Antidepressant binding site in a bacterial homologue of neurotransmitter transporters

Sodium-coupled transporters are ubiquitous pumps that harness pre-existing sodium gradients to catalyse the thermodynamically unfavourable uptake of essential nutrients, neurotransmitters and inorganic ions across the lipid bilayer. Dysfunction of these integral membrane proteins has been implicated in glucose/galactose malabsorption, congenital hypothyroidism, Bartter's syndrome, epilepsy, depression, autism and obsessive-compulsive disorder. Sodium-coupled transporters are blocked by a number of therapeutically important compounds, including diuretics, anticonvulsants and antidepressants, many of which have also become indispensable tools in biochemical experiments designed to probe antagonist binding sites and to elucidate transport mechanisms. Steady-state kinetic data have revealed that both competitive and noncompetitive modes of inhibition exist. Antagonist dissociation experiments on the serotonin transporter (SERT) have also unveiled the existence of a low-affinity allosteric site that slows the dissociation of inhibitors from a separate high-affinity site. Despite these strides, atomic-level insights into inhibitor action have remained elusive. Here we screen a panel of molecules for their ability to inhibit LeuT, a prokaryotic homologue of mammalian neurotransmitter sodium symporters, and show that the tricyclic antidepressant (TCA) clomipramine noncompetitively inhibits substrate uptake. Cocrystal structures show that clomipramine, along with two other TCAs, binds in an extracellular-facing vestibule about 11 A above the substrate and two sodium ions, apparently stabilizing the extracellular gate in a closed conformation. Off-rate assays establish that clomipramine reduces the rate at which leucine dissociates from LeuT and reinforce our contention that this TCA inhibits LeuT by slowing substrate release. Our results represent a molecular view into noncompetitive inhibition of a sodium-coupled transporter and define principles for the rational design of new inhibitors.     

Glycogenolysis induced by serotonin in brain: identification of a new class of receptor

Serotonin-containing neurones in brain have been proposed to have a role in the control of physiological mechanisms such as sleep, thermoregulation, pain perception and endocrine secretions as well as in the physiopathology of migraine or depressive illness. One difficulty in testing these possibilities lies in the scarcity of pharmacological agents able to interact selectively with the probably multiple classes of serotonin receptors in the central nervous system. Development of such agents would be facilitated by simple in vitro models in which biological responses to serotonin in mammalian brain could be quantified. Thus a serotonin-sensitive adenylate cyclase has been characterized in rat brain, but the response to serotonin is weak in newborn and practically absent in adult animals. In addition, two pharmacologically distinct classes of serotoninergic binding site have been identified using 3H-serotonin and 3H-spiperone as ligands, but their identification as receptors remains to be established. More recently, serotonin has been shown to stimulate phosphorylation of a neuronal protein in slices from the facial motor nucleus, although the receptors mediating this action were not characterized. We now report that serotonin stimulates glycogen hydrolysis in slices of cerebral cortex, that this action is mediated by a novel class of receptors and that tricyclic antidepressants are among the best competitive antagonists of the indolamine.     

Channel properties of an insect neuronal acetylcholine receptor protein reconstituted in planar lipid bilayers

A pentameric membrane protein composed of four types of polypeptide has been identified as the minimal structural unit responsible for the electrogenic action of acetylcholine on electrocytes and muscle cells. Because many populations of central and peripheral neurons also have nicotinic acetylcholine receptors (AChRs), considerable effort has recently gone into identifying the neuronal receptor. The central nervous tissue of insects contains very high concentrations of nicotinic AChRs, and we have recently purified an alpha-toxin binding protein, a putative AChR, from neuronal membranes of locusts. It is a component of high relative molecular mass, clearly composed of identical subunits, a structure predicted for an ancestral AChR protein. To verify that the purified polypeptides not only represent ligand binding sites but that they are indeed functional receptors, we have now reconstituted the isolated protein in a planar lipid bilayer. We show that in this system cholinergic agonists activate functional ion channels, that have properties comparable to those exhibited by the peripheral AChRs in vertebrates; thus, for the first time a functional acetylcholine receptor channel has been identified in nerve cells.     

