Molecular Targets for Treating Cognitive Dysfunction in Schizophrenia

Abstract

Cognitive impairment is a core feature of schizophrenia as deficits are present in the majority of patients, frequently precede the onset of other positive symptoms, persist even with successful treatment of positive symptoms, and account for a significant portion of functional impairment in schizophrenia. While the atypical antipsychotics have produced incremental improvements in the cognitive function of patients with schizophrenia, overall treatment remains inadequate. In recent years, there has been an increased interest in developing novel strategies for treating the cognitive deficits in schizophrenia, focusing on ameliorating impairments in working memory, attention, and social cognition. Here we review various molecular targets that are actively being explored for potential drug discovery efforts in schizophrenia and cognition. These molecular targets include dopamine receptors in the prefrontal cortex, nicotinic and muscarinic acetylcholine receptors, the glutamatergic excitatory synapse, various serotonin receptors, and the γ-aminobutyric acid (GABA) system.

Introduction

Schizophrenia is characterized by positive symptoms such as delusions and hallucinations, negative symptoms such as avolition and flat affect, and cognitive impairments. Although Kraepelin,¹ with his term “dementia praecox,” characterized the relationship between cognitive deficits and schizophrenia nearly a century ago, effective treatments for these deficits have not been developed. Cognitive dysfunction is estimated to occur in 75%–85% of patients with schizophrenia,² often precedes the onset of other symptoms,² and persists even after other symptoms have been effectively treated.³ Indeed, a meta-analysis of cognitive deficits suggested that indices of cognitive deficits are much better predictors of functional outcome than indices from any other symptom domain.⁴ Furthermore, the severity of cognitive deficits is predictive of poorer medication compliance,⁵ overall treatment adherence,⁶ and increased tendency for relapse in first-episode patients.⁷

Until recently, antipsychotic drug development in schizophrenia has focused mainly on developing drugs that reduce the positive symptoms of schizophrenia,⁸ and indeed, all the current medications appear to be similar in efficacy for reducing positive symptoms in typical patients with schizophrenia.^9,10 In a recent meta-analysis, patients treated with typical antipsychotics were actually shown to have small but detectable improvements in several cognitive domains¹¹; however, due to extrapyramidal side effects, many patients are also treated with anticholinergic agents that are well known to impair memory¹² and global cognitive ability.¹³ In addition, there is some evidence for the superiority of atypical antipsychotics, such as olanzapine and risperidone, over typicals in improving cognitive performance,^14–16 though the benefits are relatively small and have not been consistently reproduced.¹⁷ Overall, the widespread use of the atypical antipsychotics has likely offered some cognitive benefit for patients with schizophrenia,¹⁸ though significant deficits persist, suggesting a need for directive treatments for enhancing cognition.

Due to the continued need for improved treatment of the cognitive impairments in schizophrenia, Wayne Fenton spearheaded the National Institutes of Mental Health's joint academic and industry initiative termed MATRICS (Measurement and Treatment Research to Improve Cognition in Schizophrenia) to facilitate the development of better treatments targeted at cognition.¹⁹ An initial MATRICS conference (http://www.matrics.ucla.edu) identified 7 primary cognitive domains that are crucial for developing targets for the treatment of cognition in schizophrenia. These domains included working memory, speed of processing, verbal learning and memory, attention and vigilance, reasoning and problem solving, visual learning and memory, and social cognition.^20–22 An additional MATRICS conference (http://www.matrics.ucla.edu) identified pharmacologic strategies that hold promise for the treatment of impaired cognition in schizophrenia. The primary molecular targets identified included dopamine receptors in the prefrontal cortex, nicotinic and muscarinic acetylcholine receptors, the glutamatergic excitatory synapse, various serotonin receptors, and the γ-aminobutyric acid (GABA) system. Below, we review many of the molecular targets being studied for potential drug development strategies aimed at enhancing cognition, both generally and specifically in schizophrenia (table 1).

Table 1.

Molecular Targets for Cognitive Enhancement in Schizophrenia

Molecular Target	Example Compound	Clinical Evidence
D1 agonistsa	Dihydrexidine	Positive proof-of-concept trials
D4 antagonists	Sonepiprazole	Ineffective in acute schizophrenia
D4 agonists	A-412997	n/a
COMT inhibitors	Tolcapone	Improved cognition in Parkinson's disease
5-HT2A antagonists	M100907	Modest efficacy in acute schizophrenia
5-HT1A agonists	Tandospirone	Mixed results
5-HT1A antagonists	WAY100635	n/a
5-HT4 agonists	Tegaserod	Recently withdrawn from market in IBSb
5-HT6 antagonists	SB-271046	n/a
5-HT7 agonists	No selective ligands	n/a
Cholinesterase inhibitors	Donepezil	Multiple trials, mixed results
Nicotinic α7 agonists	DMXB-A	Positive proof-of-concept trials
Nicotinic α4β2 agonists	RJR2403	n/a
M1 agonists	NDMC (nonselective)	Positive effects with nonselective agonists
M4 agonists	No selective ligands	n/a
M5 antagonists	No selective ligands	n/a
NMDA enhancers	Glycine	Multiple small trials, mostly positive
GlyT inhibitors	Org-24598	Positive proof-of-concept trials
Ampakines	CX-516	Mixed proof-of-concept trials
mGluR2/3 agonists	Unknown	Positive proof-of-concept trial reported
mGluR5 agonists	CDPPB	n/a
α2-adrenergic antagonists	Guanfacine	Positive small trials
GABAA (α2) agonists	TPA023	Single positive proof-of-concept trial
GABAA (α5) antagonists	L-655708	n/a
Sigma agonists	No selective ligands	n/a

Note: D, dopamine; n/a, clinical evidence not available to date; COMT, catechol-O-methyltransferase; 5-HT, serotonin; IBS, irritable bowel syndrome; DMXB-A, 3-2,4-dimethoxybenzylidene anabaseine; M, muscarinic; NMDA, N-methyl-D-aspartate; GlyT, glycine transporter; mGluR, metabotropic glutamate receptor; GABA, γ-aminobutyric acid.

^aThe term “agonist” is used here to refer to agonists, partial agonists, and positive allosteric modulators.

^bwithdrawn due to increase in adverse cardiovascular events.

