The therapeutic potential of metabotropic glutamate receptor modulation for schizophrenia

Abstract

Accumulating evidence suggests that a dysregulation of the glutamatergic system exists in the brains of schizophrenia patients. The metabotropic glutamate (mGlu) receptors are being investigated as novel drug targets for this disease, and have shown promise in both preclinical and clinical studies. Activation of mGlu₅ receptors may be efficacious for several symptom domains (positive, negative, and cognitive) and the potential for targeting mGlu₅ receptors has been bolstered by recent research on mitigating toxicity profiles associated with mGlu₅ activation. Additionally, genetic profiling of schizophrenia patients suggests that genes encoding for mGlu₁ and mGlu₃ receptors are altered, prompting preclinical studies that have demonstrated potential antipsychotic and cognitive enhancing effects of agents that activate mGlu₁ and mGlu₃ receptors, respectively. Development of subtype-specific drugs for the mGlu receptors, such as allosteric modulators, could provide a path forward for more efficacious and tolerable therapeutics for schizophrenia.

Glutamate is the major excitatory neurotransmitter in the mammalian central nervous system (CNS). It mediates cellular signaling between brain cells by binding both ionotropic (i.e. N-methyl-D-aspartate (NMDA), a-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and kainate) [1] and metabotropic glutamate (mGlu) receptors [2], thus contributing to information processing through neuronal network activity and synaptic plasticity [3]. Pharmacological modulation of the glutamatergic system is therefore poised to restore neuronal signaling and provide therapeutic benefit in neurological disease states. Schizophrenia, a neuropsychiatric disorder with a prevalence of around 1% of the world population [4], is one such psychiatric disorder that is highly amenable to glutamatergic modulation [5, 6]. Schizophrenia is clinically diagnosed in the United States according the criteria set by the current edition of the Diagnostic and Statistical Manual of Mental Disorders 5 (DSM-V) [7]. Of the three main symptom clusters found in schizophrenia patients, current typical and atypical antipsychotic medications that target the monoamine receptors have shown efficacy toward the positive (e.g., auditory and visual hallucinations, delusions), but not negative (e.g., social withdrawal, flat affect) nor cognitive (e.g., impairments in executive function, working memory and cognitive flexibility) symptoms [8]. Unfortunately, the efficacy of these medications is met with significant unwanted extrapyramidal side effects and weight gain, resulting in a significant unmet clinical need for improved pharmacological therapeutics.

Ongoing basic research into the biology of glutamate receptors has yielded significant knowledge that can be leveraged to improve drug discovery efforts for schizophrenia. Although the etiology of the disease is unknown, considerable evidence suggests that NMDA receptor hypofunction may contribute to the pathophysiology of schizophrenia [9, 10]. For example, administration of the NMDA receptor antagonist phencyclidine (PCP) causes schizophrenia-like symptoms in healthy individuals [11], and the NMDA receptor antagonist ketamine can exacerbate symptoms in individuals with schizophrenia [12]. These studies suggest an important role for NMDA receptor signaling in the pathology of schizophrenia; however, modulation of NMDA receptors using orthosteric agonists produces neuronal excitotoxicity and seizures, prohibiting direct NMDA receptor targeting as a viable therapeutic option. Alternatively, targeting the mGlu receptors pharmacologically may allow for regulation of glutamatergic neurotransmission and more precise tuning of the excitatory/inhibitory balance of the brain.

The mGlu receptors are G protein-coupled protein receptors that are sub-divided into three groups (Group I: mGlu₁ and mGlu₅; Group II: mGlu₂, mGlu₃; Group III: mGlu₄, mGlu₆, mGlu₇, mGlu₈), based on sequence homology, agonist selectivity and secondary signaling cascades [13]. As pharmacology efforts have advanced, the ability to attain subtype specificity of mGlu receptors has been accomplished through development of allosteric modulators. Positive allosteric modulators (PAMs) act to potentiate endogenous glutamate signaling, while negative allosteric modulators (NAMs) reduce the receptor responsiveness to glutamate [14]. In the context of CNS disorders, drug discovery efforts have elucidated beneficial modes of action for allosteric modulators that target subtypes within each of the three groups. Hence, several mGlu receptor subtypes represent novel therapeutic targets with potential utility for treatment of schizophrenia.

