CHARACTERIZATION AND SUBCELLULAR LOCALIZATION OF BRAIN MUSCARINIC RECEPTORS LABELLED IN VIVO BY [3H]DEXETIMIDE

Abstract— [³H]Dexetimide specifically labels brain muscarinic receptors in vivo . After i.v. injection of labelled drug into rats, radioactivity specifically accumulates in brain regions containing muscarinic receptors but not in cerebellum. This accumulation is stereospecific, saturable and displaceable by unhbelled dexetimide. In contrast, [³H]levetimide, the inactive enantiomer, does not show such preferential uptake or stereospecific displacement.
An analytical approach was used to study the subcellular distribution of [³H]dexetimide binding sites. After differential centrifugation the binding sites are mainly recovered in the microsomal fraction from different brain regions but not from the cerebellum. After displacement the radioactivity is found in the supernatant. After equilibration in a density gradient the distribution pattern of [³H]dexetimide is bimodal, like that of 5'-nucleotidase, with a major peak in a region of low density.
When the microsomal fraction was treated with digitcnin, three groups of membrane were characterized by isopycnic centrifugation on the basis of their differential shift to higher densities. Evidence is provided that the postsynaptic membranes bearing muscarinic receptors belong to the class of plasma membranes. Finally, digitonin treatment may represent a useful tool to produce subfractions enriched in postsynaptic membranes which can now be identified biochemically in binding experiments.     

THE BINDING OF [3H]MEPYRAMINE TO HISTAMINE H1 RECEPTORS IN GUINEA-PIG BRAIN

The promethazine-sensitive binding of [³H]mepyramine to a membrane fraction from guinea-pig whole brain is saturable with a dissociation constant of 1.7 × 10^-9M. The maximum amount of [³H]mepyramine binding varied widely between preparations, range 122–365 pmol/g protein, with a mean value of 227 ± 52 pmol/g protein. The inhibition of [³H]mepyramine binding by a number of drugs correlated closely with their potency as histamine H₁ antagonists. (+) Chlorpheniramine was 240-fold more potent as an inhibitor of [³H]mepyramine binding than (-)-chlorpheniramine. All antagonists inhibited the binding of [³H]mepyramine to the same extent, but the Hill coefficients characterising the inhibition curves did not all approximate to unity, the value expected for a simple antagonist-receptor equilibrium. The distribution of histamine H₁ receptors, defined by the promethazine-sensitive binding of [³H]mepyramine, in 11 different brain regions was uneven with the largest amounts in cerebellum, superior and inferior colliculus and hypothalamus and the smallest in caudate nucleus, brain stem and spinal cord.     

[3H]GLYCOGEN HYDROLYSIS IN BRAIN SLICES: RESPONSES TO NEUROTRANSMITTERS AND MODULATION OF NORADRENALINE RECEPTORS

Abstract— Different agents have been investigated for their effects on [C³H]glycogen synthesized in mouse cortical slices. Of these noradrenaline, serotonin and histamine induced clear concentration-dependent glycogenolysis.
[C³H]Glycogen hydrolysis induced by noradrenaline appears to be mediated by beta-adrenergic receptors because it is completely prevented by timolol, while phentolamine is ineffective. It seems to involve cyclic AMP because it is potentiated in the presence of isobutylmethylxanthine; in addition dibutyryl cyclic AMP (but not dibutyryl cyclic GMP) promotes glycogenolysis.
Lower concentrations of noradrenaline were necessary for [C³H]glycogen hydrolysis (EC₅₀= 0.5μM) than for stimulation of cyclic AMP accumulation (EC₅₀= 8μM).
After subchronic reserpine treatment the concentration-response curve to noradrenaline was significantly shifted to the left (EC₅₀= 0.09 ± 0.02 μM as compared with 0.49 ± 0.08 μM in saline-pretreated mice) without modifications of either the basal [C³H]glycogen level, maximal glycogenolytic effect, or the dibutyryl cAMP-induced glycogenolytic response.
In addition to noradrenaline, clear concentration-dependent [³H]glycogen hydrolysis was observed in the presence of histamine or serotonin. In contrast to the partial [³H]glycogen hydrolysis elicited by these biogenic amines, depolarization of the slices by 50 mM K⁺ provoked a nearly total [C³H)glycogen hydrolysis.     

