GABA_B受体的最新进展:从药理学到分子生物学(英文)

Bicuculline-insensitive receptors for the inhibitory neurotransmitter γ-aminobutyric acid (GABA), GABAB receptors, are a distinct subclass of receptors that mediate depression of synaptic transmission and contribute to neu-ronal inhibition. When activated, these receptors reduce transmission at excitatory and inhibitory synapses, as a result of an increase in K conductance, or a decrease in voltage-dependent Ca2 currents. They are also linked to G-proteins, or intracellular effector systems in a very complex manner. The recent development of highly specific and potent agonists and antagonists for these receptors has led to a much better understanding of their physiology and pharmacology, including their heterogeneity, as well as their molecular biology. Over the past year, expression and cloning studies have contributed to major advances in characterizing GABAB receptor structure, with the discovery of the amino acid sequences of GABABRla/R1b splice variants and GABABR2 receptors. These isoforms are     

Pharmacological actions of thymol and an analogue at GABAB autoreceptors

GABA_B autoreceptors inhibit release of GABA from GABAergic nerve terminals. Agonists of these receptors (e.g. baclofen) inhibit, whereas antagonists (e.g. (+)‐(S)‐5,5‐dimethylmorpholinyl‐2‐acetic acid; Sch 50911) enhance release of the transmitter. The actions of thymol (2‐isopropyl‐5‐methylphenol) and the structurally related compound 2‐tert‐butyl‐4‐methylphenol, (4MP) on the release of [³H]‐GABA were examined in rat neocortical slices where the GABAergic nerves had been preloaded with [³H]‐GABA and subsequently stimulated electrically on two occasions (S₁ and S₂). Test agents, baclofen and Sch 50911 were added to the superfusion medium prior to the second period of stimulation (S₂). Stimulation‐induced overflow (SIO) of [³H]‐GABA as a consequence of these stimulations (SIO₁ and SIO₂) were calculated and the effects of agents determined by comparing the SIO₂/SIO₁ ratio in the presence of each agent with that in control tissue. Thymol potentiated the release of [³H]‐GABA (EC₅₀ 170 μmol/L), an action reversed by baclofen (2 μmol/L). Baclofen alone had little effect on GABA release. Release of [³H]‐GABA was inhibited by 4MP (IC₅₀ 3 μmol/L) and this effect was blocked by Sch 50911 (10 μmol/L). Alone, Sch 50911 markedly potentiated the release of GABA. These results imply that 4MP is an agonist of GABA_B autoreceptors; however, further studies are needed to confirm that thymol is indeed a GABA_B autoreceptor antagonist. Of interest are structural differences in these agents. Thymol has a propyl group in the ortho position relative to the phenolic hydroxyl, whereas in 4MP this is a butyl group and the methyl group moves from position 5 to 4. Whether one or both of these changes was responsible for the above actions is unknown.     

The action of structural analogues of γ-aminobutyric acid on binding sites in mouse brain

Nineteen structural analogues of γ-aminobutyric acid (GABA) and the anxiolytics buspirone and diazepam were tested for their ability to modify GABA function in vitro. Each compound was added to the GABA receptor binding assay, the flunitrazepam receptor binding assay, the sodium-dependent GABA binding assay, and the assay for GABA aminotransferase. It was anticipated that some of these analogues would substantially affect several of these assays. The most interesting compound to emerge was phenylthiohydantoic acid. This not only inhibited GABA receptor binding (K_i = 121 μM) but also stimulated flunitrazepam binding and inhibited the activity of GABA aminotransferase. In light of the evidence that a reduced GABA function can be a factor in anxiety, phenylthiohydantoic acid has the potential to be a GABA-mimetic and thus be useful in the treatment of anxiety.     

The use of [3H](-)-DO 710 as selective dopaminergic ligand for binding and autoradiographic studies

DO 710, a benzamide derivative previously shown to display a dopamine antagonistic potency superior to that of sulpiride, was 3H-labeled. Its use as radioligand was assessed in membrane binding and autoradiographic studies. The compound displayed relatively high affinity (Kd 2 nM) and pronounced selectivity for dopamine receptors (distinct from D-1 receptors) as well as low non-specific binding particularly evidenced in autoradiographic experiments. Hence [3H](-)-DO 710 displays distinct technical advantages over commonly used dopaminergic radioligands.     

