Effects of 5'-alkyl-benzothiadiazides on (R,S)-alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor biophysics and synaptic responses

Ovarian hormone loss impairs excitatory synaptic transmission at hippocampal CA3-CA1 synapses

Premature and long-term ovarian hormone loss following ovariectomy (OVX) is associated with cognitive impairment. This condition is prevented by estradiol (E2) therapy when initiated shortly following OVX but not after substantial delay. To determine whether these clinical findings are correlated with changes in synaptic functions, we used adult OVX rats to evaluate the consequences of short-term (7-10 d, OVXControl) and long-term (∼5 months, OVXLT) ovarian hormone loss, as well as subsequent in vivo E2 treatment, on excitatory synaptic transmission at the hippocampal CA3-CA1 synapses important for learning and memory. The results show that ovarian hormone loss was associated with a marked decrease in synaptic strength. E2 treatment increased synaptic strength in OVXControl but not OVXLT rats, demonstrating a change in the efficacy for E2 5 months following OVX. E2 also had a more rapid effect: within minutes of bath application, E2 acutely increased synaptic strength in all groups except OVXLT rats that did not receive in vivo E2 treatment. E2's acute effect was mediated postsynaptically, and required Ca(2+) influx through the voltage-gated Ca(2+) channels. Despite E2's acute effect, synaptic strength of OVXLT rats remained significantly lower than that of OVXControl rats. Thus, changes in CA3-CA1 synaptic transmission associated with ovarian hormone loss cannot be fully reversed with delayed E2 treatment. Given that synaptic strength at CA3-CA1 synapses is related to the ability to learn hippocampus-dependent tasks, these findings provide additional insights for understanding cognitive impairment-associated long-term ovarian hormone loss and ineffectiveness for delayed E2 treatment to maintain cognitive functions.

Probing the allosteric modulator binding site of GluR2 with thiazide derivatives

Ionotropic glutamate receptors mediate the majority of vertebrate excitatory synaptic transmission and are therapeutic targets for cognitive enhancement and treatment of schizophrenia. The binding domains of these tetrameric receptors consist of two dimers, and the dissociation of the dimer interface of the ligand-binding domain leads to desensitization in the continued presence of agonist. Positive allosteric modulators act by strengthening the dimer interface and reducing the level of desensitization, thereby increasing steady-state activation. Removing the desensitized state for simplified analysis of receptor activation is commonly achieved using cyclothiazide (CTZ), the most potent modulator of the benzothiadiazide class, with the flip form of the AMPA receptor subtype. IDRA-21, the first benzothiadiazide to have an effect in behavioral tests, is an important lead compound in clinical trials for cognitive enhancement as it can cross the blood-brain barrier. Intermediate structures between CTZ and IDRA-21 show reduced potency, suggesting that these two compounds have different contact points associated with binding. To understand how benzothiadiazides bind to the pocket bridging the dimer interface, we generated a series of crystal structures of the GluR2 ligand-binding domain complexed with benzothiadiazide derivatives (IDRA-21, hydroflumethiazide, hydrochlorothiazide, chlorothiazide, trichlormethiazide, and althiazide) for comparison with an existing structure for cyclothiazide. The structures detail how changes in the substituents at the 3- and 7-positions of the hydrobenzothiadiazide ring shift the orientation of the drug in the binding site and, in some cases, change the stoichiometry of binding. All derivatives maintain a hydrogen bond with the Ser754 hydroxyl, affirming the partial selectivity of the benzothiadiazides for the flip form of AMPA receptors.

Identification of a putative binding site for 5-alkyl-benzothiadiazides in the AMPA receptor dimer interface

Crystal structures of three different allosteric modulators co-crystallized with the iGluR2 ligand-binding domain are currently available. The modulators, cyclothiazide, aniracetam and CX614, bind at overlapping binding sites in the dimer interface between two iGluR2 subunits. However, pharmacological data indicate that there are one or more additional binding sites for this class of compounds. Based on differences in structure-activity relationship data we show that 5-alkyl-benzothiadiazide (5ABTD) modulators and a series of close analogs of cyclothiazide, despite having a common core structure, do not have the same binding site. In the present work, a new potential binding site for allosteric modulators has been identified in the dimer interface of the iGluR2 ligand-binding domain. By comparing different iGluR2 crystal structures including different co-crystallized agonists, this cavity is shown to be a structurally conserved part of the dimer interface. The cavity is characterized with respect to shape and potential favorable interactions with ligands and docking is used to find a reasonable binding mode for the core structure of the 5ABTDs. The extensive structure-activity data available for this series of compounds are in agreement with the proposed binding mode, supporting the conclusion that the identified cavity most likely is the binding site for the 5ABTDs.