BINDING OF [3H]KAINIC ACID, AN ANALOGUE OF l-GLUTAMATE, TO BRAIN MEMBRANES

—The specific binding of [³H]kainic acid to synaptic membranes from rat brain was saturable with a dissociation constant of about 60 nm . The apparent maximal number of binding sites was about 1 pmol/mg protein. The most effective displacer of specific [³H]kainic acid binding was quisqualic acid, a powerful excitant which is structurally similar to l -glutamate. However, quisqualic acid was one-third as potent a displacer as kainic acid itself. l -Glutamate was the next potent in displacing [³H]kainic acid binding, but also was less effective (1/25) than kainic acid itself. All other compounds including suspected neurotransmitters were at least an order of magnitude lower in potency compared to l -glutamate. When various tissues and brain regions were tested for specific [³H]kainic acid binding, we found the specified binding was localized to grey matter in the brain. In studies of subcellular fractionation of the brain, we found that crude synaptosomal membrane preparations were most enriched in specific [³H]kainic acid binding. Specific [³H]kainic acid binding in various regions of the rat brain varied 5- to 6-fold.     

HETEROGENEITY OF HISTAMINE Hi-RECEPTORS: SPECIES VARIATIONS IN [3H]MEPYRAMINE BINDING OF BRAIN MEMBRANES

[³H]Mepyramine binds with high affinity to membranes from brain of human, rat, guinea-pig, rabbit and mouse with drug specificity indicating an association with histamine H₁receptors. Considerable species differences occur in the affinity of [³H]mepyramine, with guinea-pig and human having 34 times greater affinity than rat, mouse or rabbit. The greater affinity of [³H]mepyramine in guinea-pig than in rat is attributable both to faster association and slower dissociation rates in guinea-pig. Species differences in affinity for H₁ receptor sites occur for some antihistamines but not for others. Some tricyclic antidepressant and neuroleptic drugs are extremely potent inhibitors of [³H]mepyramine binding, exceeding in potency any H₁ antihistamines examined. The tricyclic antidepressant doxepin and the neuroleptic clozapine are the most potent of all drugs examined in competing for [³H]mepyramine binding. The regional distribution of specific [³H]mepyramine binding differs considerably in the various species examined.     

ALTERATIONS IN THE BINDING OF [3H]LEUCINE-ENKEPHALIN TO STRIATAL MEMBRANES BY ENDOGENOUS FACTORS

—Saturation binding studies with [³H]leu-enk ([tyrosyl-3, 5-³H(N)]⁵leu-enkephalin) revealed the presence of high and low affinity binding sites in a paniculate fraction derived from rat striatum. The binding of [³H]leu-enk to the high affinity component (K_D= 2.0 ± 0.3 nM) was sensitive to morphine and levorphanol, while the binding to the low affinity component (K_D= 21 ± 2 nM) was not. Incubation of the membranes, prior to assay for 30 min at 37°C, followed by centrifugation at 27, 000 g for 20 min in order to pellet the membranes allowed the detection of a factor, present in the high speed supernatant, which caused a dose-dependent inhibition of the binding of [³H]leu-enk to the morphine-sensitive and insensitive binding components. Investigations into the nature of the morphine-insensitive binding component demonstrated that it was an artifact since it was not detectable when bound and free ligand were separated by centrifugation. Furthermore, [³H]leu-enk bound to Whatman glass fiber filters, used to collect bound ligand, in a morphine-insensitive manner, and under conditions where the binding of [³H]leu-enk to the morphine sensitive component diluted proportionally with serial dilutions of the membranes, the binding to the morphine-insensitive component did not. The factor present in the high speed supernatant did not dialyze and its effects were mimicked by either trypsin or soybean trypsin inhibitor, but not by bovine serum albumin. The apparent inhibition of the binding of [³H]leu-enk to these binding components is probably not of biological significance, but the fact that the artifactual morphine insensitive binding component of striatal membranes has been shown to decrease by 20–30% following lesions of the substantia nigra suggests that the influence of this endogenous factor must be controlled for.     

