Cholecystokinin induces the interaction of its receptor with a guanine nucleotide binding protein

To evaluate the relation between the pancreatic cholecystokinin (CCK) receptor and guanine nucleotide-binding protein(s) we studied the effects of nucleotides on 125I-CCK binding to pancreatic acinar plasma membranes, 125I-CCK binding to solubilized 125I-CCK receptors, and the stability of the solubilized 125I-CCK-receptor complex. In plasma membranes, guanine nucleotides both inhibited CCK binding and increased the dissociation of CCK from its receptor. The potency of the nucleotides studied was GTP gamma S = GMP-PNP greater than GTP much greater than ATP. When membranes were solubilized with digitonin, subsequent binding of CCK was insensitive to guanine nucleotides including GTP, GMP-PNP and GTP gamma S. However, if CCK binding occurred before solubilization of the membranes, guanine nucleotides increased dissociation at concentrations and with a specificity similar to that observed for effects on intact pancreatic membranes. It is concluded that guanine nucleotides act via a protein which is separable from the receptor to induce dissociation of bound CCK. Moreover, CCK binding induces an association in the plasma membrane of the CCK receptor with this guanine nucleotide binding protein.     

Characterization of cholecystokinin receptors in bullfrog (Rana catesbeiana) brain and pancreas

The binding of biologically active 125I-Bolton-Hunter-CCK-33 to bullfrog brain and pancreatic membrane particles was characterized. Both tissues exhibited time-dependent, saturable, reversible, and high affinity binding without evidence for cooperative interaction. Both bullfrog CCK receptors resembled their mammalian counterparts in having acidic pH optima for tracer binding and a Kd of about 0.5 nM. However, the receptors differed from their mammalian counterparts in that (1) the bullfrog brain membranes bound more tracer per mg protein than did the pancreatic membranes, (2) both bullfrog CCK receptors were relatively insensitive to dibutyryl cGMP, and (3) both bullfrog brain and pancreatic CCK receptors exhibited the same general specificity toward a variety of CCK and gastrin peptides. For both tissues, the relative order of receptor binding potency was CCK-8 greater than caerulein = CCK-33 greater than gastrin-17-II greater than CCK-8-ns = gastrin-17-I greater than caerulein-ns greater than gastrin-4 with the sulfated CCK peptides being 1000-fold more potent than their nonsulfated analogs. Sulfated gastrin was also relatively potent, being only 10-fold weaker than CCK-8. Gastrin-4 was 20 000-fold weaker than CCK-8 in interacting with the brain CCK receptor. The latter finding is in sharp contrast to the mammalian brain CCK receptor. We conclude that the bullfrog brain and pancreas contain similar CCK receptors of probable physiological significance and may represent an ancestral condition from which the two distinct CCK receptors present in mammalian brain and pancreas have evolved.     

The CCK receptor on pancreatic plasma membranes: binding characteristics and covalent cross-linking

The cholecystokinin (CCK) receptor in purified plasma membranes prepared from mouse pancreatic acini had a binding affinity of 1.8 nM, an acid pH optimum between 6.0 and 6.5, and an analog specificity of CCK8 greater than CCK33 greater than desulphated CCK8 greater than CCK4. Binding of CCK to its receptor was abolished by pretreatment of plasma membranes with trypsin. When [125I]CCK was cross-linked to its receptors with disuccinimidyl suberate, and the preparation solubilized and subjected to gel electrophoresis and autoradiography, the hormone was associated with Mr 80 000 protein in both the presence and absence of the reducing agent dithiothreitol.     

VIP receptors from porcine liver: high yield solubilization in a GTP-insensitive form.

Vasoactive intestinal peptide (VIP) receptors were solubilized from porcine liver membranes using CHAPS. The binding of 125I-VIP to solubilized receptors was reversible, saturable and specific. Scatchard analysis indicated the presence of one binding site with a Kd of 6.5 +/- 0.3 nM and a Bmax of 1.20 +/- 0.15 pmol/mg protein. Solubilized and membrane-bound receptors displayed the same pharmacological profile since VIP and VIP-related peptides inhibited 125I-VIP binding to both receptor preparations with the same rank order of potency e.g. VIP greater than helodermin greater than rat GRF greater than rat PHI greater than secretin greater than human GRF. GTP inhibited 125I-VIP binding to membrane-bound receptors but not to solubilized receptors supporting functional uncoupling of VIP receptor and G protein during solubilization. Affinity labeling of solubilized and membrane-bound VIP receptors with 125I-VIP revealed the presence of a single molecular component with Mr 55,000 in both cases. It is concluded that VIP receptors from porcine liver can be solubilized with a good yield, in a GTP-insentive, G protein-free form. This represents a major advance towards the purification of VIP receptors.     