Baclofen, an agonist at peripheral GABAB receptors, induces antinociception via activation of TEA-sensitive potassium channels

BACKGROUND AND PURPOSE: Central anti-nociceptive actions of baclofen involve activation of K+ channels. Here we assessed what types of K+ channel might participate in the peripheral anti-nociception induced by baclofen.Experimental approach:Nociceptive thresholds to mechanical stimulation in rat paws treated with intraplantar prostaglandin E2.(PGE2) to induce hyperalgesia were measured 3 h after PGE2 injection. Other agents were also given by intraplantar injection. KEY RESULTS: Baclofen elicited a dose-dependent (15 - 240 microg per paw) anti-nociceptive effect. An intermediate dose of baclofen (60 microg) did not produce antinociception in the contralateral paw, showing its peripheral site of action. The GABAB receptor antagonist saclofen (12.5 - 100 microg per paw) antagonized, in a dose-dependent manner, peripheral antinociception induced by baclofen (60 microg), suggesting a specific effect. This antinociceptive action of baclofen was unaffected by bicuculline, GABAA receptor antagonist (80 microg per paw), or by (1,2,5,6 tetrahydropyridin-4-yl) methylphosphinic acid, GABAC receptor antagonist (20 microg per paw). The peripheral antinociception induced by baclofen (60 microg) was reversed, in a dose-dependent manner, by the voltage-dependent K+ channel blockers tetraethylammonium (7.5 - 30 microg per paw) and 4-aminopyridine (2.5 - 10 microg per paw). The blockers of other K+ channels, glibenclamide (160 microg), tolbutamide (320 microg), charybdotoxin (2 microg), dequalinium (50 microg) and caesium (500 microg) had no effect. CONCLUSIONS AND IMPLICATIONS: This study provides evidence that the peripheral antinociceptive effect of the GABAB receptor agonist baclofen results from the activation of tetraethylammonium-sensitive K+ channels. Other K+ channels appear not to be involved.     

Characterization of specific binding sites for [3H](d)-N-allylnormetazocine in rat brain membranes

Binding of [3H](d)-N-allylnormetazocine ([3H](d)-NANM) to rat brain membranes is stereospecific, reversible, and saturable (Bmax = 260 fmol/mg of protein) and manifests moderately high affinity (Kd = 20 nM). The rank order of potency among opioidbenzomorphans and phencyclidine (PCP) analogs for competition for [3H](d)-NANM-binding sites is as follows: (d)-NANM = PCP-3-OH greater than (d)-cyclazocine greater than N-ethylphenylcyclohexylamine greater than PCP greater than (l)-cyclazocine = dextrorphan greater than (d/l)-ethylketocyclazocine greater than (d/l)-bremazocine greater than (1)-NANM greater than 1-phenylcyclohexylamine greater than levorphanol. Other opioid ligands, relatively selective for each of the types of opioid binding sites other than sigma, such as morphine (mu), H-Tyr-D-Ala(Me)Phe-NH-CH2-OH (mu), D-Ala2-D-Leu5-enkephalin (delta), tifluadom (kappa), and U 50488 (kappa) as well as etorphine and naloxone were all unable to compete with [3H](d)-NANM for specific binding even at a concentration of 1 microM. Regional distribution studies of [3H](d)-NANM-binding sites show high density in the hippocampus, thalamus, hypothalamus, and amygdala and low density in cerebellum and nonfrontal neocortex membranes of the rat brain. These binding sites are very sensitive to protein-modifying enzymes and reagents such as trypsin and N-ethylmaleimide and to heat denaturation. These results provide direct biochemical evidence for the existence of distinct (d)-NANM-binding sites in rat brain. In addition, this study supports the view that PCP and several of its analogues and the dextrorotatory isomers of psychotomimetic benzomorphans may act at a common recognition site in rat central nervous system.     

3H]zacopride: ligand for the identification of 5-HT3 recognition sites

[3H]Zacopride displayed saturable binding to homogenates of the rat entorhinal cortex as measured by the inclusion of the 5-HT3 receptor antagonist BRL43694 in the incubation media. Scatchard analysis indicated a single high affinity binding site (KD 0.76 +/- 0.08 nM, Bmax 77.5 +/- 6.5 fmol (mg protein)-1) with a Hill slope close to unity. Other 5-HT3 receptor antagonists (zacopride, ICS 205-930, GR38032F, GR65630, metoclopramide and cocaine) also competed for the binding site displacing 60% of the total [3H]zacopride binding. 5-HT and 2-methyl-5-HT also were competitive antagonists for [3H]zacopride binding whereas 5-HT1/5-HT2 agonists and antagonists, and agents acting on other neurotransmitter receptors had Ki values greater than 10(-5) M. It is concluded that [3H]zacopride may prove a useful ligand for the study of 5-HT3 recognition sites.     

