Regulation of Glutamatergic Activity via Bidirectional Activation of Two Select Receptors as a Novel Approach in Antipsychotic Drug Discovery

Schizophrenia is a mental disorder that affects approximately 1–2% of the population and develops in early adulthood. The disease is characterized by positive, negative, and cognitive symptoms. A large percentage of patients with schizophrenia have a treatment-resistant disease, and the risk of developing adverse effects is high. Many researchers have attempted to introduce new antipsychotic drugs to the clinic, but most of these treatments failed, and the diversity of schizophrenic symptoms is one of the causes of disappointing results. The present review summarizes the results of our latest papers, showing that the simultaneous activation of two receptors with sub-effective doses of their ligands induces similar effects as the highest dose of each compound alone. The treatments were focused on inhibiting the increased glutamate release responsible for schizophrenia arousal, without interacting with dopamine (D2) receptors. Ligands activating metabotropic receptors for glutamate, GABAB or muscarinic receptors were used, and the compounds were administered in several different combinations. Some combinations reversed all schizophrenia-related deficits in animal models, but others were active only in select models of schizophrenia symptoms (i.e., cognitive or negative symptoms).


Introduction
Schizophrenia is one of the most complicated mental disorders, and it is characterized by different symptoms that may enrich or disrupt normal behavior. Particular symptoms are not equally manifested in patients, and at least four groups of patients with schizophrenia have been described. However, diagnostic manuals (DSM-V and ICD-11) have recently abandoned the use of schizophrenia subtypes, as they are not stable over time, have low diagnostic value, and substantially reduce the heterogeneity of schizophrenia [1,2]. Separate diseases characterized by schizophrenia-like symptoms have also been specified. The manifestation, intensity, and occurrence of particular symptoms differ between groups (Table 1).
A large percentage of patients with schizophrenia suffer from cognitive impairments that substantially influence daily functioning. Patients with severe cases of schizophrenia or individuals with the predominant presentation of negative and cognitive symptoms are generally treatment-resistant. Other patients with schizophrenia, who respond relatively well to antipsychotic medications, develop adverse effects that lead to discontinuation of the treatment. These factors make living with schizophrenia difficult or impossible. In contrast to other mental diseases, such as depression or anxiety, the effectiveness of psychotherapy as an add-on treatment to antipsychotic medication is very limited [4,5]. Table 1. Groups of symptoms and symptom intensity in patients with schizophrenia, schizoaffective 31disorder, and psychotic disorder, where "−"no symptoms, "+"-very mild, "++"-mild, "+++"-moderate, "++++"-severe," +++++"-very severe (based on [3] The search for new treatment strategies for schizophrenia began years ago, but no spectacular achievements have been reported. This lack of success may be partially due to the ambiguous, unspecified, and complex causes of schizophrenia arousal. The specific changes responsible for schizophrenia development that contribute to the manifestation of particular symptoms have not been fully determined. For many years, the dopaminergic theory of schizophrenia dominated the field and indicated increased dopaminergic neurotransmission as the main factor responsible for the pathophysiology of the disease [8]. The theory was proposed based on observations that dopaminergic antagonists reversed the psychotic symptoms of schizophrenia [13][14][15]. The lack of effectiveness of dopamine-based drugs towards negative and cognitive symptoms of schizophrenia caused doubts regarding the theory and indicates obvious shortcomings of the hypothesis and limits of the treatment. Further research indicated that changes in dopaminergic neurotransmission were not necessarily crucial in schizophrenia arousal. At least two groups of patients were distinguished that differed in their responsiveness to treatment [16]. These groups were normodopaminergic and hyperdopaminergic subpopulations of patients. The latter group had a better response to neuroleptic medications [16]. Genetic predispositions were also indicated as important in successful treatment [17][18][19]. The observations that NMDA receptor antagonists, such as PCP, ketamine, or dizocilpine (MK-801), induced the full spectrum of schizophrenia symptoms prompted the development of the glutamatergic hypothesis of schizophrenia [20][21][22][23]. One of the first papers describing its more important relevance was released in 1987 by Javitt et al., who reviewed studies showing the induction of negative symptoms of schizophrenia in healthy subjects and animals after PCP administration and proposed a novel hypothesis of schizophrenia [24]. Other studies also presented this hypothesis and suggested that preferential hypofunction of NMDA receptors expressed on GABAergic postsynaptic sites led to a decrease in the sensitivity of these neurons to the stimulatory effect of glutamate [25,26]. Consequently, the synthesis and release of GABA becomes impaired, and the subsequent inhibitory control over glutamatergic neurons is lost. The resulting increase in glutamate release is the proposed primary cause of schizophrenia development and results from the hypofunction of NMDA receptors at critical sites in local circuits that modulate the function of a particular brain region or control projections from one region to another (e.g., hippocampal-cortical or thalamocortical projections) [25,26]. This increased glutamate efflux under specific conditions or individual predisposition results in subsequent changes in other neurotransmitters, e.g., dopamine [15].
The formulation of this theory provided new possibilities in the search for treatment strategies based on the reduction of enhanced glutamatergic transmission. Naturally occurring full or partial agonists at the glycine modulatory site of the NMDA receptor, such as glycine, d-serine, and d-cycloserine, and a glycine transporter inhibitor with low affinity, sarcosine, were investigated in add-on studies to ongoing antipsychotic treatment and primarily focused on persistent negative symptoms [27]. Improvements in negative symptoms, sometimes with improvements in cognitive and positive symptoms, were noted [28][29][30][31][32][33], although subsequent meta-analyses did not confirm these results [27,34]. However, the activation of NMDA-dependent pathways with dopaminergic system inhibition and the activation/inhibition of accidental receptors confound the therapeutic effect and increase the risk of adverse effects.
The discovery of metabotropic glutamate (mGlu) receptors in 1989 showed the possibility of regulating glutamatergic neurotransmission without directly targeting NMDA ion channels.
Extensive research on the therapeutic potency of mGlu receptors and their distribution within the CNS is summarized in a vast number of review papers. A PubMed search of "schizophrenia" and "metabotropic glutamate receptors" retrieved more than 100 review papers. The most important reviews are shown in Table 2.

