Recent advances in targeting the ionotropic glutamate receptors in treating schizophrenia

Abstract

The treatment of schizophrenia has been focused on modulation of dopamine receptors for over 50 years. Recent developments have implicated other neurotransmitter systems in the pathophysiology of this illness. The discovery and characterization of glutamate receptors and their roles in the brain has lead to novel approaches for the treatment of schizophrenia. In this article, we review drugs that modulate ionotropic gluamate receptors and discuss their efficacy for the treatment of this often debilitating severe mental illness.

A. The schizophrenia syndrome

Schizophrenia is a devastating illness that afflicts over 2 million people in the U.S. and over 50 million people worldwide [1]. This may be the most serious of all psychiatric illnesses, since more hospital beds are filled by persons with schizophrenia than due to any other medical condition [1]. Schizophrenia is a syndrome characterized by positive, negative, and cognitive symptoms [1]. Positive symptoms typically include auditory hallucinations. Patients report that they hear voices that emanate from outside of their heads, often engaged in a running commentary on the patient's behaviors or thoughts. Much more insidious and debilitating are the negative symptoms, which are associated with the loss or impairment of social behaviors. Negative symptoms may include decreased spontaneous communication, withdrawal, decreased or muted facial expression, diminished vocal inflection, diminished spontaneous movement, and decreased eye contact [1]. Cognitive deficits in this illness include abnormalities of working memory and executive function [1]. Few individuals suffering from schizophrenia have all of these symptoms, but the persistence of several characteristic symptoms, like auditory hallucinations, must be present in order for someone to be diagnosed with this disorder [1].

B. Treatment of schizophrenia with antipsychotic medications

The pharmacological revolution in psychiatry started, in part, with the discovery of chlorpromazine as a treatment for psychiatric illness [2, 3]. Prior to the discovery of this and other antipsychotic medications, patients with psychosis were frequently institutionalized for long periods of time and offered a variety of treatments with questionable benefit [4]. The superior efficacy of these original (now called “typical”) antipsychotics compared to previous treatment modalities revolutionized the clinical management of schizophrenia by providing a pharmacological intervention that provided relief from psychotic symptoms in a majority of patients [2, 3]. Subsequent research demonstrated that these effects were associated with dopamine D2 receptor antagonism, a defining pharmacological feature of earlier antipsychotics [5, 6].

While typical antipsychotics have good efficacy for positive symptoms, they also have significant side effects, including extrapyramidal symptoms (EPS), and tardive dyskinesia, which affect treatment adherence. The second generation (so-called “atypical”) antipsychotics were developed with an eye towards minimizing these often disabling side effects and improving efficacy for negative symptoms [7, 8]. The atypicals also work via dopamine D2 receptor antagonism. Although these newer medications have an improved EPS profile, they are associated with significant weight gain and adverse metabolic effects, and the overall benefit of using these medications versus the older antipsychotics is currently an area of debate [9-12]. Finally, the classes of typical and atypical medications are not homogenous with regard to side effects. For example, agents like chlorpromazine induce few extrapyramidal side effects

Although the advances in treatment of schizophrenia due to these drugs are remarkable, a significant limitation is that some patients taking these medications do not improve. Poor treatment responders are typically defined by failure to reach a defined clinical threshold determined by measuring changes in symptoms using rating scales for positive and negative symptoms [13-17]. Approximately one third of patients with schizophrenia may be classified as poor treatment responders, despite multiple trials of medications [13, 14, 18, 19]. The data for response rates are only applicable for those patients who are treatment adherent. Antipsychotic medications generally have a high discontinuation rate due to side effects and other factors. In the large CATIE trial, where no differences in efficacy between one typical and several atypical drugs were found, 74% of all patients discontinued their medications during the study [12, 20]. Taken together, the rate of poor response to antipsychotic treatment and the high rate of discontinuation suggest that the pharmacological approach to treating schizophrenia could be significantly improved.