Brain CCK receptors: species differences in regional distribution and selectivity

The binding of 125I-CCK-33 to its receptors prepared from cerebral cortex and cerebellum was studied in four species: mouse, rat, hamster, and guinea pig. Only the guinea pig showed significant binding to membranes from cerebellum and this binding was comparable to that observed for cerebral cortex. In all four species, the order of potency of unlabeled analogs to compete for the binding site was CCK-8 greater than CCK-33 greater than desulfated CCK-8 greater than CCK-4. While the affinity for CCK-8 and CCK-33 was similar in the various species, the relative affinity for desulfated CCK-8 and CCK-4 was less for hamster and guinea pig, indicating species differences in receptor specificity, as well as in regional localization.     

Characterization of cholecystokinin receptors in toad retina

The binding characteristics, structure, and pharmacologic properties of a cholecystokinin binding protein in toad retinal membranes have been studied. In competition binding studies using 125I-CCK-8, toad retinal membranes exhibited a high affinity binding site having a Ki50 of 1.5 nM using CCK-8 as competitive ligand. The relative potencies of CCK-related peptides in inhibiting radioligand binding were caerulein greater than gastrin II approximately equal to CCK-8 approximately equal to CCK-33 greater than CCK-8-DS approximately equal to gastrin I. L-364,718, a potent inhibitor of peripheral CCK receptors, was ineffective at competition binding at concentrations up to 1 microM; dibutyryl cyclic GMP was modestly effective at competing (KD approximately 10 mM). Covalent binding of 125I-CCK-33 to toad retinal membranes using chemical cross-linkers or UV irradiation resulted in the labeling of a major Mr 62,000 protein and the intermittent labeling of minor components of Mr 105,000 and Mr 40,000 as determined by SDS-PAGE and autoradiography. The binding of 125I-CCK-33 to retinal membranes and the concomitant labeling of the Mr 62,000 component was specifically inhibited by CCK-8 (KD approximately 1.5 nM). Reduction of membranes with DTT abolished specific binding of 125I-CCK. SDS-PAGE analysis of affinity cross-linked membranes under non-reducing conditions revealed that the Mr 62,000 protein migrated with an apparently lower molecular weight. These results suggest that the Mr 62,000 CCK binding protein in the toad retina contains an intramolecular disulfide bond(s). The Mr 62,000 protein was retained on a wheat germ agglutinin-agarose column and eluted with N-acetyl D-glucosamine, suggesting the glycoprotein nature of this protein. Digestion of the Mr 62,000 protein with neuraminidase together with O-glycanase resulted in a discrete product of Mr approximately 60,000. These results indicate that the Mr 62,000 protein is a glycoprotein with O-linked oligosaccharide chains. Taken together, these data indicate that the CCK receptor in toad retina has a distinct structure compared to that described in rat pancreas or brain. It will be important to establish whether this difference is reflected in differences in signal transduction mechanisms.     

Cholecystokinin-stimulated pepsinogen secretion and cholecystokinin receptors on gastric chief cells in guinea pigs

We investigated cholecystokinin (CCK) receptors on isolated gastric chief cells from guinea pig. CCK stimulated pepsinogen secretion from chief cells at the same efficacy as that induced by carbamylcholine. Binding of 125I-labeled CCK-33 (125I-CCK) to chief cells was temperature-dependent, and was saturable and reversible at 37 degrees C. Hofstee plots of the ability of CCK-8 to inhibit binding of 125I-CCK showed a linear regression line, suggesting that CCK receptors possessed one binding site. The dissociation constant of the binding site was calculated to be 3.8 x 10(-10) M. The dose-response curve of CCK for pepsinogen secretion was superimposed on that for the binding to its receptors. These results indicated that gastric chief cells from the guinea pig possess CCK receptors that relate closely to the action of CCK involved in pepsinogen secretion.     

