosine have been demonstrated to have agonist properties at theA1 receptor, and since the A1 receptor-adenylate cyclase system is known to be modulated by guanine nucleotides (33-37), we assessed the effect of Gpp(NH)p on lZ5I-APNEAbinding. In a series of three,Gpp(NH)p M ) was demonstrated to decrease B,,, by 76 f 6% ( p < 0.001) while not significantly affecting the Ku (1.6 f 0.6 nM). This result is similar to that previously reported for ['251]HPIA binding in rat cerebral cortex membranes (15). To determine if the number of At receptors quantitated by "9-APNEA in cerebral cortex membranes was similar to that observed with the previously characterized [ '251]HPIAbinding (15), we measured receptor concentrations in the same membrane preparations and found a Bmaxof 756 k 110 fmol/mg with [1251]HPIA,a value not different than thatfound for lZ5IAPNEA. Thus, ' 9 - A P N E A and ['251]HPIA recognize the same quantity of A1 adenosine receptors. Binding of lz5I-APNEAwas specific for A1 adenosine receptors. Thus, competition curves (Fig. 4) constructed using lZ5IAPNEA demonstrated the appropriate order of ligand potency expected of an A1 receptor subtype: N6-CHA (EC, = 0.3 k 0.1 nM) 2 R-PIA (ECso = 0.55 f 0.14 nM) > NECA (ECm = 4.1 k 0.1 nM) > S-PIA (EC50 = 55.9 f 9.3 nM) >> 3-isobutyl1-methylxanthine (ECEo= 16.4 f 3.8 WM). In addition, the nonpurine ligands, isoproterenol @-adrenergic), epinephrine (a-and &adrenergic), dopamine (Dl and Dz receptors), and carbachol (muscarinic receptors) at concentrations up to M failed to compete for occupancy with lz5I-APNEAat the A1 receptor (data not shown). These results demonstrate that this ligand binds with all the characteristics expected of a ligand interacting specifically with the A1 adenosine receptor.
Photoaffinity Cross-linking of lZ5I-APNEA into theA, Receptor Binding Subunit-Although the pharmacologic characterization of this radioligand was an important part of this
Adenosine A , Receptor
10809
E3
Mr
Mr
Mr
66K45K45K36K29K-
36K29K-
24K24 K-
20K-
/ /
/
/
FIG. 5 (left and center). Photoaffinity cross-linking of 12'I-APNEA into rat cerebral cortex membranes. A, cerebral cortex membranes were photoaffinity cross-linked with lZ5I-APNEAalone (control) or in the presence of the indicated concentration of competing ligand as described under "Experimental Procedures." The samples were then solubilized and electrophoresed on a 10% acrylamide gel. The relative molecular weight (M,) scale was calibrated using iodinated protein standards (Sigma): bovine albumin (M, = 66,000), ovalbumin (45,000), glyceraldehyde-3-phosphate dehydrogenase (36,000); carbonic anhydrase (29,000); trypsinogen (PMSF-treated) (24,000);and soybean trypsin inhibitor (20,100). The results shown are representative of four similar experiments. IBMX, isobutylmethylxanthine. B, the photoaffinity cross-linking was performed as described in A. FIG.6 (right). Photoaffinity cross-linking of 12"I-APNEA in rat adipocyte membranes. Adipocyte membranes were prepared as described under "Experimental Procedures." Membranes were photoaffinity crossM as described linked with lZ5I-APNEAand SANPAHeither alone (control) or in the presence of R-PIA under "Experimental Procedures." The samples were then solubilized and electrophoresed on a 10% acrylamide gel. M , calibration is as described in the legend to Fig. 5A.
