Chronic ethanol exposure delays the 'developmental switch' of the NMDA receptor 2A and 2B subunits in cultured cerebellar granule neurons.

Chronic ethanol treatment of cultured neurons from various brain areas has been found to increase NMDA receptor function and to alter the levels of some NMDA receptor subunit proteins. Because the cultured neurons are exposed to ethanol during a period when the NMDA receptor is undergoing developmental changes in subunit expression, we wished to determine whether ethanol treatment alters this developmental pattern. We found that 3 days of treatment of cerebellar granule neurons with ethanol, which was previously reported to increase NMDA receptor function, resulted in a delay in the 'developmental switch' of the NR2A and NR2B subunits, i.e. the developmental decrease in NR2B and increase in NR2A protein expression. As a result, the level of NR2B was higher, and that of NR2A was lower, in the ethanol-treated cells than in control cells. Cross-linking experiments showed that the changes in total receptor subunit proteins levels were reflected in cell-surface expressed proteins, indicating changes in the amount of functional receptors. These results were confirmed by a higher potency of glycine at the NMDA receptor in the ethanol-treated cells, as determined by NMDA/glycine-induced increases in intracellular Ca(2+). The results suggest that the mechanism by which ethanol alters NMDA receptor expression in cultured neurons, where receptors are undergoing development, differs from the mechanism of ethanol's effect on NMDA receptors in adult brain. Changes in the proportion of NR2A and NR2B subunits may contribute to effects of ethanol on neuronal development.

Chronic ethanol exposure attenuates the anti-apoptotic effect of NMDA in cerebellar granule neurons.

Ethanol, added to primary cultures of cerebellar granule neurons simultaneously with NMDA, was previously shown to inhibit the anti-apoptotic effect of NMDA. The in vitro anti-apoptotic effect of NMDA is believed to mimic in vivo protection against apoptosis afforded by innervation of developing cerebellar granule neurons by glutamatergic mossy fibers. Therefore, the results suggested that the presence of ethanol in the brain at a critical period of development would promote apoptosis. In the present studies, we examined the effect of chronic ethanol exposure on the anti-apoptotic action of NMDA in cerebellar granule neurons. The neurons were treated with ethanol in vitro for 1-3 days in the absence of NMDA. Even after ethanol was removed from the culture medium, as ascertained by gas chromatography, the protective effect of added NMDA was significantly attenuated. The decreased anti-apoptotic effect of NMDA was associated with a change in the properties of the NMDA receptor, as indicated by a decrease in ligand binding, decreased expression of NMDA receptor subunit proteins, and decreased functional responses including stimulation of increases in intracellular Ca(2+) and induction of brain-derived neurotrophic factor expression. The latter effect may directly underlie the attenuated protective effect of NMDA in these neurons. The results suggest that ethanol exposure during development can have long-lasting effects on neuronal survival. The change in the NMDA receptor caused by chronic ethanol treatment may contribute to the loss of cerebellar granule neurons that is observed in animals and humans exposed to ethanol during gestation.

Novel structure having antagonist actions at both the glycine site of the N-methyl-D-aspartate receptor and neuronal voltage-sensitive sodium channels: biochemical, electrophysiological, and behavioral characterization.

A novel series of N-substituted 4-ureido-5,7-dichloro-quinolines were synthesized to contain pharmacophores directed at voltage-sensitive sodium channels (VSNaCs) and N-methyl-D-aspartate (NMDA) receptors. These compounds were shown to act in a use-dependent manner as antagonists of VSNaCs and to act as selective competitive antagonists at the strychnine-insensitive glycine recognition site of NMDA receptors. These agents had little or no effect on alpha-adrenergic receptors, other glutamate receptors, or sites other than the glycine site on the NMDA receptor, and did not block voltage-sensitive calcium channels in vitro. In vivo, the compounds were active in preventing or reducing the signs and symptoms of neurohyperexcitability and had anxiolytic properties. Unlike benzodiazepines, N-substituted 4-ureido-5, 7-dichloro-quinolines showed little interaction with the sedative effects of ethanol, but were effective in controlling ethanol withdrawal seizures. The combined actions of these compounds on VSNaCs and NMDA receptors also impart properties to these compounds that are important for preventing and reducing excitotoxic neurodegeneration, but these compounds lack the undesirable side effects of other agents used for these purposes.