Glutamatergic Network in the Brain
Glutamate is the most abundant excitatory neurotransmitter in the brain, reaching high concentrations ranging from 5 to 15 µM per gram of tissue [104,105]. The activity of glutamatergic neurons is critical for the proper functioning of the cerebral cortex and the subcortical areas receiving glutamatergic projections.
Glutamatergic neurons are widely distributed across the CNS. At least five key glutamatergic pathways have been identified ( Figure 1) [106]. Three pathways descend from the cortex to subcortical structures, such as the brainstem, thalamus, nucleus accumbens, and striatum. One pathway ascends from the thalamus to the cortex. Intracortical loops of glutamatergic interneurons that stabilize the activity of cortical networks have also been identified. Similar loops have been observed in other brain areas, such as the hippocampus. 3. Regulation of Glutamate Release

Glutamatergic Network in the Brain
Glutamate is the most abundant excitatory neurotransmitter in the brain, reaching high concentrations ranging from 5 to 15 µM per gram of tissue [104,105]. The activity of glutamatergic neurons is critical for the proper functioning of the cerebral cortex and the subcortical areas receiving glutamatergic projections.
Glutamatergic neurons are widely distributed across the CNS. At least five key glutamatergic pathways have been identified ( Figure 1) [106]. Three pathways descend from the cortex to subcortical structures, such as the brainstem, thalamus, nucleus accumbens, and striatum. One pathway ascends from the thalamus to the cortex. Intracortical loops of glutamatergic interneurons that stabilize the activity of cortical networks have also been identified. Similar loops have been observed in other brain areas, such as the hippocampus. . Glutamatergic (red) and GABAergic (green) pathways in the human (A) and rat (B) brain. "a" and "b"-cortico-brainstem pathway, "c"-cortico-striatal pathway, "d"-cortico-accumbens pathway, "e"-cortico-thalamic pathway, "f"-thalamo-cortical pathway, and "g"-cortico-cortical pathway.
Based on these connections, glutamate is crucial in the integration of neurotransmission in the brain, including the regulation of monoaminergic nuclei located in the brainstem and cholinergic neurotransmission originating from the pedunculopontine and laterodorsal tegmental nucleus [106,107]. This excitatory system remains under the inhibitory control of GABAergic neurotransmission in a type of homeostatic balance.
Based on these connections, glutamate is crucial in the integration of neurotransmission in the brain, including the regulation of monoaminergic nuclei located in the brainstem and cholinergic neurotransmission originating from the pedunculopontine and laterodorsal tegmental nucleus [106,107].
This excitatory system remains under the inhibitory control of GABAergic neurotransmission in a type of homeostatic balance.
GABAergic neurons are spread throughout the brain and form a network that connects with the excitatory system and regulates its functions ( Figure 1) [108,109].
A variety of specific mechanisms regulate the release of neurotransmitters. One of the most important mechanisms is the presynaptic regulatory mechanism of receptors expressed on axon terminals, which may involve autoreceptors activated by the transmitters released from the host neuron or heteroreceptors activated by neurotransmitters that are synthesized by other neurons.
The activation or inhibition of receptors localized on dendritic shafts and cell bodies (postsynaptic receptors) triggers an electrical signal by regulating the activity of ion channels. The influx of ions changes the membrane potential of a neuron and results in a signal that is transmitted along the axon to regulate other neurons and the neuronal network.
The most important aspects of the pre-and postsynaptic regulation of glutamatergic networks are summarized below. Attention was placed on receptors that are likely targets for antipsychotic drug discovery.