Importance of sulfation of gastrin or cholecystokinin (CCK) on affinity for gastrin and CCK receptors

We investigated the importance of sulfation of gastrin or cholecystokinin (CCK) on influencing their affinity for gastrin or CCK receptors by comparing the abilities of sulfated gastrin-17 (gastrin-17-II), desulfated gastrin-17 (gastrin-17-I), CCK-8 and desulfated CCK-8 [des(SO3)CCK-8] to interact with CCK or gastrin receptors on guinea pig pancreatic acini. For inhibiting binding of 125I-gastrin to gastrin receptors, gastrin-17-II (Kd 0.08 nM) greater than CCK-8 (Kd 0.4 nM) greater than gastrin-17-I (Kd 1.5 nM) greater than des(SO3)CCK-8 (Kd 28 nM). For inhibiting binding of 125I-Bolton Hunter-labeled CCK-8 to CCK receptors the relative potencies were: CCK-8 much greater than des(SO3)CCK-8 = gastrin-17-II greater than gastrin-17-I. Each peptide interacted with both high and low affinity CCK binding sites. The relative abilities of each peptide to interact with high affinity CCK receptors showed a close correlation with their abilities to cause half-maximal stimulation of enzyme secretion. These results demonstrate that, in contrast to older studies, sulfation of both CCK and gastrin increase their affinities for both gastrin and CCK receptors. Moreover, the gastrin receptor is relatively insensitive to the position of the sulfate moiety, whereas the CCK receptor is extremely sensitive to both the presence and exact position of the sulfate moiety.     

Purification of the pancreatic cholecystokinin receptor

We have previously shown that the pancreatic cholecystokinin (CCK) receptor can be solubilized in 1% digitonin. In this study, digitonin-solubilized CCK receptors from rat pancreas were purified using sequential affinity chromatography on ricin-II agarose and on AffiGel-CCK. Electrophoresis of the radioiodinated purified receptors on SDS-polyacrylamide gels followed by autoradiography revealed two proteins: a major band of Mr = 80,000-90,000, and a minor band of Mr = 55,000. Through the purification procedure, the receptors preserved their agonist specificity (CCK-8 less than CCK-33 less than desulfated CCK-8 less than CCK-4) and binding affinity. Scatchard transformations of binding data for the purified receptor preparation were best fit by linear plots compatible with a single class of binding sites with Kd = 9.4 nM. The estimated purification was about 80,000 fold and consistent with the expected Bmax for a pure Mr = 80,000 protein binding one CCK molecule. This two-step purification procedure opens the possibility for molecular studies of the CCK receptor.     

Characterization of cholecystokinin receptor subunits on pancreatic plasma membranes

To determine the size and subunit structure of the pancreatic cholecystokinin (CCK) receptor, 125I-CCK33 was covalently cross-linked to its receptor on mouse pancreatic acinar plasma membranes utilizing the bifunctional cross-linker disuccinimidyl suberate. When CCK was cross-linked at pH 7.4 to either purified plasma membranes or to isolated pancreatic acini and then followed by preparation of plasma membranes, the major labeled protein band revealed by polyacrylamide gel electrophoresis was Mr = 120,000 in the absence of reducing agent and Mr = 80,000 in the presence of reducing agent. A similar banding pattern was also observed when different cross-linkers, ethylene glycol bis(succinimidyl succinate) or dithiobis (succinimidyl propionate), were employed. At pH 6.0, where CCK binding to its receptors is optimal, the labeling pattern was similar to that seen at 7.4, although the two bands were more heavily labeled. Both the binding of CCK to its receptors on plasma membranes and the appearance of the two cross-linked proteins on gels were inhibited in a parallel manner by increasing concentrations of unlabeled CCK8; similar results were observed with dibutyryl cyclic GMP, a competitive inhibitor of CCK binding and action. The data indicate, therefore, that the CCK receptor possesses subunit structure whereby an Mr = 76,000 binding subunit is linked to an Mr = 40,000 nonbinding subunit by a disulfide bond.     