CHARACTERISTICS AND REGIONAL DISTRIBUTIONS OF TWO DISTINCT [3H]NALOXONE BINDING SITES IN THE RAT BRAIN

Using [³H]naloxone at a concentration of 4.5 nm , the potent opiate agonist etorphine as well as the potent antagonist diprenorphine displace only about 75% of specific naloxone binding P₂ fractions from rat whole forebrain, without additive effect. Several other opiates and antagonists completely displace specific naloxone binding. This indicates that etorphine and diprenorphine specifically bind to one and the same naloxone binding site (type I) while leaving another naloxone binding site (type II) unaffected. Type I binding sites are much more thermo-labile than type II. [³H]Naloxone binding to type I sites is unaffected by incubation temperature in the range 10 to 25°C. while binding type II sites decreases rapidly with increasing incubation temperature, no specific type II binding being detectable at or above 20°C. The two naloxone receptor types also differ with respect to pH dependence, and affinity for naloxone with types I and II having affinity constants (K_d) of 2 and 16 nm , respectively, at 0°C. The two binding sites have different regional distributions with high relative levels of type II receptors in cerebellum and low relative levels in pons-medulla and striatum. In whole rat brain there are about 4 times as many type II receptors as type I. These results suggest that naloxone and several other opiate agonists and antagonists bind to two distinct receptor types which are probably not agonist/antagonist aspects of the same receptor.     

GLYCINE SPECIFIC [3H]STRYCHNINE BINDING IN THE PIGEON CNS

Abstract— Glycine sensitive [³H]strychnine binding has been demonstrated in the pigeon CNS. The kinetic parameters in spinal cord, tectum, telencephalon and cerebellum in this order were: K _n, (nM) 5 , 5, 2.3, 6 and B _max (fmol/mg protein) 1000, 275, 92, 20. The effect of various amino acids and strychnine analogs on the binding was measured. Retinal ablation caused a 22% decrease in the glycine sensitive [³H]strychnine binding in the tectum while lesions destroying up to 80% of the nucleus Ipc had no effect.     

CHEMICAL SYNTHESIS OF [3H]CEREBROSIDE [35S]SULFATE AND ITS IN VIVO METABOLISM IN RAT BRAIN

—Double-labeled sulfatide containing [3-³H]lignoceric acid and [³⁵S]sulfate was synthesized and injected intracerebrally into 28-day-old rats. The ³H-labeled sulfatide was synthesized by condensing (RS)-[3-³H]lignoceroyl chloride with lysosulfatide which had been obtained by saponification of sulfatide. The ³⁵S-labeled sulfatide was synthesized by using [³⁵S]sulfuric acid for sulfating 2′, 4′, 6′-tri-benzoyl-galactosyl N-fatty acyl, N-benzoyl-3-0-benzoyl-sphingosine, which had been obtained by per-benzoylation followed by solvolysis of calf brain nonhydroxycerebrosides. The perbenzoylated [³⁵S]sul-fatide was then subjected to mild alkaline saponification. Eight hours following the injection, the brain lipids contained various radioactive sphingolipids in addition to sulfatides. Fourteen per cent of the injected ³H was recovered in total lipids, and 26% of this was found in sulfatide. Nonhydroxy- and hydroxyceramides, nonhydroxy- and hydroxycerebrosides, and polar lipids contained 7, 1, 8, 3, and 22 per cent of the ³H found in total lipids, respectively. On the other hand, only 6% of the ³⁵S injected was recovered in total lipids; 63% of this was found in sulfatide, 5% in a mixture of seminolipid and cholesterol sulfate and 10% in a water-soluble material.     