Cholinergic Targets

Acetylcholine is known to play an important role not only in motor function but also in various domains of cognition, particularly attention, learning, and memory.²³ Indeed, cholinergic dysfunction has been shown to be central to the pathophysiology of Alzheimer's disease and has also been postulated to contribute to the cognitive deficits of various neuropsychiatric disorders, including schizophrenia.²³ The basal cholinergic complex sends widely diffuse afferents through 2 projections: the septohippocampal and the nucleus basalis of Meynart cortical pathways.²⁴ The septohippocampal pathway is important in working memory processes through hippocampal storage of intermediate-term memory,^25,26 whereas the nucleus basalis of Meynart cortical pathway is involved in reference memory through long-term information storage in the neocortex.^27,28 Additionally, a role for acetylcholine in the processes of attention has been demonstrated in rats and monkeys.²⁹ Pharmacologically, anticholinergic drugs, like scopolamine, produce learning impairments in healthy subjects similar to those of persons with dementia,³⁰ while cholinomimetic drugs, like physostigmine, can significantly enhance the memory functions of healthy individuals.^30,31

Although the degeneration of cholinergic neurons in the basal forebrain and the associated loss of cerebral neurotransmission that is seen in Alzheimer's disease are absent in schizophrenia,^32,33 there is evidence of decreased nicotinic³⁴ and muscarinic³⁵ acetylcholine receptors in the cortex and hippocampus of individuals with schizophrenia. Interestingly, in patients with schizophrenia, decreased activity of choline acetyltransferase, a biosynthetic enzyme for acetylcholine production, was correlated with poorer cognitive functioning as measured by the Clinical Dementia Rating.³³ In addition, some of the atypical antipsychotics, but not typical antipsychotics, can increase the release of acetylcholine in the prefrontal cortex, possibly contributing to their modest enhancement of cognition in schizophrenia.³⁶ Thus, various targets within the cholinergic system are being investigated as potential enhancers of cognition in schizophrenia.

Cholinesterases

Cholinesterase inhibitors, such as donepezil and rivastigmine, are currently the main pharmacologic approach to the treatment of Alzheimer's disease and have been shown to slow the cognitive decline in this neurodegenerative disease.³⁷ As such, it has been proposed that cholinesterase inhibitors may also be useful in the treatment of the cognitive dysfunction in schizophrenia.³⁸ Acetylcholinesterase and butyrylcholinesterase are present in a wide variety of tissues and are broadly distributed in the brain. Inhibition of cholinesterases increases the synaptic concentration of acetylcholine, thereby enhancing and prolonging the action of acetylcholine on both muscarinic and nicotinic receptors. Therefore, cholinesterase inhibitors act as indirect cholinergic agonists at muscarinic and nicotinic receptors.

Following the administration of atypical antipsychotic treatment, cholinesterase inhibitors can increase the concentration of acetylcholine in the medial prefrontal cortex by 2- to 3-fold³⁶ and have been demonstrated to produce some functional normalization of brain activity during verbal fluency task performance of schizophrenic patients characterized by a significant increase in frontal lobe and cingulate activity on functional magnetic resonance imaging (fMRI).³⁹ As such, in recent years there have been multiple small randomized controlled trials of cholinesterase inhibitors in patients with schizophrenia, though results have been disappointing and inconsistent.⁴⁰ It has been suggested that the lack of consistent effects of cholinesterase inhibitors may be due to the high rate of cigarette smoking among patients with schizophrenia⁴¹ and subsequent desensitization of nicotinic receptors,⁴² thus rendering increased acetylcholine levels ineffective.⁴³ Indeed, galantamine, an acetylcholinesterase inhibitor that is also an allosteric potentiator of α₇ nicotinic receptors,^44,45 does not cause α₇ receptor desensitization and has been shown to enhance cognitive functioning of schizophrenic patients in a 4-week double-blind placebo controlled trial.⁴⁶ Interestingly, galantamine produced cognitive enhancement in schizophrenic patients despite the fact that all the patients smoked at least 10 cigarettes per day.⁴⁶ Unfortunately, despite these initial positive findings, subsequent results with galantamine have been mixed.^47–49 Thus, while pure cholinesterase inhibitors may be of minimal benefit for enhancing cognition in patients with schizophrenia, possibly due to desensitization of their α₇ nicotinic receptors from cigarette smoking, further study may be warranted with combined acetylcholinesterase inhibitors and allosteric potentiators of the nicotinic receptor in schizophrenia. However, it is likely that selective agents at various nicotinic and muscarinic receptors may be a more effective approach to developing drugs for treatment of the cognitive impairment in schizophrenia.

Nicotinic Receptors

It is well known that the smoking rates in individuals with schizophrenia are significantly higher than in the general population, and some have suggested that these individuals may be “self-medicating” with nicotine.^50,51 Indeed, nicotine administration has been shown to improve various measures of cognition that may ease some of the side effects of antipsychotic medications.⁵¹ For example, in patients with schizophrenia, a nicotine transdermal patch could dose dependently reverse haloperidol-induced impairments in working memory, attention, and reaction time⁵² and has been shown to reduce haloperidol-induced bradykinesia and rigidity compared with a placebo patch.⁵³ However, other studies have been mixed.⁵⁴ Interestingly, in one study,⁵⁵ some of the modest effects of cigarette smoking on clinical and cognitive outcome measures could also be improved by smoking denicotinized cigarettes, suggesting that nonnicotine components of cigarette smoke may also contribute.⁵⁵ Overall, while research on nicotinic treatment of individuals with schizophrenia has shown that, in single administrations, nicotine improves some aspects of cognition, additional administrations are not effective due to rapid desensitization of nicotinic receptors.⁴² Thus, considerable research is exploring the potential use of nicotinic agents, particularly partial agonists and allosteric modulators at various nicotinic receptor subunits that would be less likely to cause rapid receptor desensitization.

Nicotinic acetylcholine receptors are ionotropic receptors with a pentameric structure composed of alpha (α₂ to α₉) and beta (β₂ to β₄) subunits and are expressed at high levels in the hippocampus, cortex, striatum, and thalamus.^34,56 The 2 most prevalent nicotinic receptors are the α₄β₂, which is a high-affinity receptor, and the α₇, which is a low-affinity nicotinic receptor, both of which have been shown to have reduced numbers in patients with schizophrenia.^34,56 In addition, functional polymorphisms exist in the promoter region of the α₇ receptor that have shown genetic linkage in schizophrenia.^57,58

The α₇ nicotinic receptor subtype is a highly studied target for the development of drugs for cognitive enhancement. Studies in rodents have shown that antagonists at the α₇ nicotinic receptor induce sensory gating deficits similar to those seen in schizophrenia,⁵⁹ a hippocampal phenomenon manifested as an inability to attend appropriately to sensory stimuli.⁶⁰ Sensory gating deficits may strongly impact cognitive performance, and it has been shown that smoking transiently normalizes these sensory gating deficits.⁶⁰ In addition, agonists at α₇ receptors such as 3-2,4-dimethoxybenzylidene anabaseine (DMXB-A) can normalize the auditory gating deficits in rodents.⁶¹ Moreover, DMXB-A had a positive effect on a cognitive battery in a small proof-of-concept trial in humans,⁶² and additional clinical trials of α₇ receptor agonists are underway. However, there is concern that long-term use of α₇ agonists may induce the desensitization of nicotinic receptors, leading to a limited duration of efficacy.⁶³ Thus, further development of α₇ receptor partial agonists (eg, GTS-21) and allosteric potentiators (eg, galantamine) that induce minimal receptor desensitization is warranted.