C. The glutamate hypothesis of schizophrenia

For decades, schizophrenia research has focused on the dopamine hypothesis of schizophrenia, which postulates that dysregulated dopaminergic neurotransmission is a key feature of the pathophysiology of the illness [21-23]. The dopamine hypothesis is based on the observation that antipsychotics block dopamine D2 receptors, and their affinity for these receptors highly correlates with the ability to ameliorate psychotic symptoms. In addition, psychostimulants that enhance dopamingergic activity can elicit positive symptoms of schizophrenia, including delusions, hallucinations, and thought disorder [24]. Although numerous studies point to dopaminergic abnormalities in schizophrenia, dopamine dysfunction may not completely account for all of the symptoms seen in schizophrenia, since antipsychotics are typically effective only for the positive symptoms of the illness [25, 26]. Consequently, alternative neurotransmitter systems that may also be involved in the pathophysiology of schizophrenia have been sought, and a growing body of evidence has implicated glutamatergic dysfunction. Specifically, the most widely held hypotheses posit altered glutamate activity in limbic brain structures in schizophrenia [27-30]. This hypothesis is broadly supported by studies demonstrating altered levels of glutamate in the cerebrospinal fluid, altered levels of glutamate and related metabolites in brain as determined by magnetic resonance spectroscopy, and postmortem studies showing altered expression in brain of glutamatergic enzymes in this illness [27, 28, 31-52].

The most compelling data implicating glutamate dysfunction in schizophrenia are the effects of phencyclidine (PCP) and similar compounds, which are uncompetitive antagonists of the NMDA subtype of glutamate receptor. These compounds can induce both the positive and negative symptoms of schizophrenia, as well as cognitive deficits [29, 30]. Moreover, these compounds can exacerbate both positive and negative symptoms in schizophrenia [53]. Thus, the original hypothesis of glutamate dysfunction in schizophrenia centered on NMDA receptor hypofunction, since PCP and other related compounds block function of this receptor. Subsequent work has focused on examination of glutamate receptor expression in schizophrenia (reviewed in: [54]). While there are only a few consistently reported changes in ionotropic glutamate receptor subunit and binding site expression in schizophrenia [55-57], their functional roles suggest these receptors are a potential target for new drug discovery. The complex biology of the ionotropic receptors has led to the development of novel classes of compounds that selectively modulate sites on various glutamate receptors.

D. Biology of the ionotropic glutamate receptors

Glutamate receptors are divided into two groups; the ionotropic receptors, which form ligand-gated ion channels, and the metabotropic receptors, which have seven transmembrane domains and are G-protein linked [58, 59]. Each of these families of receptors has multiple subtypes which are grouped both pharmacologically, by affinity for selective ligands, and structurally, based on gene sequence similarity. The ionotropic receptors facilitate molecular correlates of learning and memory, including long-term depression (LTD) and long-term potentiation (LTP) [60]. Ionotropic glutamate receptors are divided into three classes, N-methyl-D-aspartic acid receptors (NMDARs), alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPARs) and kainate receptors [58, 59]. AMPARs and NMDARs are primarily responsible for potentiating excitatory neurotransmission through selective ion transport.

NMDA receptors

The NR1, NR2A-D, and NR3A-B subunits combine in functionally distinct tetramers, usually forming a dimer of dimers. The NR1 subunit is obligatory for a fully assembled NMDA receptor, usually with the addition of a dimer of NR2 subunits. NMDA receptors exhibit subunit and splice variant specific properties, and pharmacological regulation of this receptor depends on the unique combination of glutamate, glycine/D-serine, polyamine, pH, Zn^2+, and Mg²⁺ binding sites [58, 59]. In addition, there is an intrachannel binding site for uncompetitive antagonists of the NMDA receptor, including PCP, ketamine, and MK801. The glycine/D-serine site is present on all NR1 subunits and binding to this co-agonist site is necessary for the activation of the receptor. NR2 subunits contain glutamate binding sites in addition to other modulatory sites. NR3 subunits are primarily developmentally expressed and can replace the NR1 subunit in the receptor complex [61-64].