Dihydroouabain is an antagonist of ouabain inotropic action

The Na+, K+-pump controls a wide variety of cellular systems and its inhibition by cardiac glycosides modifies important physiological functions and evokes several pharmacological effects (refs 1, 2 and refs therein). However, not all the actions of cardiac glycosides can be attributed to Na+, K+-pump inhibition and several observations show that, at low doses, cardiac glycosides stimulate the pump. It has been proposed that their positive inotropic effect could be the sum of two processes: the inhibition of the pump and a still unknown additional inotropic mechanism. In guinea pig heart, low doses of ouabain interact with high-affinity binding sites, which differ from the lower-affinity sites responsible for Na+, K+-pump inhibition. It has been suggested that ouabain interaction with these high-affinity sites could be responsible for the additional inotropic mechanism. The existence of two classes of ouabain-binding sites has been documented not only in guinea pig heart, but also in dog, rat and human heart. Dihydroouabain, a derivative of ouabain in which the lactone ring is saturated, is about 50-fold less potent than ouabain as an inhibitor of Na+, K+-pump and does not stimulate the pump at low doses. Its inotropic effect can be entirely accounted for by the inhibition of the pump. We have examined the pharmacological action of ouabain in the presence of dihydroouabain and report here that dihydroouabain reduces ouabain inotropic action but not Na+, K+-pump inhibition.     

Lithium inhibits adrenergic and cholinergic increases in GTP binding in rat cortex

Lithium is a unique drug with therapeutic as well as prophylactic value for both manic and depressive phases of manic-depressive illness. The precise mechanisms of its clinical efficacy remain unknown, but there are two main theories of its biochemical action. One proposes that lithium inhibits adrenergically activated adenylate cyclase function whereas the other suggests that it inhibits phosphatidyl inositol turnover, which is known to be activated by cholinergic agonists. Neither mechanism alone, however, can explain both the antimanic and antidepressant effects of lithium. Because of the pivotal role of G proteins in post-receptor information transduction, we have investigated the interaction of lithium with G protein function. Lithium at therapeutically efficacious concentrations completely blocked both adrenergic and cholinergic agonist-induced increases in [3H]GTP binding to membranes from rat cerebral cortex, in both in vitro and ex vivo experiments. The same lithium treatments also abolished guanine nucleotide modulation of agonist binding. Our findings suggest G proteins (Gs and Gi or Go) as the molecular site of action for both the antimanic and antidepressant effects of lithium.     

Acceleration of beta-receptor desensitization in combined administration of antidepressants and phenoxybenzamine

The delayed therapeutic effects of antidepressants (usually between 10 and 14 days) and of the tricyclic antidepressants in particular, are believed (on the basis of animal experiments) to lie in a progressive decrease of the sensitivity of cortical beta-adrenergic receptors. This is thought to be due to an increase in the synaptic concentration of noradrenaline, in turn accomplished by a decrease in the sensitivity of the presynaptic alpha 2 receptors which normally regulate noradrenaline secretion by a negative feedback mechanism. This model suggests that the desensitization of postsynaptic beta-receptors by antidepressants should be accelerated by the inhibition of the presynaptic alpha 2- adrenergic system, and we have indeed observed such an effect in preliminary studies with desipramine and phenoxybenzamine (PBZ) combined. We now show that the administration of either tricyclic or monoamine oxidase inhibitor antidepressants in combination with PBZ, an irreversible alpha-adrenergic blocker, accelerates and intensifies the desensitization of beta-adrenergic receptors. Our observations may have therapeutic implications.     