In addition to α₇ receptors, it has been suggested that α₄β₂ nicotinic receptors are involved in cognition.⁶⁴ Indeed, α₄β₂ receptors are considered to represent more than 90% of the high-affinity nicotine-binding sites in the rat brain,⁶⁵ and decreased levels of α₄β₂ receptor binding have been found in the hippocampus of patients with schizophrenia.⁵⁶ Furthermore, agonists of α₄β₂ receptors, such as RJR2403 and SIB-1553A, can produce a significant and long-lasting improvement of memory in rats and monkeys.^66–68 Additionally, α₄β₂ agonists have been shown to stimulate the release of dopamine, norepinephrine, and acetylcholine in the hippocampus and frontal cortex in rats.⁶⁶ Thus, nicotinic α₄β₂ receptor agonists may be of therapeutic benefit for the treatment of the cognitive deficits in schizophrenia by several mechanisms. In addition to α₇ and α₄β₂ receptors, other nicotinic receptors, such as α₃- and α₆-containing receptors, may be involved in cognitive performance; however, studies are limited due to the lack of selective agonists and antagonists for these receptors.

Muscarinic Receptors

In addition to the ionotropic nicotinic receptors, numerous studies have implicated metabotropic muscarinic acetylcholine receptors in schizophrenia. Muscarinic receptors are G protein–coupled receptors⁶⁹ found widely throughout the central nervous system on both cholinergic and noncholinergic cells where they function as both autoreceptors and heteroreceptors.^70–72 Of the 5 genetically distinct subtypes of muscarinic receptors (M₁–M₅), the M₁ receptor has been most closely linked to cognition and schizophrenia.⁷³ Indeed, the M₁ receptor subtype is the most abundant of the muscarinic receptors in forebrain and hippocampus,^74,75 brain regions crucial to normal cognitive functions. In addition, decreased M₁ receptor binding has been reported in postmortem studies of the prefrontal cortex, hippocampus, and striatum from patients with schizophrenia,^76–78 that is, importantly, not due to chronic antipsychotic treatment.^35,77,79 Interestingly, M₁ receptor–deficient mice demonstrate deficits in working memory and remote reference memory indicative of impaired hippocampal-cortical interactions.⁸⁰ Together, these results suggest that alterations in central M₁ receptors may have a role in the pathophysiology of schizophrenia and that M₁ receptor agonists may be beneficial in treating schizophrenia, particularly the cognitive impairments.⁷³

Action at M₁ receptors has been proposed to be a major contributor to the cognition-enhancing effects of clozapine,⁸¹ despite the fact that clozapine is an exceedingly weak partial agonist at M₁ and other muscarinic receptors.^73,82 Thus, attention has focused on various clozapine metabolites including clozapine-N-oxide and N-desmethylclozapine (NDMC). Clozapine-N-oxide is inactive while NDMC has actions at many receptors (http://pdsp.med.unc.edu/pdsp.php).⁸² Interestingly, NDMC is the major active metabolic of clozapine and has been reported to be a potent M₁ agonist that preferentially binds to M₁ receptors vs clozapine,⁸³ although more comprehensive studies fail to demonstrate selectivity of NDMC for M₁ receptors.⁸⁴ In addition, NDMC has high affinities for 5-HT_2A and 5-HT_2C receptors and is a partial agonist at D₂, D₃ receptors^85,86 and δ-opioid receptors,⁸⁷ suggesting that this metabolite of clozapine may have antipsychotic and cognition-enhancing properties via a number of mechanisms. Furthermore, NDMC, but not clozapine, increases release of dopamine and acetylcholine in the prefrontal cortex and the hippocampus⁸⁸ and potentiates NMDA receptor activity in the hippocampus.⁸³ Thus, the cognitive enhancement observed with clozapine could actually be due to its metabolite NDMC via an uncertain mechanism.

Indeed, NDMC (ACP-104) and other M₁ receptor agonists are in clinical trials as potential treatments of the cognitive dysfunction in schizophrenia. Xanomeline, a nonselective M₁ and M₄ muscarinic agonist with potent actions at a variety of additional nonmuscarinic receptors including 5-HT_1A and 5-HT_2A receptors,⁸⁹ improved cognition and psychotic-like symptoms in Alzheimer's disease.⁹⁰ In addition, monotherapy with xanomeline resulted in an improvement of positive symptoms and cognitive function in 20 subjects with schizophrenia⁹¹ but was discontinued due to poor tolerability.⁹² The relatively nonselective actions of xanomeline and NMDC at a number of receptors (http://pdsp.med.unc.edu/pdsp.php) should engender caution among schizophrenia researchers for embracing positive data from xanomeline and NMDC studies as being specifically indicative of a role for M₁ receptors in schizophrenia.

Overall, evidence suggests that M₁ receptor agonists could be useful in treating various symptom domains in schizophrenia, though the roles of the other muscarinic receptor subtypes are less clear. M₅ receptors, for example, may be relevant to schizophrenia as they are located in the brainstem and midbrain, where they have an effect on dopamine release.⁹³ Indeed, xanomeline is also an antagonist at M₅ receptors, suggesting that blockade of M₅ may be involved in the benefits seen with xanomeline.⁹⁴ In addition, M₄ receptor knockout mice have impairments of cognitive performance and elevated levels of dopamine in the nucleus accumbens, suggesting a potential role for M₄ receptor agonists in treating both the positive and cognitive symptoms of schizophrenia.^95,96 As selective agonists at muscarinic receptor subtypes have been difficult to develop, positive allosteric modulators are also being explored as potential therapeutic agents.

Glutaminergic Targets

Glutamate is the primary excitatory neurotransmitter for approximately 60% of the neurons in the mammalian brain, including all cortical pyramidal neurons,⁹⁷ and plays a principal role in modulating long-term potentiation, a likely key cellular mechanism for learning and memory.^98,99 Glutamate mediates fast excitatory postsynaptic potentials by acting on 3 ionotropic receptors, which are differentiated based upon sensitivity to the synthetic glutamate derivatives N-methyl-D-aspartate (NMDA), alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA), and kainate.¹⁰⁰ In addition, glutamate exerts slower modulatory effects by acting on various G protein–coupled metabotropic glutamate receptors (mGluRs).¹⁰¹ For example, mGluR2 and mGluR3 receptors modulate the release of glutamate, whereas the mGluR5 receptor potentiates the duration of NMDA receptor–dependent excitatory postsynaptic potentials.¹⁰²

It has been hypothesized for decades that some deficiency in NMDA function might play a role in the pathophysiology of schizophrenia.¹⁰³ Since the 1950s, the NMDA receptor antagonists phencyclidine (PCP) and ketamine were known to produce a large range of schizophrenia-like symptoms including psychotic symptoms, negative symptoms, and cognitive dysfunction,^103–105 and it has been suggested that augmenting NMDA receptor activity may have therapeutic potential in schizophrenia.¹⁰⁰ Indeed, some of the atypical antipsychotics, but not typical antipsychotics, have been shown in preclinical models to reverse the effects of ketamine and PCP,^106–109 presumably through indirect activation of NMDA receptors mediated by other neurotransmitter systems.⁸³ It is also important to note that a competing hypothesis suggests that a hyperactivity of glutamatergic neurotransmission is involved in the psychopathology of schizophrenia, leading to seemingly contradictory pharmacologic approaches being explored.¹⁰⁰ Indeed, glutamatergic excitotoxicity is thought to be a factor in the neurodegeneration of Alzheimer's disease¹¹⁰ and a weak NMDA receptor antagonist, memantine, has shown efficacy in slowing cognitive decline in moderate to advanced Alzheimer's disease.¹¹¹ Thus, the glutamate system, especially its NMDA-dependent components, is complex, and while small increases in NMDA-dependent glutamate neurotransmission might be cognitively enhancing, too much activation may result in neurodegeneration. Fortunately, the glutamatergic excitatory synapse offers multiple targets for drug development to provide the precise level of enhancement to improve cognition without excitotoxicity. Thus, we briefly review below various approaches being explored for modulating NMDA receptor neurotransmission and discuss approaches aimed at other glutamatergic mediators.