study, the major goal was the elucidation of the structure of the binding subunit of the AI receptor. To estimate the size of the ligand binding subunit, we studied the two most completely characterized membrane systems containing A, receptors, those of rat cerebral cortex and adipocytes. Fig. 5A demonstrates the results obtainedwhen the A, receptor of rat cerebral cortex membranes was cross-linked with lZ5I-APNEA using the photoaffinity cross-linking agent SANPAH followed by SDS-PAGE. Asingle major band of radioactivity is apparent with M , = 38,000 f 1,500 (Fig.5A, control). The specificity of labeling is documented by the fact that incorporation of '251-APNEAis completely blocked by 10"' M R-PIA, and that an equivalent concentration of NECA is much less effective in blocking incorporation. In addition, incorporation is stereoselective as evidenced by the fact that at lo-' M, S-PIA is ineffective in blocking incorporation compared to R-PIA. These results are in good agreement with the findings by radioligand binding (see below). Furthermore, Gpp(NH)p at very low concentrations (lo-' M) decreases the amount of covalent incorporation. Fig. 5B demonstrates that N6-CHA at M completely blocks incorporation; M 3-isobutyl-
1-methylxanthine significantly blocks incorporation; and Gpp(NH)p M) also significantly decreases labeling; all consistent withthe results of the radioligand binding studies. Several controls strengthenthe evidence that thisM , = 38,000 peptide is indeed the binding subunit. Since photoaffinity cross-linking is an inefficient process and a covalent labeling efficiency of 0.1-2% is to be expected (24, 38), we needed to document the efficiency of ligand incorporation into the receptor protein.By determining the amount of specific binding in solubilized aliquots of membranes that had been labeled with '251-APNEA alone or inthe presence of M R-PIA and thencomparing that with the amount of specific binding determined by counting SDS-PAGE gel slices, we demonstrated anefficiency of labeling of -1.2% in this system. This is in good agreement with the expected results. In addition, it was demonstrated that either the failure to include SANPAH in the reaction mixture or failure to expose the membranes to ultraviolet light leads to no covalent incorporation (data not shown). The degradation of receptors by endogenous proteases is a major problem complicating the biochemical characterization
10810
Adenosine A I Receptor
of membrane-bound receptors. Failure to inhibit such probound toreceptorsultimately undergoescovalentlinkage. teases may allow the generationof multiple peptide fragments Accordingly, it is important to demonstrate that proteins so which retainthe pharmacological attributes of thenative labeled are representative of the general receptor population. receptors (25, 39, 40). To assess if this was occurring in this A series of A, receptor ligands showedthe samepotency order system, we have used fresh uersus frozen membranes, peras inhibitorsof photoaffinity cross-linking as they hadshown formedlabeling at 25 "C rather than 37 "C, included the as inhibitors of the reversible binding of Iz5I-APNEA. Likeproteaseinhibitors,EDTA,leupeptin,pepstatin, soybean wise, Gpp(NH)p reduced the covalent incorporation of lZ5Itrypsininhibitor,PMSF,benzamidine,dithiothreitol,and APNEA. Furthermore, thespecies of peptide labeled by phoiodoacetate and could demonstrate no change in the size of toaffinity cross-linking appears to be independent of the type the radiolabeled peptide. Such results suggest that the M , = of cross-linking reagent. 38,000 peptide is not a degradation product of a larger native Themembraneprotein labeled by Iz5I-APNEA appears protein. distinct from the purinenucleoside transporter, another cell To determine whether the Mr,38,000 peptide is unique for membraneprotein which is capable of reversibly binding the A, receptor in brain cortex membranes, we also photoaf- adenosine. Irradiating plasma membranes in thepresence of finitycross-linkedthe AI receptorinrat adipocyte mem- the nucleoside transport inhibitor [3H]6-S-(4-nitrobenbranes. Fig. 6 demonstrates that exactly the same molecular zy1)thioinosine labels a peptide whose M , is variously estiweight protein islabeled in theadipocyte membrane AI recep- mated at between 47,000 and 66,000 (19-22). Nucleoside tor system as in brain cortex. Thus, in two separate tissues, transportinhibitors specifically block the labeling of this proteins of identical size are labeled which demonstrate the protein, but adenosinereceptor agonists such as N6-CHA and characteristics expected of the A, adenosine receptor. other N6-alkyladenosinesdo not. Suchdifferences in size and "Ex0 labeling," the covalent linking of a radioligand to pharmacological specificity suggest that the protein labeled proteinsadjacenttothereceptorprotein(38), is another by lz5I-APNEAis not the nucleoside transporter. potential artifact of photoaffinity cross-linking experiments. Previous work in a number of receptor systems has demThis liability is more acute in the case of cross-linking agents onstrated the importance of inhibiting endogenous proteinin which long spacer arms separate the reactive groups. In ases during membrane preparation to prevent degradation of order to testfor this artifact, we replaced SANPAH in brain receptor proteins (25,39,40). In this study, we have tested all labeling experimentswith N-5-azido-2-nitrobenzoyloxysuc- the proteinase inhibitorspreviously documented tobe imporhavevariedlabeling conditions cinimide, a cross-linking reagent shorter by six methylene tant in other systems and groups. Although the substitution of N-5-azido-2-nitroben- (temperature, fresh and frozen membranes) and can demonstrate no evidence that this M , = 38,000 protein is a degrazoyloxysuccinimide reduced radioligand incorporation to