Group I mGlu receptors

The group I mGlu receptors consist of mGlu₁ and mGlu₅, both of which are coupled to signaling proteins that stimulate phosphoinositide hydrolysis and phospholipase C [2]. For many years, mGlu₅ has been considered an appealing therapeutic target due to its interaction with NMDA receptors through structural connections with scaffolding proteins such as Homer and Shank [15], as well as functionally by potentiation of NMDA receptor responses [16]. Correspondingly, positive allosteric modulation of mGlu₅ has been shown to enhance long-term plasticity (both long-term potentiation (LTP) and long-term depression (LTD)) in the hippocampus, and has pro-cognitive effects in healthy rodents [17]. When mGlu₅ is genetically deleted from mice, NMDA receptor-dependent plasticity and learning is impaired [18]. The mGlu₅-knockout mice have also been shown to display a disruption in prepulse inhibition (PPI) [19], a model of sensory motor gating that is disrupted in schizophrenia patients, leading to the hypothesis that mGlu₅ may represent a therapeutic target for schizophrenia. Preclinical research has established that mGlu₅ PAMs could be efficacious for schizophrenia in assays with predictive validity [20–22]. Albeit promising, the development of compounds that target the mGlu₅ receptors for treatment of neurological disorders have been stymied by preclinical toxicology issues, possibly related to excessive NMDA receptor activation [23]. Mechanistic studies point to the possibility that subsets of mGlu₅ PAMs can be biased away from NMDA receptor current potentiation, thus mitigating excitotoxic/seizure-inducing profiles [24]. Interestingly, compounds such as VU0409551 that do not potentiate effects of mGlu₅ on NMDA receptor currents [25], retain their efficacy in rodent models, displaying antipsychotic-like and cognitive enhancing activity [26, 27]. These beneficial effects may be related to the ability of mGlu₅ PAMs to modify long-term plasticity through non-NMDA receptor related mechanisms such as enhanced excitability of pyramidal neurons [28], enhancement of long-term plasticity requiring protein synthesis [29], or possibly depression of inhibitory tone through local messengers such as cannabinoids [30]. Taken together, mGlu₅ represents a target that could be efficacious for positive, negative and cognitive symptoms of schizophrenia. However, this may be contingent on the potential to develop mGlu₅ PAMs without adverse side-effect liability.

While mGlu₅ has been heavily studied as a potential target for pharmacotherapy, mGlu₁ has received less attention. However, multiple mutations in the gene encoding for human mGlu₁ receptor, GRM1, have been identified in schizophrenia patients, raising the possibility that mGlu₁ may be dysregulated in schizophrenia [31]. This is supported by the finding that mGlu₁ mRNA expression is dysregulated in tissue of postmortem schizophrenia patients [32]. Interestingly, mutations that have been found to be associated with schizophrenia leads to a decrease in mGlu₁ signaling in-vitro, and a range of highly selective mGlu₁ PAMs can potentiate responses to activation of the mutant receptors [33], raising the possibility that mGlu₁ PAMs could reverse deficits in mGlu₁ signaling in schizophrenia patients that bear these mutations.

Preclinically, mGlu₁ knockout mice have a pathological phenotype that mirrors the symptomology of schizophrenia, such as deficits in PPI [34]. Given the hyperdopaminergic state of the schizophrenic brain [35, 36], drugs that antagonize the dopamine system have been a mainstay of the therapeutic treatment. Therefore, production of new drugs that act to dampen dopaminergic efflux would be predicted to have antipsychotic efficacy. It was demonstrated that group I mGlu activation depresses dopamine release in the dorsal striatum of mice using fast-scan cyclic voltammetry ex-vivo; an effect likely attributable specifically to mGlu₁ activation, as the presence of mGlu₅ antagonist or using mGlu₅–null mice did not affect group I mGlu agonist-evoked dopamine release [37]. Indeed, administration of a selective mGlu₁ PAM, VU6004909, causes dopamine release to be significantly attenuated in the dorsal striatum after electrical stimulation of the medial forebrain bundle in rats [38]. Incongruently, mGlu₁ NAMs are also beneficial in preclinical assays such as methamphetamine induced-hyperlocomotion and disruption of PPI [39]. Therefore, more research is necessary to elucidate the precise mechanisms by which mGlu₁ modulation shows efficacy in these assays. These data also highlight the important fact that underlying individual pathology such as the presence of GRM1 mutations may promote the utility of mGlu₁ specific drugs.