Cholecystokinin antagonists

The three classes of CCK antagonists illustrate the various factors governing affinity of the antagonists for the CCK receptor. The major influence in determining potency of the cyclic nucleotide derivatives, amino acid derivatives and C-terminal fragments of CCK, are hydrophobic forces. In contrast, structural requirements are the major influences in determining potency of the N-terminal fragments of CCK-26-33. The most potent CCK antagonist in each of the three classes is illustrated in Fig. 11. CBZ-CCK-27-32-NH2 is 30 times more potent than N-CBZ-cystine, which is, in turn, slightly more potent than Bt2 cGMP. All these CCK antagonists, however, are relatively weak. For example, CBZ-CCK-27-32-NH2 inhibits binding of 125I-CCK by 50% at a concentration of approximately 5 microM. In contrast, the agonist CCK-26-33 inhibits binding of 125I-CCK by 50% at a concentration of approximately 1 nM. The antagonists remain useful for analyzing those responses that are caused by CCK, though the relatively low potencies of the antagonists may limit their usefulness as antagonists of CCK in vivo systems. This limitation, however, may be only theoretical. For example, proglumide, which requires an in vitro concentration of 0.3 mM to cause half-maximal inhibition of binding of 125I-CCK, can inhibit the actions of CCK and gastrin in animals (Hutchison and Dockray 1980; Stubbs and Stabile 1985) and of gastrin in man (Lamers and Jansen 1983). Nevertheless, the identification of CCK antagonists with greater potency than those hitherto described will facilitate studies of the actions of CCK.     

Purification and characterization of the rat pancreatic cholecystokinin receptor

Cholecystokinin (CCK) is a peptide hormone that has a variety of physiologically important functions in the gastrointestinal tract, in which distinct high affinity receptors have been identified. We describe here the purification of the digitonin-solubilized rat pancreatic receptor as an initial step in the determination of its primary structure. Solubilization of total pancreatic membranes using 1% digitonin resulted in a single class of binding sites with a specific content of 4 pmol/mg as measured in a soluble binding assay using the nonpeptidyl CCK antagonist [3H]3S[-]-N-[2,3-dihydro-1-methyl-2-oxo-5-phenyl-1H-1,4- benzodiazepine-3-yl]-1H-indole-2-carboxamide [( 3H]364,718). The solubilized receptor was purified using the following chromatographic steps: 1) cation exchange; 2) Ulex europaeus agglutinin-I-agarose; and 3) Sephacryl S-300. The final preparation of the purified receptor had a specific content of 8,055 pmol/mg, which represented a 9,051-fold purification from intact membranes. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis of the purified receptor preparation under reducing conditions resulted in a predominant polypeptide with an Mr = 85,000-95,000 and minor polypeptides of Mr = 57,000 and 26,000 as determined by radiolabeling and silver staining. Solubilized pancreatic membranes were affinity labeled with the peptidyl CCK agonist 125I-D-Tyr-Gly-[(Nle28,31,6-NO2-Phe33)CCK-26-33] and chromatographed under conditions similar to those described for untreated membranes. Elution of radioactive peaks from each chromatographic column was coincident with [3H]364,718 binding activity and resulted in a labeled polypeptide having the same electrophoretic mobility as receptor derived from freshly labeled membranes and purified from untreated membranes. High performance liquid-gel exclusion chromatography of the crude digitonin-solubilized membrane preparation revealed an estimated molecular size for the [3H]364,718-binding activity of 370,000, which was consistent with the size determined by nondenaturing gel electrophoresis of the purified receptor complexed with the labeled nonpeptidyl antagonist. Binding of [3H]364,718 to the purified receptor preparation was comparable to that observed with the crude solubilized pancreatic membrane preparation; and both the homologous ligand 364,718 (Ki = 0.5 nm) and CCK-8 (Ki = 1.4 microM) competed for binding to both preparations in a similar manner.     

Autoradiographic and functional development of gastric cholecystokinin receptors in the rat

To determine the functional role for and the pharmacological specificity of developing gastrointestinal CCK receptors, in vitro pyloric contractility and autoradiographic CCK receptor binding were examined in pups aged 1–20 days. CCK contracted the gastroduodenal junction at all ages, while nonsulfated CCK-8 (d-CCK) was less potent. Autoradiographic studies revealed CCK binding localized to the gastroduodenal junction throughout development. MK-329, a specific type A CCK receptor antagonist, completely displaced ¹²⁵I CCK-8 binding at all ages, while d-CCK displaced binding at ages at which d-CCK elicited gastroduodenal contractility. The results demonstrate a physiological role for and pharmacological specificity of neonatal gastroduodenal CCK receptors.     