Ethanol sensitivity of NMDA receptor function in developing cerebellar granule neurons.

The mechanism by which ethanol inhibits the function of the NMDA subtype of glutamate receptor has not been elucidated. One possibility that has been suggested is that NMDA receptor subunit composition influences the sensitivity of the receptor to ethanol. We have taken advantage of developmental changes in subunit composition of the NMDA receptor in cultured neurons to examine possible changes in the effect of ethanol. We found an increase in expression of the NR2A subunit, and a decrease in expression of the NR2B subunit of the NMDA receptor in primary cultures of cerebellar granule neurons over time in culture, with no significant change in NR1 expression. This change in NR2 subunit expression was associated with the expected changes in functional properties of the NMDA receptor (measured as the NMDA-induced increase in intracellular Ca2+), i.e., ifenprodil sensitivity and glycine potency were higher when there was a relatively greater proportion of NR2B in the cultured neurons. However, the potency of ethanol to inhibit NMDA receptor function was lower when there was a greater proportion of NR2B subunits. Previous studies showed that ethanol inhibition of NMDA receptor function in cerebellar granule neurons resulted from an ethanol-induced decrease in potency of the co-agonist, glycine, and that this effect of ethanol was blocked by inhibitors of protein kinase C. Our current results suggest that the lower potency of ethanol to inhibit the response of NMDA receptors when cerebellar granule neurons are expressing a greater proportion of NR2B subunits is a result of the higher affinity of the NMDA receptors for endogenous levels of glycine at this point in time.

Gangliosides reduce the development of ethanol dependence without affecting ethanol tolerance.

Mice given an ethanol-containing liquid diet, as their sole source of nutrients and fluid, rapidly developed functional tolerance to and physical dependence on ethanol. The presence of physical dependence was demonstrated by measured signs of central nervous system hyperexcitability upon withdrawal of ethanol. The withdrawal hyperexcitability, which included tremors, handling-induced seizures and spontaneous clonic/tonic seizures, was more pronounced when mice consumed the ethanol-containing diet for 7 days, compared with 5 days. Daily treatment of the animals with either a ganglioside mixture (extracted bovine brain gangliosides, 250 or 500 mg/kg i.p.) or ganglioside GM1 (100 mg/kg i.p.) for the terminal two-thirds of the ethanol administration period resulted in a significant reduction in the ethanol withdrawal signs. On the other hand, tolerance to the hypnotic action of ethanol, tested 30 hr after withdrawal of ethanol, was unaffected by ganglioside treatment. Ganglioside GM1 given i.c.v. at a daily dose of 10 micrograms during the ethanol ingestion period was as effective as 100 mg/kg GM1 given i.p. in reducing signs of ethanol withdrawal. The daily administration of gangliosides during the feeding of the ethanol diet did not alter the animals' ethanol consumption, intoxication or blood ethanol levels at the time of ethanol withdrawal. A single dose of GM1 given either i.p. or i.c.v. 16 hr before withdrawal produced no effect on the measured ethanol withdrawal signs. Our prior work and the work of others has demonstrated a relationship between up-regulation of N-methyl-D-aspartate receptor numbers in brain and the manifestation of ethanol withdrawal signs. Daily administration of GM1, during the last 5 days of a 7-day period of ethanol ingestion, prevented the up-regulation of N-methyl-D-aspartate receptors in the hippocampus and reduced the ethanol withdrawal signs. Our data demonstrate that the daily administration of gangliosides during the period of ethanol consumption may prevent the development of ethanol physical dependence, while leaving ethanol tolerance intact.

Regional and subunit specific changes in NMDA receptor mRNA and immunoreactivity in mouse brain following chronic ethanol ingestion.

Chronic ethanol treatment of mice has been shown to result in increased binding of dizocilpine and glutamate to hippocampal NMDA receptors. These changes were suggested to reflect an increase in NMDA receptor number that may underlie certain signs of the ethanol withdrawal syndrome. However, there was no change in binding of a competitive NMDA receptor antagonist, or of ligand binding to the glycine co-agonist site on the receptor after chronic ethanol treatment. Differential changes in the binding of particular ligands at the NMDA receptor suggested the possibility that chronic ethanol ingestion might selectively affect the expression of particular NMDA receptor subunits. Our current work demonstrates that chronic ethanol ingestion by mice, which results in the generation of physical dependence, also produces increases in the NMDA receptor NR1 subunit protein in the hippocampus and cerebellum (approximately 50% and 95%, respectively), and produces increases in the NR2A subunit protein in the hippocampus and cortex (approximately 25% and 40%, respectively). However, the mRNA levels for these subunits were not increased in the respective brain areas by the same ethanol treatment. The changes in NMDA receptor subunit expression in discrete areas of the brain may contribute to the previously observed changes in ligand binding and, possibly, signs of ethanol withdrawal.