Cloning and expression of a complementary DNA encoding a bovine adrenal angiotensin II type-1 receptor

Angiotensin II elicits different responses which affect cardiovascular, neuronal and electrolyte transport regulation. To understand the mechanisms responsible for its various actions, the receptor for angiotensin II has long been sought, but numerous attempts to purify the receptor have been unsuccessful owing to its instability and low concentration. We report here the expression cloning of a complementary DNA encoding a bovine angiotensin II receptor to overcome these difficulties. The receptor cDNA encodes a protein of 359 amino-acid residues with a transmembrane topology similar to that of other G protein-coupled receptors. COS-7 cells transfected with the cDNA expressed specific and high-affinity binding sites for angiotensin II, angiotensin II antagonist and a non-peptide specific antagonist for type-1 receptor. Dithiothreitol inhibited ligand binding. The concentration of intracellular Ca2+ and of inositol-1,4,5-trisphosphate increased in the transfected COS-7 cells in response to angiotensin II or angiotensin III, indicating that this receptor is the type-1 receptor for angiotensin II. Northern blot analysis revealed that the messenger RNA for this receptor is expressed in bovine adrenal medulla, cortex and kidney.     

Similarity of teleocidin B and phorbol ester tumour promoters in effects on membrane receptors

The tumour promoter 12-O-tetradecanoyl-phorbol-13-acetate (TPA) and structurally related phorbol esters effect changes in avian and mammalian cell cultures that mimic transformation by oncogenic viruses or chemical carcinogens and the inhibition or induction of various types of differentiation (for review see refs 1--3). Unlike initiating carcinogens, which seem to act by binding covalently to cellular DNA, the primary site of action of the phorbol ester tumour promoters seems to be the cell membrane; indeed, specific high-affinity saturable receptors for phorbol esters have been identified in cell membranes. The recently discovered class of tumour promoters, the teleocidins, are as potent as TPA in the induction of ornithine decarboxylase in mouse skin, the inhibition of differentiation of Friend erythroleukaemia cells, the induction of HL-60 cell adhesion and the promotion of tumours on mouse skin. As the teleocidins are structurally unrelated to the phorbol esters, we set out to determine their effects on cell membranes and receptors. We found that in rodent cell cultures, teleocidin B and dihydroteleocidin B have effects similar to those of TPA and that, at nanomolar concentrations, teleocidin inhibits the binding of phorbol esters to cell-surface receptors, which suggests that the action of both classes of compounds may be mediated by the same or a similar receptor system.     

Characterization of receptors for human tumour necrosis factor and their regulation by gamma-interferon

Tumour necrosis factors, TNF-alpha and TNF-beta (previously called lymphotoxin), are the products of activated monocytes and lymphocytes, respectively, and both have recently been purified, sequenced and cloned by recombinant DNA methods, revealing 35% identity and 50% homology in the amino-acid sequence. Both proteins have been found to be specifically toxic to many tumour cells. Furthermore, it has been reported that various interferons are synergistic with TNF for anti-tumour effects in vitro, while activities attributed to the two proteins have also been shown to necrotize various tumours in vivo. We have now prepared 125I-labelled highly purified recombinant human TNF-alpha to study in detail its binding to the human cervical carcinoma cell line ME-180. Our results indicate that there is a single class of specific high-affinity receptors for TNF on this cell line which has a Kd of about 0.2 nM and an average of 2,000 receptor sites per cell. The binding of labelled TNF-alpha to these cells can be inhibited by both TNF-alpha and TNF-beta but not by gamma-interferon (IFN-gamma). However, preincubation of cells with IFN-gamma increases the total number of TNF receptors two to threefold without any significant change in the affinity constant. This is the first report that TNF-alpha and -beta share a common receptor and that the receptors can be up-regulated by interferon. Our results may explain previous observations regarding similar biological activities observed for these two cytotoxic proteins and also their synergistic action with interferons.     

Regulation of glutamate receptors by cations.