Solubilization of the Neuropeptide Y Receptor from Rat Brain Membranes

The neuropeptide Y (NPY) receptor was solubilized from rat brain membranes with the zwitterionic detergent 3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate (CHAPS). The binding of 125I-NPY to CHAPS extracts was protein, time, and temperature dependent. Unlabeled NPY and the related peptides peptide YY (PYY) and pancreatic polypeptide inhibited 125I-NPY binding to solubilized receptors with relative potencies similar to those seen with membrane-bound receptors: NPY greater than PYY much greater than pancreatic polypeptide. Scatchard analysis of equilibrium binding data showed the CHAPS extracts to contain a single population of binding sites with a KD of 3.6 +/- 0.4 nM (mean +/- SEM) and a Bmax of 5.0 +/- 0.2 pmol/mg of protein. In addition the 125I-NPY binding to the soluble receptor was not inhibited by guanosine-5'-O-(3-thiotriphosphate), in contrast to the GTP sensitivity displayed by the membrane-bound receptor. Gel filtration chromatography using Sepharose 6B revealed a single peak of binding activity corresponding to a Mr of approximately 67,000, and sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis after chemical cross-linking revealed a single band at Mr 62,000. After solubilization and gel chromatography a 50- to 100-fold purification of the NPY receptor was obtained.     

Molecular properties of solubilized CCK receptor from guinea-pig pancreas

In order to characterize the CCK receptor in guinea-pig pancreas, iodinated CCK-39 was bound to pancreatic membranes and the reversible complex was solubilized using various non-denaturing detergents. In term of recovery of ligand stabilized receptors, the relative potencies were Zwittergent 3-14 greater than CHAPS = CHAPSO greater than digitonin greater than MEGA 10 greater than octyl beta-D-glucopyranoside. The stability of receptor complexes was increased by glycerol. Chromatographic analysis revealed that digitonin was the most efficient detergent for disaggregation of CCK receptor complex since it yielded a 76 kDa component in addition to the large components obtained after solubilization with CHAPS and Zwittergent. Furthermore, CCK receptors were covalently labelled using dissuccinimidyl suberate or UV irradiation of labelled membranes by photoactivable radioiodinated CCK-39 and subsequently solubilized by CHAPS + SDS or by SDS alone. A predominant molecule was characterized by chromatography (76 kDa) and SDS-PAGE (89 kDa). In addition to this component, other components having molecular masses of 130-150 kDa, 57 kDa and 40 kDa were detected by SDS-PAGE. They correspond to minor bands. These bands, except the 40 kDa band, were protected from covalent labelling by the presence of CCK-39 (10(-6) M) during initial incubation. Reduction under beta-mercaptoethanol mainly resulted in the decrease of high molecular weight aggregates (Mr greater than 200 kDa). We concluded that for a given detergent a specific molecular weight pattern of solubilized CCK receptor complex is achieved. The minimal component had a molecular mass of 71-84 kDa according to the method of biochemical analysis used.     

Characterization of cholecystokinin receptors on guinea pig gastric chief cell membranes

The binding of cholecystokinin (CCK) to its receptors on guinea pig gastric chief cell membranes were characterized by the use of 125I-CCK-octapeptide (CCK8). At 30 degrees C optimal binding was obtained at acidic pH in the presence of Mg2+, while Na+ reduced the binding. In contrast to reports on pancreatic and brain CCK receptors, scatchard analysis of CCK binding to chief cell membranes revealed two classes of binding sites. Whereas, in the presence of a non-hydrolyzable GTP analog, GTP gamma S, only a low affinity site of CCK binding was observed. Chief cell receptors recognized CCK analogs, with an order of potency of: CCK8 greater than gastrin-I greater than CCK4. Although all CCK receptor antagonists tested (dibutyryl cyclic GMP, L-364718 and CR1409) inhibited labeled CCK binding to chief cell membranes, the relative potencies of these antagonists in terms of inhibiting labeled CCK binding were different from those observed in either pancreatic membranes or brain membranes. The results indicate, therefore, that on gastric chief cell membranes there exist specific CCK receptors, which are coupled to G protein. Furthermore, chief cell CCK receptors may be distinct from pancreatic or brain type CCK receptors.     