POTASSIUM-INDUCED RELEASE OF [3H]GABA AND OF [3H]NORADRENALINE FROM NORMAL AND RESERPINIZED RAT BRAIN CORTEX SLICES. DIFFERENCES IN CALCIUMDEPENDENCY, AND IN SENSITIVITY TO POTASSIUM IONS

The release of [³H]GABA induced by elevated extracellular potassium (K)_o, from thin rat brain cortex slices, has been compared with that of [³H]noradrenaline ([³H]NA), released by the same procedures, both from normal slices, and from slices pre-treated with reserpine and nialamide, [³H]NA being predominantly a vesicular component in the former situation, and a soluble substance in the latter one. 46 mM-(K)_o released considerably more [³H]NA from normal than from drug-treated slices, while the release of GABA was about two thirds of the latter. When 4min ‘pulses’ of increasing concentrations of potassium were applied, it was observed that the release of GABA and of [³H]NA from drug-treated slices increased in proportion to (K)_o, up to 36-46 mM and then declined considerably with higher (K)_o. The dependency of potassium-induced release on the concentration of calcium in the medium, indicated that release of [³H]NA from normal slices was proportional to calcium up to 1.5-2 mM, while that of [³H]NA from drug-treated slices increased up to 0.5 mM-Calcium, and then declined with higher concentrations. GABA release also increased up to 0.5 mM-calcium, but no further changes were observed at higher concentrations. The calcium antagonist D-600 inhibited high (K)_o-induced release of [³H]NA from normal slices to a greater extent than that of [³H]GABA or of [³H]NA from drug-treated slices. These results, in which elevated (K)_o-induced release of [³H]GABA resembles considerably that of soluble NA, but differs from that of NA present in synaptic vesicles, suggest that release of [³H]GABA also occurs from the soluble cytoplasmic compartment, and that the partial calcium requirement that is found is unrelated to that of transmitter secretion. These findings are also a further indication of the lack of specificity of elevated (K)_o as a stimulus for inducing transmitter secretions.     

INCREASED BINDING OF [3H]COLCHICINE TO VISUAL CORTEX PROTEINS OF DARK-REARED RATS ON FIRST EXPOSURE TO LIGHT

Abstract— The binding of [³H]colchicine (or a functionally similar metabolite) to acid-insoluble material in vivo was measured in the motor and visual cortices of littermates rats which were either dark-reared (D), exposed to light for 3 h or 24 h (L), or raised normally (N) in 12 h light/12 h dark animal house conditions. No significant differences were found in the binding in the motor cortex of the 3 h or 24 h L, D or N animals, but in the visual cortex after 3 h of light exposure a 23% elevation in binding was measured in L compared with D animals and a small though non-significant (10%) increase in binding was also observed in this region in L compared with N animals. After 24 h of light exposure, binding of the label in the L animals fell near to that of the N and D animals. The results of vinblastine precipitation experiments suggested that much of the radioactivity was bound to the protein tubulin, and this was confirmed when no increased binding of an analogue of colchicine, lumi-colchicine, was observed after 3 h of light exposure in L compared with D animals. It is suggested that these experiments show that colchicine can be used as a marker for changes in the tubulin population in light exposed animals, and demonstrate the transient nature of the increase in tubulin quantity, as opposed to a lasting effect on its synthesis. Further, they argue strongly in support of the idea that a component of protein flow from neuronal cell body to axons and dendrites in light exposed animals, is subject to environmental modification.     

DIFFERENT LABELLING PATTERNS IN MOUSE LYMPHOID TISSUES WITH [3H]DEOXYCYTIDINE AND [3H]THYMIDINE

The percentages of labelled lymphocytes in smear preparations of mouse thymus were higher than those in similar preparations of mesenteric lymph nodes with either generally labelled tritiated deoxycytidine, [³H]CdR, or tritiated thymidine, [³H]TdR. Lymphocytes in the thymus cortex and in germinal centres of mesenteric lymph nodes were intensely labelled with [³H]CdR, whereas with [³H]TdR lymphocytes in the peripheral region of thymus and medullary cords of mesenteric lymph nodes were heavily labelled. The majority of lymphocytes in thymic cortex and germinal centres of mesenteric lymph nodes were labelled weakly with [³H]TdR. Thus, labelling patterns with [³H]CdR differed from those with [³H]TdR in lymphoid tissues of the mouse. Mouse lymphocytes can utilize [³H]CdR as a precursor molecule for cytosine and thymine in DNA. The ratio of radioactivity of thymine to that of cytosine was measured biochemically in DNA extracted from lymphocytes labelled with [³H]CdR. This radioactivity ratio in thymus was higher than that in mesenteric lymph nodes. These results suggest that the metabolic activities of utilizing CdR for DNA synthesis differ within lymphocyte populations in various lymphoid tissues in the mouse.     