NMDA Receptors

NMDA glutamate receptors are ligand-gated ion channels with a primary glutamate-binding site and an allosteric glycine-binding site.¹⁰⁰ Interestingly, the opening of the NMDA channel appears to require both glutamate and glycine binding, with glycine binding affecting channel open time and desensitization rate but not inducing channel opening itself.¹⁰⁰ In addition, NMDA receptor activity can be allosterically modulated by multiple other substances, including Mg²⁺, polyamines, and protons.¹¹² Thus, while direct agonists of the glutamate-binding site of the NMDA receptor may not be clinically feasible due to the risk of excess excitation and neurotoxicity, the allosteric sites on the NMDA receptor complex, particularly the glycine-binding site, are promising targets for drug development. Indeed, chronic treatment of rodents with glycine has not been found to induce excitotoxicity.^113,114

Compounds that target the glycine site of the NMDA receptor complex have been studied in multiple small clinical trials. These include glycine¹¹⁵ and D-serine,¹¹⁶ which are endogenous agonists at the glycine site of the NMDA receptor complex; D-alanine; and D-cycloserine, an antituberculosis drug that also binds to the glycine modulatory site where it functions as a partial agonist.¹¹⁷ In most of these studies, the test compound was administered along with either a typical or atypical antipsychotic, and there appears to be significant effects at reducing negative symptoms and cognitive impairment in patients with schizophrenia.¹¹⁸ Of the 4 agents, D-cycloserine has been the least efficacious, likely due to it being a partial agonist that acts as an antagonist at high doses. Interestingly, when used concurrently with clozapine, glycine¹¹⁹ and D-serine¹²⁰ have been reported to be ineffective while D-cycloserine seemed to worsen symptoms,¹²¹ possibly because clozapine may already enhance glycine and glutamate neurotransmission. Overall, agonists at the glycine allosteric site of the NMDA glutamate receptor hold promise in the treatment of the negative and cognitive symptoms of schizophrenia, possibly as an augmentation of currently existing antipsychotics. A potential limitation of targeting the glycine modulatory site is that the glycine site is probably half-saturated during physiologic conditions, suggesting that treatments targeting the glycine site would theoretically only be able to effectively double NMDA neurotransmission.¹¹⁸ In addition, both glycine and D-serine must be given at gram-level doses to significantly elevate central nervous system levels, and attempts to modify glycine or D-serine to produce synthetic glycine-site agonists, have, thus far, been unsuccessful. Thus, indirect approaches to activate NMDA receptors are being explored, such as increasing extracellular glycine and glutamate and by modulating AMPA receptors and mGluRs.

Glycine Transporters

An indirect approach being explored to boost NMDA activity via the glycine allosteric site is to increase synaptic glycine by inhibiting the glycine transporter. The glycine transporters, GlyT1 and GlyT2, have been identified on both neuronal and glial cells in the central nervous system, where they are suggested to control the extracellular concentration of glycine.¹²² Thus, blockade of glycine transporters is predicted to increase extracellular glycine and thus enhance NMDA receptor neurotransmission. Indeed, preclinical data suggest that inhibition of glycine reuptake represents a feasible approach to enhance NMDA receptor activity and possibly be therapeutic in schizophrenia.¹⁰⁰ For example, selective, high-affinity inhibitors of the glycine transporter, including Org-24598, N-[3-(4′-fluorophenyl)-3-(4′-phenylphenoxy)propyl] sarcosine, and SSR-504734, have been found to reverse PCP-induced hyperactivity and dopaminergic hyperreactivity in rodents.^123,124 Furthermore, the glycine transport inhibitor glycyldodecylamide attenuated PCP-induced hyperactivity more potently than administration of glycine.^125,126

Clinical trials to date, however, have only studied the low-potency glycine transport inhibitor sarcosine (N-methyl glycine). In a clinical trial of sarcosine added to the stable antipsychotic regimen of patient with schizophrenia, there was a highly significant reduction in negative symptoms, along with smaller but significant reductions in positive and cognitive symptoms.¹²⁷ Interestingly, a subsequent study with patients on clozapine found no improvement of symptoms with the addition of sarcosine, a result similar to studies with the NMDA glycine site agonists.¹²⁸ These results strongly suggest a role of glycine transport inhibitors in the treatment of schizophrenia, though results of trials with selective, high-potency inhibitors are anticipated.

AMPA/Kainate Receptors

Another glutamatergic approach to drug development for cognitive enhancement in schizophrenia has been the development of compounds that stimulate AMPA and kainate glutamate receptors. AMPA and kainate receptors mediate the majority of the fast glutamatergic signaling in the brain, and AMPA receptors work heavily in concert with NMDA receptors.¹⁰⁰ AMPA receptors provide the primary depolarization necessary to activate NMDA receptors, while NMDA receptors are required for proper incorporation of AMPA receptors into the postsynaptic membrane, a process involved in synaptic plasticity.¹²⁹ Thus, activation of AMPA receptors is likely critically important for learning and memory; however, direct AMPA agonists are unlikely to be therapeutically useful as AMPA receptors rapidly desensitize after stimulation.¹³⁰

In an attempt to avoid desensitization of AMPA receptors, allosteric potentiators of AMPA receptor function, a class of compounds termed ampakines, are being studied as potential treatments for enhancing cognition in schizophrenia.^131,132 Indeed, ampakines have been shown to enhance glutamatergic transmission and facilitate long-term potentiation in rodents.^133,134 Furthermore, ampakines improve performance in rodents on a variety of memory tasks including spatial mazes^135,136 and learned fear¹³⁷ and have been shown to be effective in reducing age-associated memory deficits in rats.¹³⁸ In a clinical trial of schizophrenia patients on clozapine, coadministration of the ampakine CX-516 yielded significant improvements in memory and attention;¹³⁹ however, a trial of CX-516 as monotherapy in schizophrenia showed no clear beneficial effects.¹⁴⁰ Importantly, higher potency ampakines are currently under clinical development as both monotherapy for schizophrenia and adjunctive treatment for cognitive dysfunction, though results of trials are not yet available. A higher potency ampakine, farampator, has been tested in healthy elderly volunteers and improved short-term memory but appeared to impair episodic memory,¹⁴¹ and thus, it remains unclear if modulation of AMPA receptors has therapeutic value in the treatment of the cognitive dysfunction in schizophrenia although this is a highly active area of current research.