mGlu 2 Receptors
The mGlu 2 receptors are located at a distance from the synaptic cleft [110]. The glutamate potency at mGlu 2 receptors is high-0.3-20 µM-but mGlu 2 receptors are exposed to relatively low concentrations of glutamate under physiological conditions [110][111][112]. The receptors are negatively associated with adenyl cyclase activity, and their stimulation results in the inhibition of glutamate release [113].
The most intense staining for mGlu 2 receptors was detected in the neocortex and limbic cortical neurons, predominantly in the hippocampus, as shown in Figure 2A and Table 4A,B. The expression of the receptor at axon terminals was evident, but examples of postsynaptic expression of the receptor on the cell bodies and dendrites of Golgi cells in the cerebellum were also noticed [114]. influx of ions changes the membrane potential of a neuron and results in a signal that is transmitted along the axon to regulate other neurons and the neuronal network.
The most important aspects of the pre-and postsynaptic regulation of glutamatergic networks are summarized below. Attention was placed on receptors that are likely targets for antipsychotic drug discovery.

mGlu2 Receptors
The mGlu2 receptors are located at a distance from the synaptic cleft [110]. The glutamate potency at mGlu2 receptors is high-0.3-20 µM-but mGlu2 receptors are exposed to relatively low concentrations of glutamate under physiological conditions [110][111][112]. The receptors are negatively associated with adenyl cyclase activity, and their stimulation results in the inhibition of glutamate release [113].
The most intense staining for mGlu2 receptors was detected in the neocortex and limbic cortical neurons, predominantly in the hippocampus, as shown in Figure 2a and Table 4a,b. The expression of the receptor at axon terminals was evident, but examples of postsynaptic expression of the receptor on the cell bodies and dendrites of Golgi cells in the cerebellum were also noticed [114].
Some postmortem studies revealed a decrease in the expression of mGlu2 receptors in the hippocampus and increased expression in the prefrontal cortex of patients with schizophrenia ( Figure 2b, Table 3c). Ligands activating mGlu2 receptors inhibit the release of glutamate and have been extensively investigated as newer antipsychotics in animals and humans. A 2007 article showed the efficacy of a mGlu2/3 orthosteric agonist in patients with schizophrenia and provided hope for new treatment solutions [57], as described in the review "Schizophrenia drug says goodbye to dopamine" [115]. Unfortunately, the results from further clinical trials of mGlu2/3 orthosteric agonists were far from satisfactory, and work with the compound was ultimately discontinued. However, this decision may have been premature because the ligands displayed excellent activity in preclinical models [51,116] and some clinical studies [117,118]. Some postmortem studies revealed a decrease in the expression of mGlu 2 receptors in the hippocampus and increased expression in the prefrontal cortex of patients with schizophrenia ( Figure 2B, Table 3C).
Ligands activating mGlu 2 receptors inhibit the release of glutamate and have been extensively investigated as newer antipsychotics in animals and humans. A 2007 article showed the efficacy of a mGlu 2/3 orthosteric agonist in patients with schizophrenia and provided hope for new treatment solutions [57], as described in the review "Schizophrenia drug says goodbye to dopamine" [115]. Unfortunately, the results from further clinical trials of mGlu 2/3 orthosteric agonists were far from satisfactory, and work with the compound was ultimately discontinued. However, this decision may have been premature because the ligands displayed excellent activity in preclinical models [51,116] and some clinical studies [117,118].
The conflicting data may result from several factors, such as genetic diversity between humans or a prior history of antipsychotic treatment. Further studies with more homogenous groups of patients and/or without prior medical treatment are needed. Importantly, the poor oral bioavailability of the compounds due to their highly hydrophilic properties was shown to be one of the reasons for their poor efficacy in humans [57,119,120]. One of the solutions to improve the gastrointestinal absorption of compounds is to design prodrugs with better absorption properties. Peptide transporter 1 (PEPT1) regulates the bioavailability of various drugs, including some mGlu 2/3 agonists; therefore, Eli Lilly designed prodrugs to be absorbed by PEPT1 (LY544344 for LY354740 and LY2140023 for LY404039) [119,121]. The generation of these prodrugs resulted in significantly higher bioavailability of the prototypes [119,122]. However, higher exposure may induce toxicity in patients [123]. An ester-based lipophilic prodrug of another mGlu 2/3 agonist, MGS0008, was designed to avoid undesirable adverse effects [123]. MGS0274 besylate exhibited a 15-fold improvement in oral bioavailability compared to MGS0008, and its administration to patients was accompanied by fewer toxic effects caused by its unnecessary exposure [120,123,124].