Mechanism of ethanol inhibition of NMDA receptor function in primary cultures of cerebral cortical cells.

Ethanol is a potent inhibitor of the function of the N-methyl-D-aspartate (NMDA) subtype of glutamate receptor in various neuronal preparations. In primary cultures of cerebellar granule cells, ethanol was suggested to interact with the glycine co-agonist site of the receptor by a mechanism involving protein kinase C. In the present study, the interaction of ethanol with various sites on the NMDA receptor was examined in primary cultures of cerebral cortical cells from embryonic rats. NMDA receptor function was determined by measuring increases in intracellular Ca2+ with fura-2 fluorescence. Ethanol inhibited the function of the NMDA receptor in cerebral cortical cells, but in contrast to the results in cerebellar granule cells, phorbol ester treatment did not inhibit the NMDA response, and ethanol did not alter the effect of glycine on NMDA receptor function. Ethanol also did not affect inhibition of the NMDA response by Mg2+ or dizocilpine. The results support the hypothesis that the mechanism of ethanol inhibition of NMDA receptor function can vary in neurons from different brain regions.

The 71 kDa glutamate-binding protein is increased in cerebellar granule cells after chronic ethanol treatment.

Besides the N-methyl-D-aspartate (NMDA) receptor proteins NR1 and NR2, another complex of proteins which has been shown to contain ligand-binding sites characteristic of NMDA receptors is expressed in cerebellar granule cells. One of the proteins in the latter complex is the 71 kDa glutamate-binding protein (GBP). To determine the role of the GBP in the response to NMDA, primary cultures of cerebellar granule cells were treated with an antisense oligonucleotide complementary to mRNA for this protein. This treatment substantially reduced both mRNA and protein levels of the GBP, as well as the response of the cells to NMDA, measured as an increase in intracellular Ca2+ with fura-2 fluorescence. The antisense oligonucleotide treatment did not alter the Ca2+ responses to KC1 or kainate. Chronic ethanol exposure has previously been shown to increase NMDA receptor function and the density of binding sites for the NMDA receptor channel blocker, dizocilpine, in cerebellar granule cells. Chronic exposure of the cells to 100mM ethanol is now shown to result in significant increases in mRNA and protein levels for the GBP (45% and 100%, respectively). Ethanol treatment did not affect mRNA levels for NR1 or NR2A, caused only a small increase (20%) in protein levels for NR1, and resulted in a decrease (30%) in NR2A protein. Although a role of the NMDA receptor NR1/NR2 subunits cannot be ruled out, these results are compatible with the hypothesis of involvement of the GBP in the chronic ethanol-induced increase in NMDA receptor function in cerebellar granule cells.

Attenuation of glutamate-induced neurotoxicity in chronically ethanol-exposed cerebellar granule cells by NMDA receptor antagonists and ganglioside GM1.

Ethanol, acutely, is a potent inhibitor of the function of the N-methyl-D-aspartate (NMDA) subtype of glutamate receptor. After chronic exposure of animals to ethanol, however, the NMDA receptor in brain is upregulated. This upregulation is associated with the occurrence of ethanol withdrawal seizures. When cultured cerebellar granule neurons are exposed chronically to ethanol, the resulting upregulation of NMDA receptor function renders the cells more susceptible to glutamate-induced neurotoxicity. The present studies show that chronic ethanol exposure produces an increase in NMDA receptor number in the cells, measured by ligand binding to intact cells. Glutamate-induced excitotoxicity, both in control and ethanol-exposed cells, is blocked by the same NMDA receptor antagonists previously shown to block ethanol withdrawal seizures in animals. In addition, glutamate neurotoxicity is blocked by acute (2-hr) pretreatment of cells with ganglioside GM1 or by chronic (3 days) treatment with the ganglioside. Acute ganglioside treatment does not interfere with the initial rise in intracellular calcium caused by glutamate, whereas this response is downregulated after chronic ganglioside treatment. These results suggest that therapeutic agents can be developed to block both ethanol withdrawal signs and the neuronal damage that accompanies ethanol withdrawal. Furthermore, chronic ganglioside treatment during ethanol exposure has the potential to prevent changes in the NMDA receptor that lead to withdrawal seizures and enhanced susceptibility to excitotoxicity.