Metabotropic Glutamate Receptors

Agents acting at mGluRs, which serve to regulate glutamatergic neurotransmission both pre- and postsynaptically, are currently in preclinical development for treatment of the cognitive dysfunction in schizophrenia. There are 8 subtypes of mGluRs (mGluR1–8) which are categorized into 3 groups according to their second messenger-coupling and ligand-binding profiles with group I receptors (mGluR1 and mGluR5) primarily being studied for cognitive enhancement in schizophrenia.¹⁰² Group I receptors, particularly mGluR5, function predominantly to potentiate both presynaptic glutamate release and postsynaptic NMDA neurotransmission,¹⁴² and mGluR5 receptors show significant colocalization with NMDA receptors in cortical pyramidal neurons.¹⁴³ Together, these findings suggest that mGluR5 agonism may enhance NMDA activity and improve memory and cognition.¹⁴² Indeed, selective mGluR5 agonists were found to inhibit PCP-induced dopamine release in the rodent prefrontal cortex.¹⁴⁴ Direct mGluR5 agonists, however, are likely to induce receptor desensitization limiting their therapeutic usefulness. Thus, selective allosteric potentiators of mGluRs have recently been developed and hold promise as therapeutic agents.^145,146 Indeed, preliminary positive results with an mGluR2/3 agonist in phase II trials have been reported (http://www.prnewswire.com/cgi-bin/micro_stories.pl?ACCT=916306&TICK=LLY&STORY=/www/story/12-07-2006/0004487009&EDATE=Dec+7,+2006).

Dopaminergic Targets

Dopamine projections to the prefrontal cortex comprising the mesocortical dopamine system are essential for normal cognition.^147,148 Thus, it has been hypothesized that decreased dopaminergic neurotransmission in the prefrontal cortex contributes to the cognitive deficits observed in schizophrenia, especially those related to executive functions and working memory.^149–151 Indeed, postmortem and in vivo imaging studies have linked prefrontal dopamine dysfunction to cognitive impairment,^151,152 and various studies have demonstrated that direct and indirect dopamine agonists can improve prefrontal cortex cognitive functions in humans.^153,154 However, it seems that precise regulation of prefrontal dopaminergic tone is essential as, in addition to insufficient dopamine, excessive prefrontal dopamine (eg, resulting from acute stress) may be deleterious to cognition.^155–157 Thus, dopamine function in the prefrontal cortex seems to follow an “inverted-U” dose-response curve whereby increases or decreases from an optimal level result in cognitive impairment.¹⁵⁸ Additionally, indirect dopamine agonists could potentially exacerbate psychosis by increasing neurotransmission at mesolimbic dopamine D₂ receptors.

D₁ Receptors

Dopamine D₁ receptors are expressed at high levels on the distal dendrites of pyramidal neurons in the prefrontal cortex that are thought to be involved in working memory processes.^158,159 Indeed, evidence suggests an important role of dopamine D₁ receptors in the cognitive dysfunction of schizophrenia.¹³⁸ For example, there is a decreased level of D₁ receptor–like binding in the prefrontal cortex of drug-naive patients with schizophrenia as measured with positron emission tomography imaging, and this decrease was found to be correlated with the severity of negative symptoms and cognitive dysfunction but not with the severity of positive symptoms.¹⁶⁰ In addition, in nonhuman primates, chronic blockade of D₂ receptors results in a downregulation of D₁ receptors in the prefrontal cortex and consequently produces severe impairments in working memory.¹⁶¹ This downregulation of D₁ receptors may explain why long-term treatment with typical antipsychotics can contribute to the cognitive dysfunction in schizophrenia. Thus, significant efforts are underway to examine the possible role of D₁ receptor agonists in treating cognitive dysfunction in schizophrenia. Indeed, low doses of selective D₁ agonists, such as dihydrexidine, A77636, and SKF81297, have cognition-enhancing actions in nonhuman primates,^162–164 and short-term administration of the D₁ selective agonist, ABT-431, reversed the cognitive deficits in monkeys treated chronically with a D₂ receptor antagonist.¹⁶¹

Although novel compounds that, directly or indirectly, stimulate D₁ receptors may be of immense value in treating cognitive deficits in schizophrenia, several potential pitfalls may need to be overcome. First, D₁ receptor activity follows the “inverted-U” dose-response curve, where either too little or too much D₁ stimulation impairs working memory.¹⁵⁸ In addition, chronic treatment with a D₁ agonist may actually lead to the downregulation of D₁ receptors which could, potentially, worsen cognition in the long term. Thus, an optimized level of D₁ receptor activation at the apex of the “inverted-U” may be required to obtain maximal cognitive benefits,¹⁵⁷ which may be accomplished by partial agonists or an intermittent pattern of administration.^161,165 An additional obstacle is the powerful hypotensive effects of direct-acting D₁ agonists on peripheral D₁ receptors,¹⁶⁶ which may necessitate the use of indirect D₁-activating agents such as catechol-O-methyltransferase (COMT) inhibitors (see below).

D₄ Receptors

The high affinity of clozapine for dopamine D₄ receptors led to speculation that D₄ receptors may be clozapine's “magic receptor.”¹⁶⁷ Clinical trials of selective D₄ antagonists, however, have not demonstrated any appreciable efficacy in the treatment of acute schizophrenia,^168–170 though it is possible that D₄ receptor blockade in collaboration with action at other neurotransmitter receptors may be clinically beneficial. Indeed, studies of the physiological roles for the D₄ receptor are finding that D₄ receptors may play an important role in impulsivity and working memory.¹⁷¹

The mechanism by which D₄ blockade could improve cognition is not fully known,¹⁷² though D₄ receptors are present on both pyramidal neurons and GABA-producing interneurons in the prefrontal cortex and hippocampus,¹⁷³ areas important for cognitive function. Studies have demonstrated that activation of D₄ receptors decreases NMDA receptor activity in the hippocampus¹⁷⁴ and inhibits glutamatergic signaling in the prefrontal cortex.¹⁷⁵ Additionally, D₄ receptor knockout mice show enhanced activity of cortical pyramidal neurons, an effect mimicked in wild-type mice by administration of the D₄ antagonist PNU-101387G.¹⁷⁵ Together, these results suggest that D₄ antagonism may be a valuable approach to improve cognition in schizophrenia. Indeed, the D₄ antagonist NDG96-1 was reported to reverse PCP-induced deficits in object retrieval tasks in monkeys,¹⁷² and another D₄ antagonist, PNU-101387G, reversed deficits in the delayed response task induced by the pharmacologic stressor FG7142 (a benzodiazepine inverse agonist).¹⁷⁶ Interestingly, PCP-induced cognitive deficits are exacerbated by haloperidol, suggesting that strong blockade of other dopamine receptors may counter the beneficial effects of D₄ blockade on cognitive functioning.¹⁷²

Seemingly contradictory evidence also suggests that D₄ receptor agonists may improve cognitive function. For example, the selective D₄ agonist A-412997 showed dose-dependent improvement in social recognition in rats, a model of short-term memory,¹⁷⁷ and the D₄ agonist PD168077 was shown to facilitate memory consolidation of an inhibitory avoidance learned response in mice.¹⁷⁸ These effects have been hypothesized to be due to D₄ receptor modulation of inhibitory GABAergic signaling in the prefrontal cortex.¹⁷⁹ Indeed, the D₄ agonist PD168077 reduces GABA_A inhibitory currents in pyramidal neurons, which could be blocked by the D₄ antagonist L-745870 as well as clozapine.¹⁷⁹ Thus, in contrast to D₄ antagonist–induced enhancement of NMDA currents in the hippocampus, D₄ agonists may suppress GABA_A inhibitory currents in the prefrontal cortex and thereby indirectly enhance cortical excitability. Taken together, D₄ receptor–selective agents may be valuable in the treatment of the cognitive deficits in schizophrenia, though a balance between D₄ receptor modulation of prefrontal GABA_A and hippocampal NMDA receptors may be necessary.