Group II mGlu receptors

The group II mGlu receptors, consisting of mGlu₂ and mGlu₃, couple to G-protein subunits that inhibit adenylate-cyclase activity [2]. Activation of these receptors by group II mGlu receptor orthosteric agonists has been shown to enhance the function of NMDA receptors [40], as well as regulate LTD and LTP in the prefrontal cortex and the hippocampus, respectively [41–43]. Furthermore, activation of group II mGlu receptors reduces excessive activity at glutamatergic synapses in the PFC that is induced by hallucinogens such as the 5-HT₂A receptor agonist DOI or administration of NMDA receptor antagonists [44, 45]. Therefore, these receptors are prime targets for the treatment of schizophrenia where NMDA receptor dysregulation and plasticity impairments are thought to be occurring in these brain areas. Indeed, seminal work testing target viability of the group II receptors for schizophrenia was conducted in rats and revealed that the mGlu_2/3 agonist, LY354740, was capable of rescuing deficits in stereotypy, locomotion, spatial working memory and cortical glutamate efflux induced by the NMDA-receptor antagonist PCP [46]. This finding was strengthened by the translation of mGlu_2/3 agonist efficacy in attenuating NMDA receptor-induced deficits in working memory within human subjects [47]. Clinically, the mGlu_2/3 agonist LY2140023, developed by Eli Lilly and Company, entered a double blind-placebo controlled phase II clinical trial and showed safety and tolerability; without weight gain or extrapyramidal side effects seen with typical antipsychotics. Most importantly, there was also a significant improvement in positive and negative symptoms of schizophrenia compared to placebo [48]. However, the effects of LY2140023 were washed out by a large placebo effect in a follow-up phase II study [49]. Combined with a subsequent placebo controlled study where the atypical antipsychotic risperidone improved scores on the Positive and Negative Symptom Scale, the mGlu_2/3 agonist LY2140023 failed to show efficacy [50]. While the reasons for clinical failure can be manifold, the importance of dose-response was exemplified by a recent non-human primate study that demonstrated mGlu_2/3 agonist enhancement of working memory follows an inverted-U dose-response, where low doses improve working memory but higher doses have varied effects [51]. Nonetheless, given the mixed results from these trials and others, Eli Lilly terminated the further development of this compound as a clinical candidate. While these trials were not successful, the fact that the drug did have benefit in some patients suggests that perhaps mGlu₂ or mGlu₃ specific drugs could yield better results clinically.

Selective mGlu₂ PAMs have been heavily pursued as drug candidates for schizophrenia given that preclinical data suggest that mGlu₂ primarily mediates the antipsychotic-like effects of mGlu_2/3 agonists in animal models. This is evidenced by the finding that the mGlu_2/3 agonist, LY314582, fails to reduce NMDA receptor-antagonist induced hyper-locomotion in mGlu₂ KO mice [52]. Several mGlu₂ PAMs, including LY487379 and biphenyl-indanone A (BINA), have shown efficacy in preclinical animal models of schizophrenia [53, 54] (for in-depth review see [55]). This research paved the way for mGlu₂ PAM testing in clinical trials. However, a proof of principle phase II study of AstraZeneca’s mGlu₂ PAM AZD8529 did not demonstrate efficacy of the drug candidate compared to placebo, while the atypical antipsychotic risperidone demonstrated significant efficacy [56]. These results taken together illustrate the difficulty of translating preclinical research findings, and highlight the potential shortcomings in animal modeling approaches. Certainly, the underlying pathology within clinical populations of schizophrenia patients is heterogeneous and disease etiology is likely to be multifaceted; possibly making activation of a single target insufficient to provide therapeutic benefit for a wide population sample. Therefore, the ability to stratify patient population based on biomarkers or genotype may enable pharmacotherapy to be tailored and boost efficacy for individual patients.