Protein kinase C activation attenuates N-methyl-D-aspartate-induced increases in intracellular calcium in cerebellar granule cells.

Activation of the N-methyl-D-aspartate (NMDA) subtype of glutamate receptor increases levels of intracellular calcium and can lead to stimulation of protein kinase C activity. Several reports have demonstrated that stimulation of protein kinase C can, in turn, increase electrophysiological responses to NMDA in certain cells or in oocytes expressing certain NMDA receptor subunits. In the present study, the effects of protein kinase C activation on NMDA receptor-mediated increases in intracellular Ca2+ level were investigated in primary cultures of rat cerebellar granule cells using fura-2 fluorescence spectroscopy. Pretreatment of the cells with the protein kinase C activator phorbol 12-myristate 13-acetate (PMA), but not the inactive analogue 4 alpha-phorbol 12-myristate 13-acetate, inhibited NMDA-induced increases in intracellular Ca2+ levels. Coincubation of cells with PMA and the kinase inhibitor staurosporine or calphostin C blocked the PMA effect. The potency of NMDA was reduced twofold, and the potency of the NMDA receptor co-agonist, glycine, to enhance the response to NMDA was decreased fourfold by pretreatment of cells with PMA. The effect on glycine was mimicked by pretreatment with okadaic acid, a protein phosphatase inhibitor. PMA treatment did not significantly alter Mg2+ inhibition of the NMDA response but decreased the potency of the competitive antagonist CGS-19755. These data suggest that, in cerebellar granule cells, the function of the NMDA receptor may be subject to feed-back inhibition by protein kinase C stimulation. Under physiological conditions, this inhibition may result from a decreased effectiveness of the endogenous co-agonists, glutamate and glycine.

Involvement of protein kinase C in ethanol-induced inhibition of NMDA receptor function in cerebellar granule cells.

Ethanol inhibits N-methyl-D-aspartate (NMDA)-stimulated increases in intracellular Ca2+ in cerebellar granule cells apparently by reducing the potency of glycine to act as a co-agonist at the NMDA receptor. The inhibitory effect of ethanol on the NMDA response in these cells can be reversed not only by a high concentration of glycine, but also by the protein kinase inhibitors, staurosporine and calphostin C. We previously showed that activation of protein kinase C in cerebellar granule cells also resulted in inhibition of the NMDA response, and in decreased potency of glycine at the NMDA receptor. Furthermore, the inhibitory effects of ethanol and protein kinase C activation are not additive. These results suggest a role for protein kinase C in ethanol inhibition of NMDA responses in cerebellar granule cells. In contrast, although ethanol can inhibit the response to kainate in these cells in a "competitive" manner, this response is not affected by activation of protein kinase C.

Ethanol inhibition of NMDA receptor function in primary cultures of rat cerebellar granule cells and cerebral cortical cells.

Ethanol inhibits the function of the N-methyl-D-aspartate (NMDA) subtype of glutamate receptor in various neuronal systems, but the mechanism of the inhibition has not been elucidated. Previous work, using primary cultures of rat cerebellar granule cells, showed that both exposure to alcohol and activation of protein kinase C (PKC) by the phorbol ester PMA reduced the potency of the co-agonist, glycine, to enhance NMDA receptor function (measured as an increase in intracellular Ca2+), resulting in inhibition of the NMDA response at low glycine concentrations. Inhibition of NMDA receptor function by PMA and ethanol could also be overcome by PKC antagonists, implicating PKC in the inhibitory effect of ethanol. We have now compared the effects of ethanol and PKC activation of NMDA receptor function in primary cultures of rat cerebral cortical cells. The receptor in these cells was much less sensitive to ethanol inhibition, and the inhibition was not overcome by high concentrations of glycine. Furthermore, PMA treatment resulted in an increased response to NMDA at low glycine concentrations. The results indicate that PKC does not mediate ethanol inhibition of NMDA receptor function in cerebral cortical cells, and that the mechanism of ethanol inhibition can vary among brain regions and/or cell types. Possible determinants of the differing mechanisms of ethanol's actions include the subunit composition of the NMDA receptor and/or the isoforms of PKC present in the different cells.