Current evidence suggests that glutamate is a major excitatory neurotransmitter in the mammalian central nervous system (CNS); particularly, glutamate excites most neurones in the CNS. Until recently this effect was widely used to study glutamate receptors and to distinguish them from those of other excitatory amino acids. The development of ligand binding studies for many neurotransmitters has facilitated the study of receptors at the molecular level and using these methods we recently reported the existence in hippocampal membranes of pharmacologically distinct sodium-dependent and sodium-independent glutamate binding sites, the former related to high-affinity uptake and the latter exhibiting several characteristics of postsynaptic receptor sites. We now report that, as with other neurotransmitters, several ions regulate the Na-independent binding of glutamate; the monovalent cations induce a decreased binding while certain divalent cations enhance this Na-independent binding. Additionally, since some of these effects appear to be irreversible, we propose that the regulation of glutamate binding by cations might account for the extremely long-lasting potentiation of synaptic responses found in the hippocampus following bursts of repetitive electrical stimulation (see ref. 9 for a review).     

Anatomical distributions of four pharmacologically distinct 3H-L-glutamate binding sites

Glutamate is thought to serve as a major excitatory neurotransmitter throughout the central nervous system (CNS); electrophysiological studies indicate that its action is mediated by multiple receptors. Four receptors have been characterized by their selective sensitivity to N-methyl-D-aspartate (NMDA), kainic acid (KA), quisqualic acid (QA) and 2-amino-4-phosphonobutyric acid (APB). Electrophysiological evidence indicates that these receptors are all present in the rat hippocampus and that the anatomically discrete synaptic fields within the hippocampus exhibit differential sensitivity to the selective excitatory amino acid agents. Thus, we have used the hippocampus as a model system to investigate possible subpopulations of 3H-L-glutamate binding sites. By using quantitative autoradiography, the pharmacological specificity of 3H-L-glutamate binding in discrete terminal fields was determined. We report here that there are at least four distinct classes of 3H-L-glutamate binding sites which differ in their anatomical distribution, pharmacological profile and regulation by ions. Two of these sites seem to correspond to the KA and NMDA receptor classes, and a third site may represent the QA receptor. The fourth binding site does not conform to present receptor classifications. None of these binding sites corresponds to the major glutamate binding site observed in biochemical studies.     

Correlation between benzodiazepine receptor occupation and anticonvulsant effects of diazepam.

The benzodiazepines are potent anticonvulsants for a wide variety of experimental and clinical seizure disorders. The demonstration of saturable, high-affinity and stereospecific binding sites for the benzodiazepines in the mammalian central nervous system suggests the presence of pharmacological receptors mediating the anticonvulsant properties of these compounds. The good correlation between the anticonvulsant potencies of a series of benzodiazepines and their ability to inhibit 3H-diazepam binding in vitro further supports this hypothesis, but evidence for a direct interaction between benzodiazepines and their receptors, and a subsequent inhibition of seizure activity (or elevation of seizure threshold) is lacking. Recent reports from our laboratory and others have demonstrated the feasibility of labelling benzodiazepine receptors in vivo following parental administration of tritiated benzodiazepine. This technique permits one to study the relationship between the anticonvulsant activity of the benzodiazepines in vivo and the number of 'drug-occupied' receptors in vitro. We now report that there is an excellent correlation between benzodiazepine receptor occupancy by diazepam and protection against pentylenetetrazol-induced seizures. Furthermore, these results demonstrate that only a small fraction of benzodiazepine receptors need be occupied to produce a complete anticonvulsant effect.     

Effect of age on benzodiazepine-induced behavioural convulsions in rats

The benzodiazepines are a class of drugs used to alleviate anxiety. As such they constitute one of the most commonly prescribed compounds, due in part to their efficacy and safety. The physiological effect of these drugs is probably through interactions with a low affinity benzodiazepine binding site and two (types 1 and 2) higher affinity sites. The ontogenesis of these latter two binding sites in the rat differs, with the type 2 binding site being predominant at birth and the type 1 binding site increasing in number after the second week after birth. The differential development of these two receptor types is important because the immature organism may have different physiological and behavioural responses from the adult. Here we demonstrate an important difference: that a prototypic benzodiazepine, chlordiazepoxide, and a water-soluble benzodiazepine, flurazepam, produce behavioural convulsions in the preweanling rat. The convulsions are antagonized by the benzodiazepine blocker Ro-15-1788. The triazolopyridizine CL-218872, specific to the type 1 receptor, does not share this action. We suggest that this paradoxical convulsant effect of chlordiazepoxide and flurazepam is due to activation of the type 2 receptor in the absence of the type 1 receptor in the immature rat.     