Group III mGlu Receptors
The third group of mGlu receptors consists of the mGlu 4 , mGlu 7 , and mGlu 8 subtypes. All of these receptors are expressed presynaptically and are negatively associated with adenyl cyclase activity [110]. The potency of glutamate at mGlu 4 receptors is slightly lower than at mGlu 2 receptors (3-38 µM), and these receptors are mainly located in the center of the synaptic cleft [110,111], near the site of fusion with synaptic vesicles. Therefore, these receptors are exposed to high glutamate concentrations [112].
Similar to mGlu 2 [125], the mGlu 4 receptor is expressed predominantly on glutamatergic terminals that oppose other glutamatergic projection neurons [126,127]. At least two splice variants of mGlu 4 receptors were identified [128], and stimulation of these receptors resulted in antipsychotic efficacy in several studies [51,53]. The receptor is expressed at relatively low levels in the hippocampus and cortex, and the most intense mGlu 4 labeling is observed in the globus pallidus and cerebellum, as shown Figure 3 and Table 4A,B. Postmortem studies have not shown altered expression of mGlu 4 receptors in patients with schizophrenia (Table 3C).
The ability of mGlu 2/3 and mGlu 4 receptors to inhibit glutamate release in the cortex was confirmed in patch-clamp experiments, in which an orthosteric agonist or positive allosteric modulator (PAM) abolished the frequency (but not the amplitude) of DOI-induced spontaneous EPSCs [129][130][131].
The mGlu 7 and mGlu 8 receptors are the least recognized mGlu receptors. Five subtypes of mGlu 7 [132] and three subtypes of the mGlu 8 receptor were cloned [133]. Due to the limited number of available ligands activating or inhibiting these receptors, data on their pharmacological activity are scarce. Available publications indicate a lack of efficacy of activation of mGlu 7 receptors in animal models of schizophrenia [134]. However, the only available mGlu 7 PAM, AMN082, was only tested in MK-801-induced hyperactivity and DOI-induced head twitches. Therefore, the data are incomplete. In contrast, the efficacy of negative allosteric modulators of the mGlu 7 receptor was observed in a wide range of tests [135]. efficacy in several studies [51,53]. The receptor is expressed at relatively low levels in the hippocampus and cortex, and the most intense mGlu4 labeling is observed in the globus pallidus and cerebellum, as shown Figure 3 and Table 4a,b. Postmortem studies have not shown altered expression of mGlu4 receptors in patients with schizophrenia (Table 3c). The ability of mGlu2/3 and mGlu4 receptors to inhibit glutamate release in the cortex was confirmed in patch-clamp experiments, in which an orthosteric agonist or positive allosteric modulator (PAM) abolished the frequency (but not the amplitude) of DOI-induced spontaneous EPSCs [129][130][131].
The mGlu7 and mGlu8 receptors are the least recognized mGlu receptors. Five subtypes of mGlu7 [132] and three subtypes of the mGlu8 receptor were cloned [133]. Due to the limited number of The mGlu 7 receptor is a presynaptic receptor located on glutamatergic axons. However, mGlu 7 -like immunoreactivity was also observed on GAD-expressing neurons in the islands of Calleja or striatum, suggesting that the receptor is also a heteroreceptor on GABAergic neurons [136]. The functional roles of these receptors are not clear because their low affinity for glutamate stimulation at distant synapses by a diluted signal is doubtful.