Radioligand binding to the N-methyl-D-aspartate receptor/ionophore complex: alterations by ethanol in vitro and by chronic in vivo ethanol ingestion.

The effects of ethanol on the binding of [3H]MK-801, [3H]L-glutamate, [3H]glycine and [3H]CGS 19755 to the N-methyl-D-aspartate receptor were determined in membranes from mouse cortex and hippocampus. Under equilibrium conditions, ethanol in vitro (100 mM) did not alter the apparent affinity or binding site density for any of these ligands. However, in the presence of glutamate and the selective glycine antagonist, 5,7-dichlorokynurenic acid, ethanol inhibited the non-equilibrium binding of [3H]MK-801. This inhibition could be reversed in a time- and concentration-dependent manner by addition of glycine. These data suggest that ethanol may inhibit N-methyl-D-aspartate receptor-mediated responses by altering the kinetics of channel activation. Chronic in vivo ethanol ingestion by mice, that resulted in tolerance to and physical dependence on ethanol, produced an increased density of hippocampal [3H]MK-801 and [3H]L-glutamate binding sites, but not [3H]glycine or [3H]CGS 19755 binding sites. It is possible that chronic ethanol ingestion may influence the subunit composition of the NMDA receptor complex.

Comparison of the effects of the uncompetitive N-methyl-D-aspartate antagonist (+-)-5-aminocarbonyl-10,11-dihydro-5H-dibenzo[a,d] cyclohepten-5,10-imine (ADCI) with its structural analogs dizocilpine (MK-801) and carbamazepine on ethanol withdrawal seizures.

The ability of [(+-)-5-aminocarbonyl-10,11-dihydro-5H-di-benzo [a,d]cyclohepten-5,10-imine (ADCI) and its structural analogs dizocilipine (MK-801) and carbamazepine to block ethanol withdrawal seizures was tested in mice made physically dependent upon ethanol. Three injections of either ADCI (ranging from 1.0-10.0 mg/kg), dizocilpine (ranging from 0.1-1.0 mg/kg) or carbamazepine (ranging from 17-50 mg/kg) were administered during the first 7 hr of ethanol withdrawal. The severity of ethanol withdrawal seizures was rated during the first 11 hr of withdrawal and again at 24 hr after withdrawal of ethanol. ADCI and dizocilpine suppressed the severity and occurrence of the withdrawal seizures in a dose-dependent fashion, whereas carbamazepine was ineffective in blocking the withdrawal seizures. The relative potencies of dizocilpine, ADCI and carbamazepine in suppressing ethanol withdrawal seizures corresponded with the relative potencies of the compounds in displacing [3H]dizocilpine from mouse cortical membrane preparations. These findings are consistent with the suggestion that blockade of N-methyl-D-aspartate-mediated neurotransmission is an effective treatment for decreasing ethanol withdrawal seizures. ADCI also blocked the occurrence of withdrawal-associated whole body tremors, whereas dizocilpine and carbamazepine were ineffective in blocking the tremors. The doses of ADCI, dizocilpine and carbamazepine that resulted in motor incoordination on an accelerating rotarod task were determined in groups of naive mice. Dizocilpine in doses as low as 0.3 mg/kg produced a decreased ability to remain on the rotarod, whereas ADCI up to 30 mg/kg did not affect rotarod performance.(ABSTRACT TRUNCATED AT 250 WORDS)

Pharmacological profile of NPC 12626, a novel, competitive N-methyl-D-aspartate receptor antagonist.