Catechol-O-methyltransferase

COMT is a postsynaptic enzyme that methylates and thereby deactivates synaptically released catecholamines, particularly dopamine.¹⁸⁰ Historically, monoamine oxidase was considered the primary enzyme for the initial deactivation of synaptic dopamine,¹⁸¹ though mounting evidence suggests that COMT may be especially important for the breakdown of dopamine, particularly in the prefrontal cortex.¹⁸² For example, COMT knockout mice show increased baseline levels of dopamine, but not other catecholamines such as norepinephrine, specifically in the frontal cortex.¹⁸³ In addition, the COMT knockout mice also showed enhanced memory performance,¹⁸³ suggesting a potential role of COMT inhibition in improving cognition. Indeed, a selective, reversible inhibitor of COMT, tolcapone, has been reported to improve working memory in rodents¹⁸⁴ and has been shown to improve cognitive dysfunction in patients with advanced Parkinson's disease,¹⁸⁵ though use is limited due to a risk of liver failure.¹⁸⁶ Other COMT inhibitors are currently being investigated for treatment of the cognitive dysfunction in schizophrenia.

Interestingly, a common single nucleotide polymorphism in the gene encoding COMT, Val158Met, results in the transcription of a variant of the COMT enzyme with approximately 40% less enzymatic activity in humans.¹⁸⁷ The reduced COMT enzymatic activity associated with the Met variant presumably results in greater availability of dopamine in the prefrontal cortex and, thus, may improve cognition, hypotheses supported by findings from the COMT knockout mice,¹⁸³ and an fMRI study in humans.¹⁸⁸ Furthermore, the COMT locus at chromosome 22q11 has been identified as a susceptibility locus for schizophrenia in several linkage studies^189,190 and 2 meta-analyzes,^189,191 though this remains controversial.^192–194 However, accumulating evidence predicts that patients with schizophrenia who have the Met allele may have improved cognitive response to clozapine.¹⁹⁵ Interestingly, a recent proof-of-concept experiment in normal human subjects treated with tolcapone demonstrated significant improvements on measures of executive function and verbal episodic memory in individuals with a Val/Val genotype but a diminished performance of individuals with the Met/Met genotype.¹⁹⁶ Thus, overall, the potential of pharmacologic inhibition of COMT in the long-term treatment of the cognitive dysfunction in schizophrenia remains to be determined.

Serotonergic Targets

5-HT_2A Receptors

5-HT_2A receptors are particularly abundant in the pyramidal neurons from cortical layer V,¹⁹⁷ where they have been described to colocalize with NMDA glutamate receptors,^198,199 suggesting a role in modulating cognitive functions. Indeed, studies have demonstrated that 5-HT_2A receptors interact with postsynaptic density protein 95, a protein involved in anchoring NMDA receptors to postsynaptic densities,²⁰⁰ and it has been suggested that activation of 5-HT_2A receptors increases the release of glutamate onto pyramidal cells.²⁰¹ Interestingly, the selective 5-HT_2A receptor antagonist M100907 blocked the cognition-impairing effects of MK-801, an NMDA receptor antagonist.^202,203 Similar findings have been reported for another 5-HT_2A antagonist AC-90179.²⁰⁴ Taken together, these studies suggest an intimate association between NMDA and 5-HT_2A receptors and imply that drugs with potent 5-HT_2A antagonistic actions may prove beneficial at improving cognition in schizophrenia, perhaps by normalizing NMDA receptor functioning.²⁰³

In addition, 5-HT_2A receptors are located on dopaminergic neurons in the ventral tegmental area,²⁰⁵ where they may modulate dopamine neuronal activity.²⁰⁶ Indeed, it is likely that a predominant role of 5-HT_2A receptors in antipsychotic action is to modulate dopaminergic tone, particularly along the mesocortical pathway.^206–208 For example, clozapine, olanzapine, and ziprasidone, but not haloperidol or risperidone, can preferentially augment dopamine and norepinephrine release in the prefrontal cortex relative to the subcortical areas.²⁰⁹ In rats, however, repeated administration of the selective 5-HT_2A antagonist M100907 can alter the activity of midbrain dopamine neurons in rats, though there is disagreement as to whether cortical dopamine levels are potentiated²¹⁰ or inhibited.²¹¹ Thus, it has been hypothesized that the ultimate effect of 5-HT_2A antagonists on dopaminergic neurotransmission might be to stabilize it.^208,212

Clinical studies with selective 5-HT_2A receptor compounds have also demonstrated a role of 5-HT_2A receptors in cognitive functioning in schizophrenia. For example, in a study of 30 hospitalized patients with schizophrenia, administration of mianserin, a 5-HT_2A/2C and α₂-adrenergic antagonist, improved scores on the Automated Neuropsychological Assessment Metrics at 4 weeks, though no significant improvement was found on the Wisconsin Card Sorting Test.²¹³ These results suggest that 5-HT_2A receptor antagonism may improve cognitive function in schizophrenia,^214,215 though additional clinical studies are needed. However, because nearly all approved atypical antipsychotic drugs have potent 5-HT_2A antagonist actions, it is unlikely that adding on a drug with potent 5-HT_2A antagonism will provide any significant boosting of cognition in treated patients.²¹² The potential cognition-enhancing effects resulting from 5-HT_2A receptor antagonism with the currently available atypical antipsychotics may be masked by other drug actions, such as anticholinergic effects, known to cause cognitive impairment.²¹⁶

5-HT_1A Receptors

5-HT_1A receptors are densely concentrated in the hippocampus, lateral septum, amygdala, and cortical limbic areas, as well as both the dorsal and median raphe nuclei.^217,218 On raphe neurons, 5-HT_1A receptors function as somatic autoreceptors providing inhibitory feedback control of 5-HT release.²¹⁹ However, the highest concentrations of 5-HT_1A receptors are on cortical and hippocampal pyramidal neurons²²⁰ suggesting a role of 5-HT_1A receptors in mediating cognitive functions. 5-HT_1A receptors on cortical pyramidal neurons are colocalized with 5-HT_2A receptors,²²¹ though while 5-HT_2A receptor activation is excitatory, 5-HT_1A receptor activation inhibits pyramidal neurons.²²²

Because they act on different locations of the receptors, both 5-HT_1A partial agonists and full antagonists have shown a positive effect on cognitive activity in animals.²²³ Thus, 5-HT_1A partial agonists, presumably acting on pyramidal neurons, improve cognition in animals, while 5-HT_1A antagonists also improve cognition, probably by acting at the raphe autoreceptors.^208,223 Indeed, atypical antipsychotic drugs modestly enhance cognition, and several atypical antipsychotic drugs have 5-HT_1A partial agonist actions (eg, aripiprazole, clozapine, olanzapine, ziprasidone, quetiapine),^224,225 while others are 5-HT_1A antagonists (eg, risperidone and sertindole).²²⁴ Preclinical experiments also show that both 5-HT_1A partial agonists and antagonists can improve cognition. For example, intraprefrontal infusion of 8-OH-DPAT, a nonselective 5-HT_1A agonist, improved visual-spatial attention and decreased impulsivity in rats,²²⁶ while WAY100635, a 5-HT_1A antagonist, blocked the deleterious effects of MK-801 and NMDA antagonist in rats.²²⁷ Thus, the preclinical literature is mixed regarding whether 5-HT_1A agonists or antagonists enhance cognition.