Dopaminergic D-3 binding sites are not presynaptic autoreceptors

Postsynaptic dopamine (DA) receptors have been classified biochemically and pharmacologically into two types: D-1 receptors mediate adenylate cyclase stimulation, demonstrating micromolar affinity for DA and butyrophenone antagonists; D-2 receptors mediate adenylate cyclase inhibition, demonstrating nanomolar affinity for DA and butyrophenone antagonists. D-1 receptors are labelled by 3H-thioxanthene antagonists, while D-2 receptors are labelled by both 3H-agonists and all 3H-antagonists. A third class of dopaminergic binding site, termed D-3, represents high-affinity 3H-agonist binding sites demonstrating low, micromolar, affinity for butyrophenones. In the rat striatum, D-3 sites were decreased 50% by 6-hydroxydopamine (6-OHDA) lesions of the nigrostriatal DA pathway, suggesting that such D-3 binding labels presynaptic DA autoreceptors on nigrostriatal terminals. However, nigrostriatal denervation produces a concomitant depletion of striatal DA. Here we demonstrate that a reserpine-induced depletion of DA produces a decrease in D-3 binding comparable to that seen with nigrostriatal denervation, independent of presynaptic terminal degeneration. This loss in binding, or that caused by 6-OHDA lesions, is recovered by preincubating the striatal membranes with DA or with the supernatant from control striatal membrane preparations. We therefore suggest that the loss of D-3 binding following 6-OHDA lesions results from the depletion of endogenous DA rather than the degeneration of terminals and their putatively associated autoreceptors.     

beta-Carboline-3-carboxylic acid ethyl ester antagonizes diazepam activity

Analogous to the progression of events in the opiate receptor-enkaphalin area, the first reports that benzodiazepines have selective and specific high-affinity binding sites in brain have stimulated a search for the endogenous 'ligand' or substance that might normally act at these sites. Braestrup and co-workers have extracted from human urine a gamma-fraction (ref. 10) which they have recently identified as beta-carboline-3-carboxylic acid ethyl ester (beta CEE). They reported that this substance is extremely potent in displacing 3H-diazepam from brain binding sites and proposed that a beta-carboline-3-carboxylic acid derivative might, in part, be the endogenous ligand for the brain benzodiazepine receptor. We have examined several synthetically derived beta-carboline-3-carboxylic acid analogues and now present data obtained from testing only the beta CEE described by Braestrup et al. In addition to confirming these workers' observation that this compound is a potent displacer of 3H-diazepam from brain tissue, our pharmacological data indicate that beta CEE has activity that is opposite to, rather than similar to, that of diazepam.     

Importance of a novel GABAA receptor subunit for benzodiazepine pharmacology

Neurotransmission effected by GABA (gamma-aminobutyric acid) is predominantly mediated by a gated chloride channel intrinsic to the GABAA receptor. This heterooligomeric receptor exists in most inhibitory synapses in the vertebrate central nervous system (CNS) and can be regulated by clinically important compounds such as benzodiazepines and barbiturates. The primary structures of GABAA receptor alpha- and beta-subunits have been deduced from cloned complementary DNAs. Co-expression of these subunits in heterologous systems generates receptors which display much of the pharmacology of their neural counterparts, including potentiation by barbiturates. Conspicuously, however, they lack binding sites for, and consistent electrophysiological responses to, benzodiazepines. We now report the isolation of a cloned cDNA encoding a new GABAA receptor subunit, termed gamma 2, which shares approximately 40% sequence identity with alpha- and beta-subunits and whose messenger RNA is prominently localized in neuronal subpopulations throughout the CNS. Importantly, coexpression of the gamma 2 subunit with alpha 1 and beta 1 subunits produces GABAA receptors displaying high-affinity binding for central benzodiazepine receptor ligands.     