The novel compound 2-amino-4,5-(1,2-cyclohexyl)-7-phosphonoheptanoic acid (NPC 12626) was evaluated for activity in a variety of tests associated with receptors for excitatory amino acids. NPC 12626 failed to inhibit the specific binding of RS-[3H] amino-3-hydroxy-5-methyl-isoxazole-4-propionic acid or [3H] kainic acid to brain membranes in vitro but displaced both agonist and antagonist binding to N-methyl-D-aspartic acid (NMDA) receptors. Like cis-(+/-)-3-(2-carboxypiperazine-4-yl)propyl-1-phosphonic acid, NPC 12626 competitively blocked NMDA-induced enhancement of [3H]-1-thienylcyclohexyl)piperidine binding. In the voltage-clamped frog oocyte expression system, NPC 12626 was a competitive inhibitor of NMDA-evoked inward current with a pA2 of 6.24. After both i.c.v. or i.p. administration, NPC 12626 was a potent anticonvulsant in the pentylenetetrazol, maximal electroshock and NMDA seizure models. Furthermore, low doses (25 mg/kg) of NPC 12626 given i.v. were effective in preventing damage to the CA1 region of hippocampus in the gerbil model of global ischemia. Unlike the noncompetitive NMDA antagonist, phencyclidine, but like cis-(+/-)-3-(2-carboxypiperazine-4-yl)propyl-1-phosphonic acid and pentobarbital, NPC 12626 only partially substituted for phencyclidine in a drug discrimination study. The results of the current study indicate that NPC 12626 is a novel, systemically active and competitive NMDA receptor antagonist.

Characterization of the binding of radioligands to the N-methyl-D-aspartate, phencyclidine, and glycine receptors in buffy coat membranes.

Rapid vacuum filtration assays were used to quantitate radioligand binding to phencyclidine (PCP), N-methyl-D-aspartate (NMDA), and strychnine-insensitive glycine receptors in a simple buffy coat preparation of rat cortical membranes. KD and Bmax values for [3H]glycine binding were very similar to those previously reported by workers who used centrifugation for the separation of free and bound [3H]glycine. We also found that this preparation had a high percentage of NMDA-displaceable L-[3H]glutamate binding sites, which demonstrated a pharmacology very similar to that previously observed in more purified synaptic plasma membranes. Hill analysis of the displacement curves indicated that glutamate bound to a single class of sites, but that NMDA and NMDA antagonists may interact with this site in a negatively cooperative fashion. This preparation was also found to be suitable for the study of NMDA and glycine receptor regulation of the associated ion channel, as these effectors, alone and in combination, increased the affinity with which [3H]TCP bound to the PCP receptor believed to be located within the ion channel. Thus, the ability to measure radioligand binding to these three sites in the same simple membrane preparation should greatly facilitate the study of the interaction between them.

Effects of nicotinic agonists and antagonists on N-methyl-D-aspartate-induced 3H-norepinephrine release and 3H-(1-[1-(2-thienyl)cyclohexyl]-piperidine) binding in rat hippocampus.

The nicotinic agonists dimethylphenylpiperazinium iodide (DMPP) and carbachol (CARB) as well as (-)nicotine [-)NIC) were tested alone and in combination with N-methyl-D-aspartate (NMDA) for their abilities to enhance the efflux of 3H-norepinephrine (NE) from slices of rat hippocampus. CARB and (-)NIC produced small, transient increases in NE efflux, while DMPP produced larger, longlasting increases. Inasmuch as the nicotinic antagonists mecamylamine (MECA) and hexamethonium (C6) did not consistently inhibit the increases in NE efflux produced by these agonists, the role of a nicotinic receptor in mediating these responses is uncertain. CARB and DMPP enhanced the ability of NMDA to stimulate NE release, while (-)NIC did not. MECA, but not C6, was found to selectively antagonize NMDA-stimulated NE release that did not appear to involve a nicotinic receptor. Binding studies indicated that MECA and the related nicotinic antagonist pempidine produced an inhibition of NMDA-stimulated NE release by an action at the PCP receptor that is known to be linked to the NMDA receptor-ionophore complex. These data suggest that the actions of these ganglionic blocking agents on excitatory responses in the hippocampus involve inhibition of excitatory amino acid as well as nicotinic receptors.

Structural requirements for activation of the glycine receptor that modulates the N-methyl-D-aspartate operated ion channel.

It has recently been demonstrated that glycine can potentiate several measures of N-methyl-D-aspartate (NMDA)-induced channel opening, including radioligand binding to the PCP receptor. These data suggest that the NMDA/PCP receptor complex may be allosterically modulated by a binding site for glycine. We report here that several other monocarboxylic amino acids enhance NMDA-induced [3H]TCP binding and displace [3H]glycine binding with similar apparent affinities and stereoisomerism. The results are discussed with relation to the structural requirements for compounds to bind to this site.