Preparation of enriched plasma membranes from bovine gallbladder muscularis for characterization of cholecystokinin receptors

A method for the preparation of enriched plasma membranes from bovine gallbladder muscularis was developed, validated, and applied to the characterization of receptors for the gastrointestinal hormone cholecystokinin (CCK) on this target. Binding of radioiodinated CCK ligands to this preparation was rapid, reversible, temperature-dependent, saturable, and specific. Only structurally related peptides inhibited CCK binding, and good correlation existed between relative potencies for binding inhibition and for stimulating gallbladder contraction. Computer analysis of CCK-binding data using a nonlinear model-fitting program best fit a model with a single class of sites, with Kd 756 pm and binding capacity 4.5 +/- 1.3 pmol/mg of protein. This degree of enrichment for plasma membranes was adequate for the initial biochemical characterization of this CCK receptor. Affinity labeling using 125I-Bolton Hunter-CCK-33 and m-maleimidobenzoyl-N-hydroxysuccinimide ester identified proteins with Mr = 70,000-85,000, Mr = 120,000-125,000, and Mr = 200,000. Labeling was inhibited in a concentration-dependent manner, with an IC50 of 1 nM CCK-8, and the electrophoretic mobility of these bands was not different under reducing and nonreducing conditions. The major labeled band of Mr = 70,000-85,000 has a lower apparent Mr than that of the analogous band in pancreas labeled with similar methods, supporting the molecular heterogeneity of CCK receptors on these two target tissues.     

Brain CCK receptors are structurally distinct from pancreas CCK receptors

Brain and pancreas cholecystokinin (CCK) receptors differ markedly in their selectivity for CCK analogs. To determine the size and subunit structure of the brain CCK receptor and compare it to that of the pancreas, 125I-CCK33 was covalently cross-linked with ultraviolet light to its receptor on mouse brain particles and purified pancreatic plasma membranes. When CCK was crosslinked to brain membranes, a single consistent major labeled protein band of Mr = 55,000 was observed in both the presence and the absence of DTT. These data with brain receptors contrast to results with pancreatic receptors where two bands of Mr = 120,000 and 80,000 are labeled in the absence and presence of DTT, respectively. These studies indicate, therefore, that the brain and pancreas CCK receptors are structurally and functionally distinct.     

Pancreatic CCK receptors: Characterization of covalently labeled subunits

¹²⁵I-CCK was crosslinked with ultraviolet light to its receptor on pancreatic plasma membranes. The predominant labeled species following polyacrylamide gel electrophoresis had a molecular weight of 120,000 in the absence, and 80,000 in the presence of the reducing agent dithiothreitol. The M_r = 120,000 labeled band could be extracted, reduced and converted to M_r = 80,000. Moreover, peptide mapping with Staph aureus V8 protease showed a similar pattern for the 120,000 and 80,000 dalton bands. The crosslinked receptor could be solubilized with Triton X-100, absorbed to wheat germ agglutinin and eluted with N-acetylglucosamine. The results indicate, therefore, that the CCK receptor is a glycoprotein with subunits coupled by disulfide bonds.     

A novel affinity purification of D-1 dopamine receptors from rat striatum

When rat striatal membranes were pretreated with the sulfhydryl (-SH) modifying reagent, N-ethylmaleimide (NEM) in the presence of the D-1-specific agonist, SKF R-38393, the D-1 dopamine receptor was completely protected from NEM-mediated inactivation. The D-1 receptors, solubilized from these membranes with 1% sodium cholate in the presence of phospholipids, bound with high efficiency (greater than 90%) to mercury-agarose columns. The bound receptors were eluted from the affinity column with a -SH reducing agent, beta-mercaptoethanol. Upon removal of beta-mercaptoethanol from the eluted fractions by inclusion chromatography, the receptor was reconstituted into phospholipid vesicles and assayed for ligand binding activity. The affinity purified receptor exhibited saturable and specific binding of the D-1-specific ligand 125I-SCH 23982, with a Kd of 1.6 nM comparable to that measured in intact membranes and solubilized extracts. The binding capacity of these receptors for 125I-SCH 23982 was 11,000 pmol/mg protein, representing greater than an 8000-fold purification over the starting membrane preparation. The purity of the affinity eluted receptors was estimated to be 78%. The purified receptors retained the pharmacological properties of membrane-bound receptors, including the ability to distinguish between active and inactive enantiomers of specific dopaminergic antagonists. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis and silver staining revealed the presence of two major polypeptides of 74 and 54 kDa. These two polypeptides were absent in those affinity eluted fractions which did not display 125I-SCH 23982-binding activity and also were not detected in preparations obtained from membranes which were NEM-treated in the absence of D-1-specific agonist. The molecular weights of these polypeptides were similar to those of membrane-bound D-1 receptors, when labeled with a D-1-specific photo-affinity ligand, 125I-8-hydroxy-3-methyl-1-(4-azidphenyl)-2,3,4,5-tetrahydro-1H-3-b enzazepine. These two polypeptides may represent glycosylated and deglycosylated forms of the D-1 dopamine receptor.