AMPA receptors

The AMPA receptor consists of four subunits, GluR1-4, that are typically assembled as dimers into a tetrameric complex [58, 65, 66]. Each subunit contains a binding site for glutamate [67]. Binding of glutamate to two sites on the receptor causes a confirmational change and opening of the ion channel [68]. The subunit composition of each receptor confers different properties to the ion channel. For example, the presence or absence of the GluR2 subunit in a receptor can affect the permeability of the receptor to calcium ions. Assembled AMPA receptors contain discrete binding sites for glutamate, competitive antagonists such as CNQX, and desensitization modulators such as aniracetam.

Kainate receptors

There are five kainate receptor subunits, GluR 5-7 and KA1 and KA2, that are similarly assembled as tetrameric complexes [58, 59, 69]. These subunits also undergo alternative splicing and post-translational editing. Their structure is very similar to the AMPA receptor subunits, and each subunit has a glutamate binding site [69]. Unlike AMPA receptors, kainate receptors show strong desensitization and slow response recovery [70]. Also, unlike other types of ionotropic glutamate receptors, kainate receptors are located both pre- and post-synaptically [71, 72]. Presynaptic kainate receptors are hypothesized to control glutamate release [73, 74].

Ionotropic receptors in glutamate synapses

The process of release, activity as a ligand, and reclamation of glutamate involves three distinct cell types: the astrocyte, the presynaptic neuron and the postsynaptic neuron [75]. In the presynaptic neuron, glutamine can be converted to glutamate by the enzyme glutaminase, and packaged into vesicles by a family of vesicular glutamate transporters (VGLUT1-3) for release into the synapse [76, 77]. Glutamate is released into the synapse and may occupy and activate ionotropic or metabotropic glutamate receptors on both neurons and astrocytes [58]. Activation of AMPA receptors depolarizes the postsynaptic neuron sufficiently to facilitate removal of baseline Mg²⁺ block from NMDA receptors, permitting activation of these receptors by glutamate and glycine/D-serine [61]. Following removal of the Mg2+ blockade, Ca2+ is able to enter the postsynaptic cell through the NMDA receptor channel [78]. The influx of calcium activates several kinases which in turn phosphorylate AMPA receptor subunits to increase their activity [79, 80]. Additional proteins mobilize the lateral diffusion of an extra-synaptic pool of AMPA receptors to the synapse [81, 82]. This increase in AMPA receptors at the synapse may potentiate NMDA activity. NMDA receptor activity is also regulated by other receptor subtypes. Kainate receptors that are located presynaptically may alter the release of glutamate from the presynaptic neuron [83, 84]. Metabotropic glutamate receptors (mGluRs) are often anchored near NMDA receptors and also can modulate their activity [85, 86]. One group of mGluRs may increase the activity of NMDA receptors while others may decrease activity [87-89]. The interplay of other receptor subtypes with NMDA receptors is essential in maintaining the function and tone at the glutamate synapse.

If diminished NMDA receptor function is associated with schizophrenia, the most straightforward approach to modulate this system to treat this illness would be to give glutamate or a related agonist. However, excessive activation of glutamate receptors can lead to excitotoxicity and cell death. Toxicity and pharmacodynamic/kinetic issues have led to the development of alternative pharmacological approaches to modulate NMDA and other glutamate receptors.

NMDA receptor co-agonist-site modulators

D-serine is a full agonist of NMDA receptor glycine binding site, and is likely the endogenous ligand. In the presence of glutamate, D-serine can increase NMDA receptor activation in the treatment of schizophrenia. In the absence of glutamate, D-serine does not activate glutamate receptors. Several early studies have examined D-serine and found that at a dose of 30mg/kg/day, patients had improved positive, negative, cognitive and depression symptoms [90, 91]. Recently, a double-blind, randomized, placebo-controlled study was conducted in which no difference between D-serine and placebo was found [92]. Although the lack of efficacy in this placebo controlled study diminishes enthusiasm for D-serine treatment in schizophrenia, initial studies were promising.