3H]acetylcholine and [3H](-)nicotine label the same recognition site in rat brain

High affinity binding sites for [3H]acetylcholine and [3H](-)nicotine in rat brain were compared with respect to key characteristics, any one of which should distinguish them if they are different. The density of binding sites for each ligand varied approximately 4-fold in five areas of rat forebrain, but in each of these areas and in human cerebral cortex as well, the densities of [3H]acetylcholine- and [3H](-)nicotine-binding sites were indistinguishable. The affinity of [3H](-)nicotine was higher than that of [3H]acetylcholine, but nicotinic cholinergic drugs competed for the sites labeled by the two ligands with similar affinities; and in each case, the site labeled displayed marked stereoselectivity for the enantiomers of nicotine. The binding of [3H]acetylcholine and [3H](-)nicotine was decreased to the same extent by preincubation of tissues with dithiothreitol, and the binding was restored by subsequent treatment with 5,5'-dithiobis-2-nitrobenzoic acid, indicating that a disulfide bond is required at or near the binding site for each ligand. Treatment of rats with nicotine for 10 days increased the density of binding sites for both ligands, and treatment with the cholinesterase inhibitor soman for 9 days decreased the density of binding sites for both ligands. Taken together, these results indicate that [3H]acetylcholine and [3H](-)nicotine bind to the same nicotinic cholinergic recognition site in rat brain.     

(-)PPAP: a new and selective ligand for sigma binding sites

Most agents employed for the investigation of sigma (sigma) binding sites display relatively low affinity for these sites, bind both at sigma sites and at either phencyclidine (PCP) sites or dopamine receptors with similar affinity, and/or produce some dopaminergic activity in vivo. We describe a new agent, (-)PPAP or R(-)-N-(3-phenyl-n-propyl)-1-phenyl-2-aminopropane hydrochloride, that binds with high affinity and selectivity at sigma (IC50 = 24 nM) versus either PCP sites (IC50 greater than 75,000 nM) or D1 and D2 dopamine receptors (IC50 greater than 5,000 nM). The sigma affinity of this agent is comparable to that of the standard ligands (+)-3-PPP and DTG. Furthermore, although (-)PPAP is structurally related to amphetamine, it neither produces nor antagonizes amphetamine-like stimulus effect in rats trained to discriminate 1 mg/kg of S(+)amphetamine from saline.     

Autoradiographic distribution of [3H]-suriclone binding sites in the rat brain

Suriclone is a potent non-benzodiazepine anxiolytic belonging to the cyclopyrrolone family. This study examines the distribution and properties of [³H]-suriclone binding sites in rat brain by autoradiography. The binding of [³H]-suriclone to rat brain sections was displaced completely by 1 m?M flumazenil; saturation experiments revealed a single class of binding sites with a K_D value of 1.19 nM and a Bmax of 1.05 pmol/mg protein. The pharmacological specificity of [³H]-suriclone binding was close to that of the benzodiazepine binding site on the GABA_A receptor. Unlike benzodiazepine (BZ) agonists, [³H]-suriclone binding to rat brain sections was not increased in the presence of GABA, and unlike imidazopyridines such as alpidem, suriclone was unable to discriminate between the BZ₁ and BZ₂ phenotype of the GABA_A receptor. The autoradiographic distribution of [³H]-suriclone binding sites in the rat brain was heterogeneous and relatively similar to that previously described for GABA_A receptors labelled by benzodiazepines; some regions, however, such as certain cortical areas, displayed a different labelling. The highest densities were found in frontal, occipital, parietal, and cingulate cortices (layers I, II, III, IV), in the molecular layer of the cerebellum, in the superior colliculus and in the hippocampus. Moderate binding densities appeared in numerous areas such as the inferior colliculus, the central grey, the dorsal raphé, and the hypothalamus whereas caudate putamen, globus pallidus, and thalamic nuclei displayed a low density of [³H]-suriclone binding sites. ©1995 Wiley-Liss, Inc.     