Therapeutic potential of positive AMPA receptor modulators in the treatment of neuropsychiatric disorders

Drugs that potentiate the activity of the alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor cause a complex cascade of consequences in experimental models, ranging from enhancement of long-term potentiation to induction of neurotrophic factors. Animal studies characterising the pharmacological and behavioural effects of these substances have provided the rationale for several initial attempts to use these drugs in neuropsychiatric clinical settings. Applications in schizophrenia, Alzheimer's disease and mild cognitive impairment have been initiated. Other trials with these compounds include the treatment of Fragile X syndrome, and possible future applications may be in the field of Parkinson's disease. The literature published to date is limited mostly to small phase I or II trials, so there is no conclusive evidence for or against the use of these drugs. Substantial questions remain concerning which compounds to use, in what dose, for what condition and for how long.

Use of [3H]fluorowillardiine to study properties of AMPA receptor allosteric modulators

Compounds which modulate AMPA receptor function through allosteric mechanisms were examined for their effect on the binding of the agonist [3H]fluorowillardiine (FW). Benzamide-type positive modulators (ampakinestrade mark) under all experimental circumstances increased [3H]FW binding to native receptors in rat brain membranes. Benzothiadiazide drugs had more variable effects ranging from large reductions with cyclothiazide and JM-13 to increases produced by more recent compounds like PEPA, D1 and LY392098. These effects on binding were moderately influenced by the assay conditions, including temperature and the presence or absence of thiocyanate. Significant changes in agonist binding were also produced by other modulatory agents such as noncompetitive blockers (GYKI 53655, SYM 2206), polycationic compounds (spermine, Naspm, philanthotoxin) and polyanionic compounds (Evans Blue, suramin, PPNDS). EC50 values usually were similar to those from physiological studies, which validates using binding tests to assess drug potencies. Moreover, direction and magnitude of the binding change (Emax) provide information about which kinetic aspects are affected by a drug. For example, the magnitude of the binding increase produced by positive modulators was strongly correlated with their ability to slow response deactivation in excised patch recordings. Binding also provides a reliable method to examine whether interactions between agents are competitive. Thus, thiocyanate did not significantly influence the EC50 of cyclothiazide, suggesting distinct sites of action. Taken together, [3H]FW binding can yield important information about drug-receptor and drug-drug interactions for a wide range of modulatory agents. One potential limitation of [3H]FW is a large preference for subunits GluR1 and GluR2 (KD 4-10 nM) over GluR3 and GluR4 (160-600 nM) which implies that tests with brain membranes preferentially reveal drug effects produced at the former two subunits. Lastly, data are shown which highlight the importance of optimizing experimental conditions in filtration assays, for instance by always including thiocyanate in wash buffers.

C-terminal truncation affects kinetic properties of GluR1 receptors

GluR1flop receptors in which the C-terminal 52 amino acids had been recombinantly removed were characterized with whole-cell recording and binding assays. Compared to wildtype GluR1, truncated receptors showed faster desensitization and deactivation and they recovered more slowly from desensitization. The EC50 for glutamate was increased 2-fold. In binding tests, K(D)s for [3H]fluorowillardiine were 1.5 times larger for truncated receptors. According to receptor simulations, most differences can be explained if the C-terminal domain is assumed to stabilize the ligand-bound closed and open states. The effects on response waveforms are different from those caused by phosphorylation, suggesting that the C-terminus influences receptor function in multiple ways. Truncated forms of GluR1 identical or similar to the one examined here may also be generated by calcium-activated proteases during intense synaptic activity. The lowered affinity and faster inactivation of these receptors suggests that their presence does not represent a risk for neuronal viability.

Prolonged positive modulation of alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors induces calpain-mediated PSD-95/Dlg/ZO-1 protein degradation and AMPA receptor down-regulation in cultured hippocampal slices