STUDY OF RNA IN SUBCELLULAR FRACTIONS FROM RAT BRAIN: SIMULTANEOUS INCORPORATION OF [14C]URIDINE AND [3H]METHYL-l-METHIONINE

Abstract— By using a combination of subcutaneous and intraventricular injections of [¹⁴C]uridine and [³H]methyl- l -methionine we have obtained maximum incorporation in about 40 min of both radioactive precursors into nuclear RNA from rat brain. In this nuclear fraction we found at least two different types of RNA that were rapidly labelled. One of them incorporated both [¹⁴C]uridine and [³H]methyl groups and seemed to correspond to species of rRNA and their precursors. The other RNA fraction was less methylated or non-methylated and exhibited sedimentation coefficients distributed along a continuous 8–30 % sucrose density gradient. At least part of the latter type of RNA very probably was mRNA, but much of it must conespond to a different RNA similar to that recently described in HeLa cells by P enman , V esco and P enman (1968).
We also found that labelled 185 and 285 rRNA components began leaving the nucleus for the cytoplasm within 24 to 33 min after the radioactive precursors had been injected, and, in the cytoplasmic fraction, the patterns of incorporation for [¹⁴C]uridine and [³H]-methyl groups were similar for the 18S and 28S rRNA components. We estimate that in this fraction of rat brain the 18S rRNA component was 1·4 times more methylated than the 28S component. We also detected a lower sedimentation coefficient for the non- or slightly methylated, species of soluble RNA found in the cytoplasmic fraction.     

DEVELOPMENT OF THE UPTAKE AND STORAGE OF L-[3H]NOREPINEPHRINE IN THE RAT BRAIN

The uptake and storage of L-[³H]norepinephrine at various stages of development was examined in homogenates of rat brain. For the adult animal, active uptake accounted for 80 per cent of the total uptake. At 14 days of gestation, no active uptake was demonstrable At 18 days of gestation, saturable uptake of L-[³H]norepinephrine with a K_m of 3 × 10 ^?7m was first demonstrable; the K_m value did not vary during subsequent development. The V_max. of uptake increased five-fold between 18 days of gestation and 28 days postnatally, at which stage it was the same as the adult value. The development of saturable uptake paralleled but preceded the increase in endogenous norepinephrine. When homogenates were incubated with l -[³H]norepinephrine and subjected to centrifugation on linear sucrose gradients, there was a peak of tritium in the synaptosomal fractions; the magnitude of the peak increased with maturation of the brain. The increase in the peak of tritium paralleled the increase in particulate LDH activity and was distinct from the peak of MAO activity. Desipramine, a compound that blocks the initial uptake of norepinephrine, first exhibited inhibition of uptake at 19 days of gestation; the degree of inhibition did not vary during subsequent development. In contrast, reserpine, a compound which inhibits the intra-neuronal storage of norepinephrine, exhibited a progressive increase of inhibition with maturation of the brain at and subsequent to 19 days of gestation.     

Heterogeneity of [3H]Dipyridamole Binding to CNS Membranes: Correlation with [3H]Nitrobenzylthioinosine Binding and [3H]Uridine Influx Studies