3H] (-)sulpiride binding in rat striatum, cortex and anterior pituitary: an improved assay

The interaction of [3H] (-)sulpiride with D2 dopamine (DA) receptors in the striatum, anterior pituitary and medial-prefrontal cortex was studied in rats, using an improved [3H] (-)sulpiride radioreceptor binding technique. Incubation on ice and a fast filtration resulted in higher specific binding (approximately 85% of total binding) a lower affinity constant (about 3 nM) and higher Bmax than reported with previous procedures. Pharmacological interactions confirmed the high selectivity of [3H] (-)sulpiride for D2 DA receptors.     

Anti-cancer drug diffusion within living rat brain tissue: an experimental study using [3H](6)-5-fluorouracil-loaded PLGA microspheres.

This study was performed (i) to monitor the diffusion of the anti-cancer drug 5-fluorouracil (5-FU) and (ii) to elucidate the fate of poly(lactide-co-glycolide) (PLGA) based microspheres within living rat brain tissue upon intracranial implantation. Drug-loaded microparticles were prepared using a solvent emulsion/extraction process and administered into healthy and C6 glioma-bearing Sprague-Dawley rats. The same surgical procedure was carried out with magnetite-loaded microspheres. To monitor 5-FU diffusion from the implantation site, tissue combustion was performed on animals implanted with tritiated drug microspheres. T2-weighted nuclear magnetic resonance imaging was undertaken on animals implanted with magnetite-loaded microspheres to determine microsphere localization after deposit. Results show that an important microparticle backflow occurs in healthy rats, whereas the microspheres remain at the site of administration in C6 glioma-bearing rats. Drug diffusion is limited to the vicinity of the implantation site.     

Novel sites for phenylalkylamines: characterisation of a sodium-sensitive drug receptor with (-)-[3H]emopamil.

(-)-Emopamil ((S)-emopamil, (2S)-2-isopropyl-5-(methylphenethylamino)- 2-phenylvaleronitrile hydrochloride) is a Ca(2+)-antagonistic phenylalkylamine which also blocks serotonin (5-HT2) receptors and has antiischemic properties. The (-)-[3H]emopamil tissue distribution profile of specific binding is in striking contrast to that observed for (+)-[3H]PN 200-110 or (-)-[3H]desmethoxyverapamil: (-)-[3H]emopamil labels membrane fractions from guinea-pig liver much greater than adrenal gland greater than kidney approximately lung approximately ductus deferens approximately brain approximately skeletal muscle. Binding to liver membrane was saturable (KD = 12.8 nM, Bmax = 35 pmol/mg of protein), stereoselective, reversible (K-1 = 0.22 min-1 at 25 degrees C) and inhibited by tetraethylammonium (IC50: 1.8 mM) greater than Li+ (IC50: 12.5 mM) approximately Na+ (IC50: 13.6 mM) and [NH4+] (IC50: 79.3 mM) but not by Rb+, Cs+ or K+. The high-affinity liver membrane binding sites have a pharmacological profile that is distinct from the phenylalkylamine receptor domain of the voltage-dependent L-type Ca2+ channel. Similar sites exist in brain and other tissues, albeit with a lower density. Amiodarone, butoprozine and amiloride derivatives bind with high affinity whereas 1,4-dihydropyridines do not interact at all. It is suggested that the novel phenylalkylamine site is linked to a sodium-dependent carrier or transport system.     

Assessment of the chloroethyl analog of 3-PPP (N-n-propyl-3-(3-hydroxyphenyl)-piperidine), 3-PPP-C1 as an irreversible ligand for central dopaminergic recognition sites

NCA, the chloro analog of the potent dopamine agonist NPA is an irreversible ligand at dopamine receptors in mammalian brain. The chloroethyl analog of the recently described putative dopamine autoreceptor agonist 3-PPP, 3-PPP-C1, was evaluated for its potential use as an irreversible autoreceptor ligand. N-Chloroethylation of 3-PPP reduced the intrinsic affinity of the agonist seven-fold and, consequently, in contrast to NCA, it was found that 3-PPP-C1 was not a good irreversible ligand at dopamine receptors.     