Clinical data in humans are equally mixed. For example, activating 5-HT_1A receptors with a single dose of tandospirone, a 5-HT_1A partial agonist, diminished explicit memory function²²⁸ in demented patients, though chronic administration of tandospirone enhanced verbal memory in patients with schizophrenia.^229,230 Interestingly, the 5-HT_1A agonist NAE-086 actually induced hallucinations and nightmares in normal individuals after repeated doses,²³¹ suggesting that 5-HT_1A agonists may not be tolerated well in schizophrenic individuals. Taken together, these results demonstrate that additional clinical studies are needed but suggest that 5-HT_1A receptors may need to be differently modulated to optimally enhance cognition in various pathologic conditions (ie, dementia vs schizophrenia) and that 5-HT_1A partial agonists with a high level of efficacy may present a significant risk of exacerbating positive symptoms in schizophrenia.²³¹

5-HT₄ Receptors

Serotonin 5-HT₄ receptors are found at high densities in the hippocampus, frontal cortex, and amygdala, suggesting a role of these receptors in cognitive functions.^232,233 Indeed, 5-HT₄ receptors have been shown to be markedly decreased in patients with Alzheimer's disease.²³⁴ In addition, 5-HT₄ receptor agonists have shown promise in the treatment of cognitive impairment by enhancing cholinergic transmission in the hippocampus.^212,235 For example, a 5-HT₄ receptor partial agonist, SL65.0155, improved learning and memory performance in chemically induced rat model of amnesia,²³⁶ and this improvement may be due in part to lengthening of the excitatory postsynaptic potential in hippocampal CA1 pyramidal neurons.²³⁷ Interestingly, a recent study also showed that the activation of 5-HT₄ receptors in a neuronal culture inhibited the secretion of β-amyloid peptide and enhanced neuronal survival.²³⁸ Thus, while 5-HT₄ receptor–selective agonists are mostly being studied for their role in the treatment of Alzheimer's disease, they may also be of benefit in the treatment of the cognitive dysfunction in schizophrenia.

5-HT₆ Receptors

Several atypical antipsychotics, including clozapine and olanzapine, and some tricyclic antidepressants, such as amoxapine, amitriptyline, and clomipramine, were found to have high affinity for 5-HT₆ receptors^239–241 prompting significant efforts to understand its possible role in schizophrenia and other neuropsychiatric disorders. When antisense oligonucleotides were used to decrease the level of 5-HT₆ receptor expression in rats, the rats exhibited an increased number of yawns and stretches that could be blocked by atropine, suggesting a role of the 5-HT₆ receptor in the control of cholinergic neurotransmission.^242,243 In addition, the selective 5-HT₆ receptor antagonist SB-271046 has been shown to improve memory retention in the water maze test of spatial learning and memory.^244,245 Thus, it appears likely that 5-HT₆ receptors may have an important future role in the treatment of cognitive deficits in neuropsychiatric illnesses such as Alzheimer's disease and schizophrenia.²¹²

5-HT₇ Receptors

The 5-HT₇ receptor exhibits a distinct distribution in the central nervous system with relatively high levels in the thalamus, hypothalamus, and hippocampus and lower levels in the cortex and amygdala.^246–248 In addition to possible roles in regulating circadian rhythms and sleep,^249–251 5-HT₇ receptors may also have an important role in hippocampus-dependent functions such as learning and memory.²⁵² For example, 5-HT₇ receptor knockout mice have been found to exhibit a specific impairment in contextual fear conditioning in which the animal learns to associate the environment with an aversive stimulus, a process generally believed to require the hippocampus.²⁵³ Electrophysiological studies have also shown that 5-HT₇ receptor activation modulates the excitability and intracellular signaling of pyramidal neurons in the CA1 region of the hippocampus.^254,255 Additionally, in 5-HT₇ receptor knockout mice, there is a reduced ability to induce long-term potentiation in the CA1 region of the hippocampus.²⁵³ Together, these findings suggest an important role for the 5-HT₇ receptor in hippocampus-dependent functions, including learning and memory.²⁵² Thus, selective 5-HT₇ receptor activators might prove therapeutically useful for the treatment of the cognitive dysfunction of schizophrenia.²¹²

Other Neurotransmitter Targets

α₂-Adrenergic Receptors

The central noradrenergic system projects from the locus ceruleus to the prefrontal cortex where α₂-adrenergic receptors appear to play an important role in cognitive functioning.²⁵⁶ Indeed, treatment with the α₂-adrenergic receptor agonists, clonidine and guanfacine, has been shown to improve cognitive performance without exacerbating positive symptoms in small trials of patients with schizophrenia.^257,258 In addition, patients randomized to risperidone plus guanfacine showed significant improvement on tasks of working memory and attention compared with patients receiving typical antipsychotics plus guanfacine.²⁵⁸ However, clozapine and other atypicals have potent antagonist properties at α₂-adrenergic receptors,²⁵⁹ which may contribute to the atypicality of atypicals by preferentially enhancing dopaminergic transmission in the frontal cortex over subcortical dopaminergic pathways.²⁶⁰ Indeed, combined treatment of a typical antipsychotic with the highly selective α₂-adrenergic receptor antagonist, idazoxan, has been reported to produce a profile of antipsychotic activity similar to clozapine.²⁶¹ Thus, as α₂-adrenergic receptor activity may be important in developing new drugs for schizophrenia that can improve cognition, balancing α₂-adrenergic receptor activity to achieve both antipsychotic and procognitive efficacy may be challenging.