Mutations in GRM3, the gene encoding for mGlu₃, have been shown to be associated with schizophrenia and deficits in cognitive function [57, 58]. The generation of knockout mice and development of specific mGlu₃ tool compounds has enabled researchers to begin to uncover the basic functions of mGlu₃ in the CNS. Within the prefrontal cortex (PFC), an area of the brain important for working memory and cognition, mGlu₃ functions postsynaptically to induce long-term depression (LTD) of excitatory transmission as well as induce calcium mobilization [41] and increase pyramidal cell excitability through inhibition of postsynaptic cAMP-K⁺ channel actions [51]. Furthermore, mGlu₃ modulates plasticity within the hippocampus, a declarative memory center that is altered in schizophrenia [59]. mGlu₃ is required for induction of LTD in the dentate gyrus subregion [60], and facilitates LTP through post-synaptic mechanisms in the CA1 subregion [43]. However, under some conditions activation of mGlu₃ can inhibit induction of LTP. Specifically, activation of mGlu₃ in hippocampal astrocytes, during corresponding beta-adrenergic receptor activation leads to an attenuation of LTP enhancement through a novel form of glial-neuronal communication [42], suggesting that mGlu₃ modulates plasticity differently depending on the activity of the network.

Behaviorally, mGlu₃ alterations could contribute to phenotypes associated with schizophrenia given that mGlu₃ knockout mice show hyperactivity as well as deficits in reference memory and working memory tests [61], paralleling some known functions of the group I mGlu, mGlu₅. Interestingly, recent preclinical evidence uncovered intriguing interactions of mGlu₃ and mGlu₅, where activation of mGlu₃ potentiates mGlu₅ signaling. Furthermore, mGlu₃-dependent LTD in the PFC was shown to be dependent on mGlu₅ activity [62]. Combined with the fact that mGlu₃ may provide neuroprotection in response to NMDA receptor-mediated toxicity [63], a drug activating mGlu₃ may reduce the risk for cytotoxicity while improving excitatory signaling. Therefore, mGlu₃ represents a novel target for cognitive dysfunction in schizophrenia patients, and development of specific mGlu₃ potentiators could provide valuable adjunct therapy options.

Group III mGlu receptors

The group III receptors are coupled to signaling cascades that inhibit adenylate cyclase activity, and thus primarily function as presynaptic autoreceptors. Aside from mGlu₆, which is restricted to the retina, mGlu₄, mGlu₇, and mGlu₈ are all potential targets to normalize glutamatergic tone within the CNS of schizophrenia patients [13]. Basic research in the field is in its early stages of validating these receptors as specific targets, however, group III agonists such as ACPT-I possess antipsychotic-like effects in animal models by reducing PCP-induced hyperlocomotion in rodents [64]. In addition, the selective mGlu₄ PAM, ADX88178, can improve indicators of all three symptom clusters preclinically [65]. Evidence that mGlu₄ could heterodimerize with mGlu₂ in vivo [66] supports the possibility of targeting this mGlu_2/4 complex to provide even greater subtype selectivity; although the antipsychotic efficacy of such a drug has yet to be tested experimentally. While mGlu₇ and mGlu₈ both play a role in cognitive functions [67, 68] and glutamatergic signaling in the hippocampus [69, 70], direct evidence for these receptors as targets for schizophrenia is lacking. In fact, current data may suggest that mGlu₇ PAMs could exacerbate positive symptoms [71]. More research is certainly required using more extensive preclinical modeling before the value of these targets can be ruled out, especially as more advanced selective compounds become available.

Conclusion

The fact that the negative and cognitive symptom domains of schizophrenia are often insufficiently controlled in the clinical setting substantiates the need for new pharmacotherapies for this disorder. The mGlu receptors represent several novel targets for the treatment of schizophrenia, with receptors from each of the three mGlu sub-groups showing promise. Based on preclinical data, mGlu₅ PAMs that are devoid of toxicity and seizure liability could yield effective treatment for all three symptom domains. Furthermore, the genetic alterations in the genes encoding mGlu₁ and mGlu₃ within the schizophrenia patient population support recent preclinical data suggesting potential antipsychotic and cognitive enhancing attributes of drugs targeting these receptors. It will be of interest to see how future investigations expound upon this data. Together, the studies outlined in this review illuminate the ongoing basic and clinical research that will undoubtedly drive discovery efforts to make safer and more efficacious mGlu targeted drugs for this disorder.

Highlights.

• mGlu₅ PAMs are efficacious for positive, negative, and cognitive symptoms in animal models

• Clinical trials for mGlu₂ PAMs have not been successful

• Genetic studies have identified GRM1 and GRM3 as dysregulated in the schizophrenic brain

• mGlu₁ PAMs reduce dopamine release in preclinical studies and could be potentially beneficial for positive symptoms

• mGlu₃ activation appears to be pro-cognitive and neuroprotective preclinically