Alpha 2-adrenoceptor in HT29 human colon adenocarcinoma cell-line: study of [3H](-)-adrenaline binding

The HT29 human adenocarcinoma cell-line was used to investigate the binding of [3H](-)-adrenaline at alpha 2-adrenoceptors. All aspects of the study indicated that alpha 2-adrenoceptors were specifically labeled. [3H](-)-adrenaline bound with high affinity (KD = 2.28 +/- 0.41 nM) to a single population of non-interacting sites. The rank order of adrenoceptor antagonists (yohimbine greater than phentolamine much greater than prazosin) and agonists (UK-14,304 greater than clonidine greater than (-)-adrenaline greater than (-)-noradrenaline greater than (+)-adrenaline greater than (+)-noradrenaline greater than amidephrine) to compete with [3H](-)-adrenaline binding showed that the labeled sites were alpha 2-selective and stereospecific. Comparison of the binding parameters of [3H](-)-adrenaline with those of [3H]clonidine (partial-agonist) and [3H]yohimbine (antagonist) indicated that [3H](-)-adrenaline and [3H]clonidine labeled a similar number of sites (156 +/- 13 versus 175 +/- 21 fmol/mg protein) and that [3H]yohimbine (Bmax = 246 +/- 22 fmol/mg protein) labeled more sites than the 3H-agonists. Data on the inhibition of [3H]yohimbine binding by (-)-adrenaline was better fitted to a two-site model and revealed (1) that the KiL/KiH ratio was higher for (-)-adrenaline than for clonidine (2) that both agonists recognized the same percentage of high-affinity receptors. The results from a kinetic study of [3H](-)-adrenaline binding were apparently inconsistent with the equilibrium data. Both the association and dissociation were bi-exponential, suggesting a relative heterogeneity of the labeled sites. The tritiated physiological full-agonist was moreover able to induce tight-binding. The practical consequences of this property are discussed.     

Labelling of human platelet alpha 2-adrenoceptors with the full agonist [3H](-)adrenaline

[3H](-)Adrenaline has been used to characterize alpha 2-adrenoceptors on human platelets. Although (-)adrenaline is a good substrate for the platelet enzyme MAO-B, enzymatic inhibition was not a prerequisite to quantify the specific binding of the radioligand to platelet membranes. At 25 degrees C the binding was rapid (t1/2 of association: 10.3 min), reversible (t1/2 of dissociation: 4.0 min) and linearly dependent on the amount of protein present in the assay. The binding sites for [3H](-)adrenaline showed the specificity required for an alpha 2-adrenoceptor. The rank order of potency of inhibitors of [3H](-)adrenaline binding was oxymetazoline greater than idazoxan congruent to phentolamine congruent to clonidine congruent to (-)adrenaline greater than (-)noradrenaline greater than yohimbine much much greater than phenylephrine much greater than prazosin greater than (+)propranolol. Moreover, the nucleotide guanosine triphosphate (GTP) inhibited in a concentration-dependent manner (10(-9)-10(-4) M) the specific binding of [3H](-)adrenaline, suggesting that the radioligand preferentially labelled the high affinity state of the alpha 2-adrenoceptor. Linear (Scatchard) and non-linear analyses of [3H](-)adrenaline binding indicated the existence of a single population of non-interacting sites (KD = 2.5-2.7 nM; Bmax = 49-53 fmol/mg protein). The binding characteristics for [3H](-)adrenaline were not affected by age and sex of the donors or by freezing of platelet-rich plasma. In the same subjects alpha 2-adrenoceptor density (Bmax) for the full agonist [3H](-)adrenaline was 2.9-fold lower than that quantitated by the selective antagonist [3H]yohimbine. The inhibition constants (Ki) of adrenergic drugs and of various antidepressant drugs in competing with [3H](-)adrenaline were correlated with the inhibition constants of these drugs in competing with [3H]clonidine (r = 0.96; P less than 0.001) which suggests that both radioligands labelled the same alpha 2-adrenoceptor on the human platelet. The binding of the full agonist [3H](-)adrenaline to human platelet membranes might be a useful tool for the study of dysfunctions related to the high affinity state of the alpha 2-adrenoceptor.     

Synthesis and anti-diabetic activity of （RS）-2-ethoxy-3-{4-[2-（4-trifluoro-methanesulfonyloxy-phenyl）-ethoxy]-phenyl}-propionic acid

AIM: To synthesize and study the anti-diabetic activity of (RS)-2-ethoxy-3-{4-[2-(4-trifluoromethanesulfonyloxy-phenyl)-ethoxy]-phenyl}-propionic acid (compound I). METHODS: Compound I was prepared in 6 steps, using 4-(2-hydroxy-ethyl)-phenol as the starting material. The in vitro selectivity and potency of target compound I, rosiglitazone and WY-14643 on human PPARalpha and PPARgamma were determined in reporter gene assays. In vivo, rosiglitazone and compound I were administered orally to KK(Ay) mice for 14 d. Insulin tolerance tests and oral glucose tolerance tests were performed on the 10th and 14th day of treatment, respectively. At the end of the treatment, sera were collected for biochemical analysis. RESULTS: In vitro, compound I significantly activated both PPARalpha and PPARgamma. In vivo, compound I corrected the impaired insulin and glucose tolerance of KK(Ay) mice, and produced a significant reduction in plasma triglyceride levels after 14 d of treatment. The effect produced was significant compared with the control group. CONCLUSION: Both in vitro and in vivo anti-diabetic activity studies for compound I were conducted and the data suggest that this compound is a potentially effective anti-diabetic agent.     