GABA_A Receptors

Appropriately synchronized GABA neurotransmission in the dorsolateral prefrontal cortex is required for adequate working memory,²⁶² suggesting that impairments in GABA-mediated inhibition could contribute to the cognitive impairments in schizophrenia. Indeed, postmortem studies have shown reduced GABAergic transmission in schizophrenia.^263–265 In addition, recent observations^266,267 have noted decreases in messenger RNA levels for glutamic acid decarboxylase 67, the synthetic enzyme for GABA, selectively in the prefrontal cortex of patients with schizophrenia. Interestingly, a recent study revealed that GABA alterations in the dorsolateral prefrontal cortex of schizophrenic patients may be restricted to certain cell classes, such as the chandelier cells, which synchronize the activation of pyramidal neurons via GABA_A receptor subtypes.²⁶⁸ Thus, the use of new benzodiazepine-like agents—selective for the α₂ subunit of the GABA_A receptor—in cognitive disorders could be both interesting and revealing. Indeed, there is evidence that reduced GABA neurotransmission in chandelier cells may be secondary to altered NMDA receptor function and could represent a “final common pathway” of prefrontal dysfunction in schizophrenia.²⁶⁹ Thus, drugs targeted to mitigate the disturbances in inhibition might be particularly effective in improving cognitive performance in schizophrenia. For example, positive allosteric modulators selective for GABA_A receptors containing α₂ subunits (eg, a GABA_A α₂-selective benzodiazepine) may improve working memory function in schizophrenia.²⁷⁰ However, drugs that directly activate α₂-containing GABA_A receptors independent of the presence of GABA may disrupt the critical synchronization of this circuit and impair working memory.²⁶⁹ In addition, activation of GABA_A receptors containing other subunits (eg, α₁ or α₅), such as by currently available benzodiazepines, may impair cognitive function. Indeed, a recent study in healthy volunteers showed that, contrary to lorazepam, a GABA_A α₂/α₃ subtype–selective partial agonist, TPA023, caused no detectable memory impairment.²⁷¹

As activation of GABA_A receptors containing α₅ subunits, such as by currently available benzodiazepines, may impair cognitive function and cause sedation, inhibitors of these receptors have been hypothesized to enhance cognition. Indeed, functionally selective inverse agonists at α₅-containing GABA_A receptors have been demonstrated to enhance performance in animal models of cognition,^272–274 apparently without lowering the seizure threshold as seen with nonselective GABA_A inverse agonists.^274,275 In addition, α₅ subunit knockout mice demonstrate increased hippocampal activity due to the release of tonic GABAergic inhibition.²⁷⁶ Thus, antagonists or inverse agonists at α₅-containing GABA_A receptors may hold promise in the treatment of the cognitive dysfunction in schizophrenia.

Neurosteroids and Sigma Receptors

The sigma (σ) receptor was initially designated as a subtype of opioid receptors²⁷⁷ but was later found to be a distinct pharmacological entity due to lack of binding of the classical opiate receptor antagonists naloxone and naltrexone.²⁷⁸ Indeed, when the σ1 receptor was isolated and cloned, it was found to have no structural similarity to the opioid receptors.²⁷⁹ The functions of these receptors are poorly understood and endogenous ligands have yet to be identified, though it has been proposed that steroid hormones (eg, progesterone and testosterone), drugs of abuse (eg, cocaine, heroin, PCP), and psychiatric drugs (haloperidol, imipramine, and sertraline) may interact with σ receptors.^280,281 In addition, it is well documented that σ1 receptor ligands increase the NMDA receptor response in the hippocampus,^282–284 suggesting a role in enhancing cognition. Indeed, σ1 receptor agonists can reverse the memory impairments induced by the NMDA antagonists MK-801 in rodents.²⁸⁵

Neurosteroids, such as dehydroepiandrosterone (DHEA) and allopregnanolone, have been implicated in neuroprotection^286–288 and enhancement of NMDA receptor neurotransmission,^287,289,290 possibly through interaction with σ1 receptors,²⁹⁰ suggesting therapeutic potential for enhancing cognition in schizophrenia. Consistent with the enhancement of NMDA neurotransmission, DHEA can enhance memory in rodents.^291–294 In humans, a double-blind study of DHEA as an adjunct to antipsychotic treatment in chronic schizophrenic patients with prominent negative symptoms suggests some efficacy at improving negative symptoms, especially in women,²⁹⁵ though further studies are needed. Thus, neurosteroids may have therapeutic potential for improving the cognitive deficits observed in schizophrenia, though long-term treatment with steroids is problematic. In addition, while the contribution of σ receptor agonism to the actions of neurosteroids is not entirely known, highly selective σ receptor agonists are needed and hold therapeutic promise.

Potential Future Targets

There are a number of additional largely theoretical pharmacotherapeutic approaches for the treatment of cognition and schizophrenia. For example, significant progress has been made in recent years on elucidating various susceptibility genes in schizophrenia, including dysbindin, neuregulin 1, COMT, DISC1, and others.²⁹⁶ Interestingly, many of these genes appear to be related to the control of synaptic plasticity and glutamate transmission (particularly NMDA receptor function) and thus may allow for hypothesis-driven approaches for developing of actual disease-modifying drugs for schizophrenia and cognitive disorders.²⁹⁷ Another strategy involves the role of neurotrophic factors in the pathophysiology of schizophrenia and the theory that schizophrenia may involve a neurodegenerative process.²⁹⁸ Neurotrophic factors, such as brain-derived neurotrophic factor, may play a role in neuronal and glial differentiation, proliferation, and regeneration and influence synaptic organization, neurotransmitter synthesis, and the maintenance of synaptic plasticity.^299,300 Thus, strategies to enhance neurotrophic factor action may be able to prevent progression of schizophrenia.

Altering neurotransmitter signaling by targeting intracellular signaling cascades has long been suggested to be a future approach to novel therapeutic agents.³⁰¹ Though there has been concern about the feasibility of this approach, lithium is a signal transduction modifier that has been used safely for decades. Some targets being investigated include protein kinase C isoforms and glycogen synthase kinase 3.³⁰² In addition, subtype-selective phosphodiesterase (PDE) inhibitors, particularly at PDE10A, are actively being explored for the treatment of various symptom domains in schizophrenia.³⁰³ Another interesting approach at the receptor level would be developing ligands that differentially activate the various signaling pathways mediated by a single receptor, a process termed “functional selectivity”.³⁰⁴ Indeed, functional selectivity has been described in serotonin, opioid, dopamine, vasopressin, and adrenergic receptor systems³⁰⁴ and may be initiated by different ligand-induced conformational states, as shown for the β₂-adrenergic receptor.³⁰⁵ Thus, the possibility of selecting or designing novel ligands that differentially activate only a subset of receptor functions is intriguing as an approach to drug discovery that may optimize therapeutic action.

Conclusions

Cognitive dysfunction is a major feature of schizophrenia that contributes significantly to the long-term functional impairment that patients experience. While the past half-century of antipsychotic development has had a profound effect on the treatment of schizophrenia, the cognitive deficits in schizophrenia have been insufficiently addressed. Therefore, it is critical to continue the pursuit of diverse molecular targets for discovering new pharmacotherapeutic agents for the treatment of schizophrenia. For the past 20 years, psychopharmacologic research in schizophrenia has aimed for the development of new antipsychotic drugs with a more rapid onset of action, lower risk of side effects, and improved efficacy in the domains of negative and cognitive symptoms from a single compound. Currently, however, it seems unlikely that a single drug will have the desired effect across all these domains, and thus, optimal treatment of schizophrenia will likely rely on individualized polypharmacy and augmentation strategies. The ultimate goal, of course, will be the development of “cure therapeutics”²⁹⁷ which will require significant advances in our understanding of the underlying pathophysiology of schizophrenia, highlighting the need for continued basic research efforts at identifying and validating diverse and novel molecular targets.

Funding

National Institutes of Health (RO1MH61887 to B.R., RO1MH57635 to B.R.); National Institute of Mental Health (NO1MH32004 to B.R.).