Regulation of glutamate receptor binding by the cytoskeletal protein fodrin

The erythrocyte cytoskeleton, which consists primarily of a meshwork of spectrin and actin, controls cell shape and the disposition of proteins within the membrane. Proteins similar to spectrin have recently been found in diverse cells and tissues, and it is possible that they mediate the capping of cell-surface receptors, although this has not been demonstrated directly. In neurones, the spectrin-like protein fodrin lines the cortical cytoplasm and may link actin filaments to the membrane. Fodrin has been hypothesized to regulate the number of receptor binding sites on neuronal membranes for the putative neurotransmitter L-glutamate. Micromolar calcium concentrations activate the thiol protease calpain I, induce fodrin degradation and more than double the density of glutamate binding sites; these effects are all blocked by thiol protease inhibitors. We have now used specific antibodies to examine further the role of fodrin proteolysis in regulating glutamate receptors. We report that fodrin antibodies block the fodrin degradation and increase in glutamate binding normally induced by calcium, and so provide direct evidence for control of membrane receptors by a non-erythroid spectrin.     

Gonadotropin-releasing hormone analogue binds to luteal cells and inhibits progesterone production.

Although gonadotropin-releasing hormone (GnRH) is believed to mediate the hypothalamic control of pituitary gonadotropin secretion, continuous or repeated administration of GnRH or its agonist analogues has been shown to cause paradoxical antifertility effects in several species, including primates. GnRH-induced interruption of reproductive cycles and pregnancy is associated with diminished progesterone production, implying defective function of the corpus luteum. These luteolytic effects have been attributed to the well recognized desensitising actions of elevated luteinising hormone (LH) levels on ovarian LH receptors and steroidogenesis, subsequent to GnRH-induced gonadotropin release from the anterior pituitary. However, treatment with high doses of exogenous LH did not cause suppression of serum progesterone levels during early pregnancy in rats, whereas a highly active GnRH analogue was effective in this regard. These observations suggested that GnRH and its agonist analogues, given in high or sustained doses, can exert a direct action on the ovary that is independent of the pituitary. This hypothesis was further supported by the ability of GnRH and its agonists to inhibit human chorionic gonadotropin (HCG)-induced ovarian and uterine weight gain in hypophysectomised rats and to delay the onset of puberty in intact female rats. Also, GnRH and its agonist analogues have recently been shown to inhibit steroidogenesis induced by follicle-stimulating hormone (FSH) in cultured granulosa cells, confirming the direct action of such peptides on the ovarian follicle. The marked inhibitory effects of GnRH and its agonists on corpus luteum function suggest that these compounds could exert direct actions by binding to specific receptors on luteal cells. The present experiments, which examine the effects of GnRH agonists on luteal steroidogenesis, demonstrate the existence of such actions and their mediation by specific high-affinity receptor sites present in luteal cell membranes.     

A site for the potentiation of GABA-mediated responses by benzodiazepines

The benzodiazepines have been well characterised as minor tranquillizers and attempts to explain their unique spectrum of activity have included suggestions that they may interact with a variety of neurotransmitter systems. Recently, a possible interaction with the gamma-aminobutyric acid (GABA) system has received most attention. Benzodiazepines potentiate the actions of both synaptically released and exogenously administered GABA on mammalian neuronal preparations but the site of action within the GABA response mechanism has not been determined. Binding studies suggest that benzodiazepines combine with highly specific sites in the neuronal membrane and that these sites have some indirect association with GABA receptors. To investigate this association further in a functioning GABA system, quantitative studies have been made in vitro on neuronal depolarisations mediated by GABA receptor activation. Evidence has already been presented that bicuculline is most probably a competitive antagonist at the GABA receptor while picrotoxin acts as an antagonist at a separate site. Here flurazepam is shown to attenuate preferentially the action of picrotoxin rather than bicuculline and a model is suggested for the site of action of these drugs within the GABA response mechanism.