Glycine has also been examined as a treatment in schizophrenia, given that it shares properties with D-serine as a critical coagonist for NMDA receptor activity. Initial trials of glycine (5-30g/day) in either drug-free or antipsychotic treated patients provided evidence for glycine therapy efficacy [93-96]. Many additional studies have confirmed the benefit of adjunctive glycine treatment[96-100]. The effect of adjunctive glycine in patients on atypical antipsychotics was predominantly improvement in negative symptoms, although one study found no benefit when added to clozapine [96-98, 100, 101]. Overall, adjunctive glycine treatment appears promising for the treatment of negative symptoms in schizophrenia.

D-cycloserine is a partial agonist of the glycine binding site of the NMDA receptor. Clinical trials have been conducted using this drug as an adjunctive treatment in patients with schizophrenia. D-cycloserine treatment improved negative symptoms when added to typical antipsychotics, but within a narrow therapeutic window with the best effects around 50mg/day [102-105]. Larger doses of D-cycloserine (250mg/day) resulted in worsening of positive symptoms [106]. A more recent large and long term treatment study found no difference in outcome between D-cycloserine and placebo [107]. Additionally, studies of D-cycloserine as an adjunct therapy to atypical antipsychotics such as clozapine reported worsened negative symptoms [104, 108]. Although modulation of the NMDA receptor glycine site with a partial agonist is a conceptually intriguing treatment option, trials of D-cycloserine have not indicated that this form of therapy is likely to be efficacious.

One study has been conducted using D-alanine, a full agonist of the glycine site, as a treatment in schizophrenia [109]. A dose of 100mg/kg/day added to stable antipsychotic dosing resulted in additional symptom improvement.

Only a few studies have examined the therapeutic effects of melacemide for the treatment of schizophrenia. The improvement of negative symptoms observed with adjunctive glycine treatment suggests that melacemide, a prodrug of glycine which readily crosses the blood brain barrier, might be efficacious and have fewer side effects than glycine. Two open label studies have been performed on drug free patients with daily doses of 0.4g/day and 1.2g/day [110, 111]. With the lower dose no symptomatic improvement was seen, and patients given the higher dose had a worsening of symptoms. These preliminary findings suggest that melacemide is probably not a good candidate for the treatment of schizophrenia.

In summary, adjunctive treatment with glycine site modulators across multiple studies results in about a 15% improvement in negative symptoms [112]. These data support further exploration of this mechanism for the treatment of schizophrenia. One additional promising avenue is inhibition of glycine and/or D-serine reuptake from the synaptic cleft. N-methylglycine (sarcosine) is a glycine-transporter 1 inhibitor that has recently been evaluated for adjunctive treatment in schizophrenia. The ability to prevent glycine from normal reuptake pathways may make it more available to NMDA receptors, thus increasing function. Several studies have examined the efficacy of sarcosine and found 2g/day can improve symptoms in patients with schizophrenia [92, 113-115]. One of these trials studied medication free patients, and the others treated patients stable on antipsychotics, raising the possibility of that glycine transport inhibitors may be used as monotherapy. Another study examined treatment response in patients on clozapine therapy, and found that similar to the results found with clozapine plus glycine, adjunctive sarcosine therapy conferred no additional therapeutic benefit [116].

AMPA receptor modulators: Ampakines

Ampakines are compounds that interact with AMPA receptors to modulate synaptic activity. Ampakines allosterically bind to AMPA receptors to facilitate synaptic transmission [117]. CX516 is a prototypical ampakine that was given as monotherapy to patients with schizophrenia with no clear improvement in psychosis or cognition [118]. However, when added to clozapine, CX516 improved measures of attention and memory [119]. Meta-analyses of studies using multiple adjunct therapies found that ampakines were ineffective in reducing the positive symptoms of schizophrenia [120-122]. A recent, large, placebo-controlled trial used CX516 as adjunct therapy to clozapine, risperidone, or olanzapine over a 4-week period. CX516 did not improve measures of cognition in these patients, and the placebo group had significantly greater improvement in positive and negative symptoms than group treated with CX516 [120]. While modulation of synaptic transmission with ampakines seemed like a promising target for pharmacotherapy of schizophrenia, the currently available compounds do not appear to improve symptoms in this illness.