Characterization of forskolin binding sites in rat brain membranes using [14,15-3H]14,15-dihydroforskolin as a ligand

Summary [14, 15-³H]14, 15-Dihydroforskolin ([³H]DHF) has been used as a radioactive ligand to identify forskolin binding sites in rat brain membranes. The binding was saturable and reversible. The binding sites showed positive cooperative properties as evident from an upward convex Scatchard plot and a Hill coefficient of 1.6. The equilibrium dissociation constants (K _D) were in the range between 10 M and 10 nM as estimated from the limiting slopes of the curved Scatchard plot. Half-maximal saturation of the binding sites was observed at a ligand concentration of 225 nM. The binding kinetics were very rapid: Binding equilibrium was reached in less than 2 min and a large excess of cold forskolin displaced 80% of the radioligand within 2 min. The dissociation reaction was not first order, characterized by a decreasing dissociation rate constant. Bound [³H]DHF could be displaced with forskolin (IC₅₀ 0.3 M), 14,15-dihydroforskolin (IC₅₀ 0.8 M) and 7-desacetylforskolin (IC₅₀ 3 M). However, nucleotides (ATP, GTP) and other receptor ligands (adenosine, isoproterenol) had no effect on the binding. Although the density of the forskolin binding sites (3.2 pmole/mg protein) is similar to those of other adenylate cyclase linked receptors, discrepancies between the K _D and the ED₅₀ obtained in adenylate cyclase studies and the finding that activation of the enzyme by forskolin is negative cooperative makes it difficult to clearly relate the binding sites to adenylate cyclase.     

Interactions of spironolactone with (+)-[3H]-isradipine and (-)-[3H]-desmethoxyverapamil binding sites in vascular smooth muscle.

Spironolactone partially inhibited the specific binding of (+)-[3H]-isradipine and (-)-[3H]-desmethoxyverapamil to vascular smooth muscle membranes. It is suggested that spironolactone interacts at a binding site of the calcium channel complex and allosterically modulates ligand binding at receptor sites in the channel.     

3H]Ro 19-6327: a reversible ligand and affinity labelling probe for monoamine oxidase-B

This study demonstrated the existence of specific binding sites for [3H]Ro 19-6327 in human platelet membranes. This compound is a novel, time-dependent inhibitor of monoamine oxidase type B (MAO-B) and is structurally closely related to [3H]Ro 16-6491. The density of the sites labelled with high affinity by [3H]Ro 19-6327 was similar to that observed in previous studies with [3H]Ro 16-6491 as ligand. Binding was reversible at 20 degrees C and showed a relatively slow dissociation (t1/2 = 220 min). The dissociation rate was markedly decreased (t1/2 = greater than 24h) at 0 degrees C. MAO-B, but not MAO-A inhibitors, effectively prevented the binding of [3H]Ro 19-6327. Like [3H]Ro 16-6491, [3H]Ro 19-6327 is recognized as a substrate by MAO-B, being eventually deaminated by the enzyme. Since the deaminated aldehyde derivative of Ro 19-6327 did not inhibit MAO-B, a still unidentified reversible adduct, formed at the MAO-B active site, might explain the high potency and selectivity of [3H]Ro 19-6327. Incubation of the radioligand-enzyme complex from platelet and brain membranes with NaBH3CN and acetic acid (to pH 4.5) caused the irreversible incorporation of the radioactivity into a single polypeptide as shown by SDS-PAGE analysis. This polypeptide had a molecular weight identical to that of the MAO-B subunit, i.e. 58,000. The presence of unlabelled MAO-B inhibitors in the incubation mixture prevented the covalent incorporation of [3H]Ro 19-6327. The irreversible MAO-B inhibitor, [3H] pargyline, labelled a protein with a molecular weight identical to the protein labelled by [3H]Ro 19-6327.(ABSTRACT TRUNCATED AT 250 WORDS)