Prolonged exposure of cultured hippocampal slices to CX614 [2H,3H,6aH-pyrrolidino[2'',1''-3',2']1,3-oxazino[6',5'-5,4]-benzo[e]1,4-dioxan 10-one], a positive alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor (AMPAr) modulator, decreases receptor response to synaptic stimulation, an effect that could reflect reduced receptor expression. The present study investigates this down-regulation and its underlying mechanisms using cultured rat hippocampal slices. Chronic treatment with CX614 gradually reduced levels of glutamate receptor (GluR)1 and GluR2/3 AMPAr subunits and of their anchoring proteins synapse-associated protein 97 (SAP97) and glutamate receptor interacting protein 1 (GRIP1) through 48 h. Decline in SAP97 and GRIP1 levels was associated with increased abundance of lower molecular weight bands, suggesting degradation of these proteins. CX614 effects were partially reversible after drug removal. GluR1 and GluR2/3 down-regulation and their slow recovery were associated with similar changes in SAP97 and GRIP1 levels. Treatment with CX614 for 48 h significantly reduced AMPAr mRNA levels in hippocampus, whereas 8-h exposure did not. Blockade of ionotropic glutamate receptors prevented CX614-induced decrease in AMPAr subunits and mRNA, with regional selectivity, although an AMPAr blocker was more efficacious than an N-methyl-D-aspartate receptor blocker. Blockade of calpain activity reduced CX614-induced degradation of SAP97 and GRIP1 and prevented decreases in AMPAr subunit but not mRNA levels. Treatment with CX614 alone or in combination with glutamate receptor blockers or calpain inhibitor III did not modify lactate dehydrogenase release into culture medium, implying the absence of cell toxicity. We conclude that CX614-induced AMPAr protein loss is primarily mediated by AMPAr activation and involves calpain-dependent proteolysis of SAP97 and GRIP1. CX614-induced suppression of AMPAr gene expression is, however, calpain-independent, and all these effects are not associated with cell damage.

Divergent effects of the purinoceptor antagonists suramin and pyridoxal-5'-phosphate-6-(2'-naphthylazo-6'-nitro-4',8'-disulfonate) (PPNDS) on alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors

Suramin is a large naphthyl-polysulfonate compound that inhibits an array of receptors and enzymes, and it has also been reported to block currents mediated by glutamate receptors. This study shows that suramin and several structurally related compounds [8,8'-[carbonylbis(imino-3,1-phenylenecarbonylamino)]bis-(1,3,5-naphthalenetrisulfonic acid), 6Na (NF023), 8,8'-(carbonylbis(imino-4,1-phenylenecarbonylimino-4,1-phenylenecarbonylimino))bis-1,3,5-naphthalenetrisulfonic acid, Na (NF279), and 4,4',4'',4'''-[carbonyl-bis[imino-5,1,3-benzenetriyl-bis-(carbonylimino)]]tetrakis-benzene-1,3-disulfonic acid, 8Na (NF449)] reduce binding of [3H]alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and [3H]fluorowillardiine to rat brain membranes and homomeric GluR1-4 receptors, with IC50 values in the range of 5 to 180 microM. Inhibition often was less than complete at saturating drug concentrations and thus seems to be noncompetitive in nature. Pyridoxal-5'-phosphate-6-(2'-naphthylazo-6'-nitro-4',8'-disulfonate) (PPNDS) is a potent antagonist of purinoceptors that shares some structural elements with suramin yet is smaller than the latter. PPNDS also had potent effects on AMPA receptors (EC50 value of 4 microM) but of a kind not seen with the other compounds in that it increased binding affinity for radioagonists severalfold. In addition, PPNDS slowed association and dissociation rates more than 10 times. In physiological experiments with GluR2 receptors, PPNDS at 50 microM reduced the peak current by 30 to 50% but had only small effects on other waveform aspects such desensitization and steady-state currents. This pattern of effects differentiates PPNDS from other compounds such as thiocyanate and up-modulators, which increase agonist binding by enhancing desensitization or slowing deactivation, respectively. Receptor model simulations indicate that most effects can be accounted for by assuming that PPNDS slows agonist binding/unbinding and stabilizes the bound-closed state of the receptor. By extension, suramin is proposed to stabilize the unbound state and thereby to reduce affinity for agonists. These drugs thus act through a novel type of noncompetitive antagonism.

Modulation of AMPA receptor kinetics differentially influences synaptic plasticity in the hippocampus