Other glutamatergic agents

Lamotrigine

Lamotrigine is an anticonvulsant drug that is also used in maintenance treatment of bipolar disorder. While its exact mechanism of action is unknown, lamotrigine may act by blocking sodium channels by binding and stabilizing inactive channels and preventing trains of action potentials [123]. It has been postulated that dysregulation of cortical activity and glutamate release in schizophrenia may be attenuated with use of lamotrigine as an add-on therapy [124]. One study found that lamotrigine was effective in improving positive symptoms and general psychopathology scores when used in conjunction with clozapine [125]. Several other studies reported a similar improvement in positive symptoms and general psychopathology scores [126, 127]. While one meta-analysis concluded that lamotrigine was effective in improving both positive and negative symptoms of schizophrenia, another meta-analysis found no robust evidence for improvement in these symptoms [128, 129]. No improvement of symptoms was noted in two recent small studies [130, 131]. A large, multicenter, double-blind, placebo-controlled study used lamotrigine as add-on therapy to atypical antipsychotic-resistant schizophrenia. While one site reported slight improvements in cognitive symptom scores, overall there was no change improvement in positive and negative symptoms [132].

Memantine

Approved for the treatment of Alzheimer's disease as a cognitive enhancer, memantine modulates blockade of current flow through NMDA receptors [133]. Several preliminary studies found improvement in negative symptoms using memantine as an adjunctive treatment [134-136]. However, a larger double blind, placebo controlled study found no effect on symptoms and a higher rate of adverse effects compared to placebo [137]. While memantine may not be useful as an adjunctive treatment for positive and negative symptoms in schizophrenia, it may be useful for the treatment of catatonia. There are several reports of memantine being used to successfully treat catatonia in treatment refractory patients [138-140]. Improvement of catatonia with memantine might involve blockade of excitatory inputs to inhibitory GABAergic circuits, decreasing inhibitory tone, leading to increased arousal and behavioral spontaneity.

Acamprosate

Indicated for the treatment of alcoholism, acamprosate modulates NMDA receptor function by enhancing activation when receptor activity is low, and blocking activation when receptor acvtivity is high [141]. While there are no published reports of using acamprosate to treat schizophrenia, at least one clinical trial is underway. This compound has been successfully used to treat alcoholism in a patient with schizophrenia [142].

Summary

Since the discovery of chlorpromazine nearly sixty years ago, there have been few major advances in the treatment of schizophrenia. Chlorpromazine and other typical antipsychotics are effective in the treatment of the positive symptoms of schizophrenia, but they have significant motor side effects. Although the atypical antipsychotics have a different side effect profile with fewer motor symptoms, they do not substantially improve the negative and cognitive symptoms of schizophrenia beyond the typical antipsychotics. The observation that glutamatergic modulation may result in symptoms similar to those seen in schizophrenia, including negative and cognitive symptoms, has prompted the development and study of compounds that modulate glutamatergic receptors. Several studies have tried with variable success to modulate the co-agonist site of NMDA receptors or to modulate AMPA receptor function in patients with schizophrenia. Other studies have examined the use of medications that may have glutamatergic activity. While there has been much excitement and theoretical reasons to think these approaches will work, with the exception possibly of the glycine modulatory site, most of the glutamatergic drugs have yielded negative data. However, the complex biology of glutamate receptors, including multiple subtypes, binding sites, and interacting proteins, provides many possible targets for altering receptor function. Compounds that modulate metabotropic glutamate receptors are currently being evaluated for the treatment of schizophrenia and represent a promising new lead. While there are no singular compounds to date that treat all deficits of schizophrenia, treatment with multiple medications may prove to be efficacious. To this end, modulation of glutamatergic neurotransmission is a promising area of research in the development of medications capable of treating negative and cognitive symptoms of schizophrenia.