The relationship between the nucleoside transport system and the nitrobenzylthioinosine-sensitive and -resistant [3H]dipyridamole binding sites was examined by comparing the characteristics of [3H]dipyridamole binding with those of [3H]nitrobenzylthioinosine binding and [3H]-uridine influx in rabbit and guinea pig cerebral cortical synaptosomes. Two distinct high-affinity synaptosomal membrane-associated [3H]dipyridamole binding sites, with different sensitivities to inhibition by nitrobenzylthioinosine, were characterized in the presence of 3-[(3-cholamidopropyl) dimethylammonio]-1-propanesulfonate (CHAPS, 0.01%) to prevent [3H]dipyridamole binding to glass tubes and filters. The nitrobenzylthioinosine-resistant [3H]-dipyridamole binding sites represented a greater proportion of the total membrane sites in guinea pig than in rabbit (40 vs. 10% based on inhibition studies). In rabbit, nitrobenzylthioinosine-sensitive [3H]dipyridamole binding (KD = 1.4 +/- 0.2 nM) and [3H]nitrobenzylthioinosine binding (KD = 0.30 +/- 0.01 nM) appeared to involve the same membrane site associated with the nitrobenzylthioinosine-sensitive nucleoside transporter. By mass law analysis, [3H]-dipyridamole binding in guinea pig could be resolved into two components based on sensitivity to inhibition by 1 microM nitrobenzylthioinosine. The nitrobenzylthioinosine-resistant [3H]dipyridamole binding sites were relatively insensitive to inhibition by all of the nucleoside transport substrates and inhibitors tested, with the exception of dipyridamole itself. In guinea pig synaptosomes, 100 microM dilazep blocked nitrobenzylthioinosine-resistant [3H]uridine transport completely but inhibited the nitrobenzylthioinosine-resistant [3H]dipyridamole binding component by only 20%. Furthermore, a greater percentage of the [3H]dipyridamole binding was nitrobenzylthioinosine resistant in guinea pig compared with rabbit, yet both species had a similar percentage of nitrobenzylthioinosine-resistant [3H]uridine transport.(ABSTRACT TRUNCATED AT 250 WORDS)     

Binding Sites for [3H]Glutamate and [3H]Aspartate in Human Cerebellum

Abstract The binding of [³H]aspartate and [³H]glutamate to membranes prepared from frozen human cerebellar cortex was studied. The binding sites differed in their relative proportions, their inhibition by amino acids and analogues, and by the effects of cations. A proportion (about 30%) of [³H]glutamate binding was to sites similar to those labelled by [³H]aspartate. An additional component of [³H]gluta-mate binding (about 50%) was displaced by quisqualate and aL-amino-3-hydroxy-5-methylisoxazole-4-propionic acid, and may represent a “quisqualate-preferring” receptor. Neither N-methyl-d-aspartic acid-sensitive nor dl-2-amino-4-phosphonobutyric acid-sensitive [³H]glutamate binding was detected.     

[1251]2α-BUNGAROTOXIN AND [3H]QUINUCLIDINYLBENZILATE BINDING IN CENTRAL NERVOUS SYSTEMS OF DIFFERENT SPECIES

Abstract— [¹²⁵I]Diiodo α-bungarotoxin ([¹²⁵I]₂BuTx) and [³H]quinuclidinylbenzilate ([³H]QNB) binding sites were measured in post-nuclear membrane fractions prepared from whole brains or brain regions of several species. Species studied included Drosophila melanogaster (fruit fly), Torpedo californiea (electric ray), Carassius auratus (goldfish), Ram pipiens (grass frog), Kana cutesheiana (bullfrog), Rattus norvegicus (rat, Sprague-Dawley), Mus muscalus (mouse, Swiss random, C58/J, LG/J), Oryctolagus cuniculus (rabbit, New Zealand Whitc), and Bos (cow). Acetyl-CoA: choline O -acetyltransferase (EC 2.3.1.6) levels were also determined in the post nuclear supernatants and correlated with the number of binding sites.
All species and regions except Drosophila had 16–150 fold more [³H]QNB binding sites than [¹²⁵I]₂BuTx binding sites. Brain regions with the highest levels of [¹²⁵I]₂BuTx binding were Drosophila heads (300 fmol/mg), goldfish optic tectum (80fmol/mg), and rat and mouse hippocampus (3040 fmol/mg). The highest levels of [³H]QNB binding were seen in rat and mouse caudate (1.3–1.6 pmol/mg). Lowest levels of [³H]QNB and [¹²⁵I]₂BuTx binding were seen in cerebellum. The utility of [¹²⁵I]₂BuTx and [³H]QNB binding as quantitative measures of nicotinic and muscarinic acetylcholine receptors in CNS is discussed.     