Prior studies showed that positive alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor modulators facilitate long-term potentiation (LTP) and improve the formation of several types of memory in animals and humans. However, these modulators are highly diverse in their effects on receptor kinetics and synaptic transmission and thus may differ also in their efficacy to promote changes in synaptic strength. The present study examined three of these modulators for their effects on synaptic plasticity in field CA1 of hippocampal slices, two of them being the benzamide drugs 1-(quinoxalin-6-ylcarbonyl)piperidine (CX516) and 1-(1,4-benzodioxan-6-ylcarbonyl)piperidine (CX546) which prominently enhance synaptic transmission yet differ in their relative impact on amplitude versus duration of the synaptic response. The third drug was cyclothiazide which potently blocks AMPA receptor desensitization. Effects on plasticity were assessed by measuring (i) the likelihood of obtaining stable potentiation when using theta-burst stimulation with three instead of four pulses per burst, (ii) the maximum amount of potentiation under optimal stimulation conditions, and (iii) the effect on long-term depression (LTD). Both benzamides facilitated the formation of stable potentiation induced with three-pulse burst stimulation which is normally ineffective. CX546 in addition increased maximally inducible potentiation after four-pulse burst stimulation from about 50% to 100%. Burst response analysis revealed that CX546 greatly prolonged the duration of depolarization by slowing the decay of the response which thus presumably leads to a more continuous N-methyl-D-aspartate (NMDA) receptor activation. Cyclothiazide was ineffective in increasing maximal potentiation in either field or whole-cell recordings. CX546, but not CX516, also enhanced nearly two-fold the NMDA receptor-dependent long-term depression induced by heterosynaptic 2 Hz stimulation. Tests with recombinant NMDA receptors (NR1/NR2A) showed that CX516 and CX546 have no direct effects on currents mediated by these receptors. These results suggest that (1) modulation of AMPA receptors which increases either response amplitude or duration can facilitate LTP formation, (2) modulators that effectively slow response deactivation augment the maximum magnitude of LTP and LTD, and (3) receptor desensitization may have a minor impact on synaptic plasticity in the hippocampus. Taken together, our data indicate that AMPA receptor modulators differ substantially in their ability to enhance synaptic potentiation or depression, depending on their particular influence on receptor kinetics, and hence that they may also be differentially effective in influencing higher-order processes such as memory encoding.

Identification of a site in GluR1 and GluR2 that is important for modulation of deactivation and desensitization

The alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid subtype of ionotropic glutamate receptors consists of rapidly gating ion channels. Positive modulation of channel gating may slow gating kinetics through at least two distinct mechanisms, evidenced by the predominant slowing of either the rate of receptor desensitization or the rate of offset after agonist withdrawal (deactivation). This study compares the actions of two positive allosteric modulators [cyclothiazide, which modulates desensitization, and 1-(1,4-benzodioxan-6-ylcarbonyl)piperidine (CX546), which modulates deactivation] in a mutant shown previously to impede modulation by cyclothiazide. These experiments test the hypothesis that the point mutation, GluR1(S493T), would also cause a loss of modulation by CX546. Wild-type GluR1 through -4 receptors were modulated by CX546, as assayed by the potentiation of steady-state currents in the Xenopus laevis oocyte expression system. CX546 potentiated steady-state currents of both splice isoforms of GluR1. Modulation by CX546 was completely abolished in GluR1(S493T) and its homolog, GluR2(S497T), although this mutation did not affect apparent agonist affinity in the absence of CX546. Thus, the GluR1(S493T) mutation has a similar impairment of modulation by either cyclothiazide or CX546, indicating that some residues at the subunit interface of glutamate receptors play an important role in channel deactivation and desensitization.

Benzamide-type AMPA receptor modulators form two subfamilies with distinct modes of action

CX516 (BDP-12) and CX546, two first-generation benzamide-type AMPA receptor modulators, were compared with regard to their influence on AMPA receptor-mediated currents, autaptic responses in cultured hippocampal neurons, hippocampal excitatory postsynaptic currents, synaptic field potentials, and agonist binding. The two drugs exhibited comparable potencies in most tests but differed in their efficacy and in their relative impact on various response parameters. CX546 greatly prolonged the duration of synaptic responses, and it slowed 10-fold the deactivation of excised-patch currents following 1-ms pulses of glutamate. The effects of CX516 on those measures were, by comparison, small; however, the drug was equally or more efficacious than CX546 in increasing the amplitude of synaptic responses. This double dissociation suggests that amplitude and duration of synaptic responses are governed by different aspects of receptor kinetics, which are differentially modified by the two drugs. These effects can be reproduced in receptor simulations if one assumes that CX516 preferentially accelerates channel opening while CX546 slows channel closing. In binding tests, CX546 caused an approximately 2-fold increase in the affinity for radiolabeled agonists, whereas CX516 was ineffective. More importantly, even millimolar concentrations of CX516 did not influence the dose-response relation for CX546, suggesting the possibility that they bind to different sites. Taken together, the evidence suggests that benzamide modulators from the Ampakine family form two subgroups with different modes and sites of action. Of these, CX516-type drugs may have the greater therapeutic utility because of their limited efficacy in prolonging synaptic responses and in attenuating receptor desensitization.