SUBCELLULAR DISTRIBUTION OF [3H]QUINUCLIDINYL BENZYLATE BINDING ACTIVITY IN VERTEBRATE RETINA AND ITS RELATIONSHIP TO OTHER CHOLINERGIC MARKERS

Abstract— The subcellular distribution of binding sites for tritiated quinuclidinyl benzylate ([³H]QNB) was studied in cow retina. Primary fractions showing higher specific activity than homogenate were P₂ (synaptosomal-mitochondrial) and P₃ (microsomal). P₂ was subfractionated on a Ficoll gradient with the P₂B subfraction exhibiting the greatest enrichment in [³H]QNB binding. Similar subfractionation of P₂ on a discontinuous sucrose gradient showed that fractions of particles banding in 0.8m and in the 0.8-1.0 m -sucrose interface also exhibit the greatest enrichment of [³H]QNB binding. When subjected to Scatchard analysis, this reaction shows a density of sites equal to 0.212-0.294 pmol per mg of protein. By plotting the apparent dissociation constant (K_D) values vs protein concentration a‘true’K_D value of 0.73 nM was obtained. Only one set of binding sites was found using three different concentrations of protein. The reaction was specificially antagonized by atropine (DI₅₀= 7 nM) and scopolamine (DI₅₀= 9 nM) whereas carbamylcholine and d-tubocurarine exhibited DI⁵⁰'s of 0.4 and 0.15 mM, respectively. For P₃ the binding of [³H]QNB is characterized by one set of binding sites with n_i= 0.250 pmol per mg of protein and an apparent K_D of 8.2 nM, and a DI₅₀ for atropine of 15 nM. The [³H]QNB binding sites showed a subcellular distribution similar to that of acetylcholinesterase and choline acetyltransferase. P₁ fractions accounted for 40–60% of the total activity of the three cholinergic markers. Purification of the crude P₁ yielded an additional fraction in which the cholinergic markers showed an enrichment with respect to homogenate and P₁. Synaptosomes isolated from this fraction exhibited the unusual ultrastructure expected from nerve endings in the outer synaptic layer of retina. The possible location of the muscarinic cholinergic transmitter system in the vertebrate retina is discussed.     

THE SUBCELLULAR DISTRIBUTION OF [14C]GABA AND [3H]DOPAMINE IN THE RETINA

Abstract— Rabbit retinae were homogenized in isotonic sucrose and subjected to differential and density gradient centrifugation. Preliminary electron microscopic examination of some of the fractions indicated that in addition to the subcellular particles usually observed in brain homogenates, the photoreceptor cells gave rise to several characteristic fragments. These included fragmented outer limbs, aggregations of mitochondria from the inner segments, and photoreceptor terminals. Unlike the synaptosomes formed from the conventional type of synapses in the retina, these photoreceptor terminals appeared to sediment mainly in the low speed crude nuclear pellet (P1).
Retinae were incubated with low concentrations of [¹⁴C]GABA and/or [³H]dopamine prior to subcellular fractionation and in these experiments the P2 pellet was further fractionated on sucrose density gradients. Analysis of the radioactivity in the fractions showed that labelled GABA was accumulated by osmotically sensitive particles which had the sedimentation characteristics of synaptosomes. The panicles accumulating [³H]dopamine appeared to belong to a different, slightly lighter, population than those accumulating [¹⁴C]GABA. It is tentatively suggested that the particles accumulating labelled GABA were synaptosomes because the fractions containing these particles also possessed most of the GAD activity of the gradient. In contrast, GABA-T and MAO activity was found in the dense fractions of the gradients usually associated with mitochondria.
When retinae were incubated with a high concentration of labelled GABA a'lighter'population of particles seemed to accumulate the amino acid than when a low external GABA concentration was used. These results suggest that the high and low affinity uptake processes for GABA in the retina may have different cellular sites.