Presynaptic Regulation of Glutamate Release-Heteroreceptors
Heteroreceptors are activated by neurotransmitters other than those synthesized by the neurons on which the receptors are expressed.
The large number of heteroreceptors involved in the regulation of glutamate release makes a discussion of each type challenging. According to recent data, GABA B and muscarinic M 4 receptors are of particular importance in the pathophysiology of schizophrenia and antipsychotic drug discovery.

GABA B Receptor
GABA B receptors, similar to group II and III mGlu receptors, are associated with adenyl cyclase activity and the inhibition of cAMP production. Glutamatergic terminals contain large numbers of this receptor, and its stimulation inhibits glutamate release [137]. Therefore, GABA B receptors, together with mGlu receptors, are one of the most important pathways regulating the release of glutamate. The GABA B receptor is found in all brain areas, and the receptor is expressed at relatively high levels in all brain structures. The labeling of the receptor is higher in the hippocampus and the cortex than in the striatum, with an additional increase in the hippocampus compared with the cortex ( Figure 4A and Table 4A,B).
Available postmortem studies revealed decreased expression of GABA B receptors in both the hippocampus and prefrontal cortex of patients with schizophrenia ( Figure 4B and Table 3B).
According to preclinical studies, the GABA B receptor is a promising target in antipsychotic drug discovery. The efficacy of PAMs of this receptor has been shown in animal models of positive, negative, and cognitive symptoms [137,138]. Notably, the use of PAMs instead of agonists is recommended because of the lower risk of developing adverse effects, such as myorelaxation or sedation, which may be induced after orthosteric agonist administration [139,140]. together with mGlu receptors, are one of the most important pathways regulating the release of glutamate. The GABAB receptor is found in all brain areas, and the receptor is expressed at relatively high levels in all brain structures. The labeling of the receptor is higher in the hippocampus and the cortex than in the striatum, with an additional increase in the hippocampus compared with the cortex (Figure 4a and Table 4a,b).
Available postmortem studies revealed decreased expression of GABAB receptors in both the hippocampus and prefrontal cortex of patients with schizophrenia ( Figure 4b and Table 3b).

Muscarinic M 4 Receptor
Recently, researchers investigating schizophrenia have focused on muscarinic receptors after the administration of xanomeline was reported to exhibit antipsychotic efficacy in patients with schizophrenia [141]. Xanomeline is a nonselective agonist of muscarinic receptors that preferentially binds to M 1 and M 4 receptors [142]. Therefore, this drug also induced adverse effects due to stimulation of peripherally expressed M 2 and M 3 receptors [143]. Treatment with selective ligands to activate muscarinic receptor subtypes that are preferentially expressed in the brain, such as M 1 , M 4 , or M 5 , should result in a lower risk of peripherally driven effects. The M 4 subtype is located at presynaptic sites and may be a heteroreceptor on glutamatergic terminals [144,145].
The M 4 receptor is negatively associated with adenyl cyclase activity. It functions as an autoreceptor in the striatum, but it is expressed as a heteroreceptor on glutamatergic axon terminals and regulates glutamate release, predominantly in the cortex and hippocampus [146][147][148][149][150][151]. Patch clamp recordings confirmed its ability to reduce excessive glutamate efflux in the cortex [152]. The expression of the receptor in the structures involved in schizophrenia pathophysiology is shown in Figure 5A and Table 4A,B. Postmortem studies indicate decreased expression of M 4 receptors in the hippocampus and parietal cortex of patients with schizophrenia ( Figure 5B and Table 3A).
According to preclinical studies, the GABAB receptor is a promising target in antipsychotic drug discovery. The efficacy of PAMs of this receptor has been shown in animal models of positive, negative, and cognitive symptoms [137,138]. Notably, the use of PAMs instead of agonists is recommended because of the lower risk of developing adverse effects, such as myorelaxation or sedation, which may be induced after orthosteric agonist administration [139,140].

Muscarinic M4 Receptor
Recently, researchers investigating schizophrenia have focused on muscarinic receptors after the administration of xanomeline was reported to exhibit antipsychotic efficacy in patients with schizophrenia [141]. Xanomeline is a nonselective agonist of muscarinic receptors that preferentially binds to M1 and M4 receptors [142]. Therefore, this drug also induced adverse effects due to stimulation of peripherally expressed M2 and M3 receptors [143]. Treatment with selective ligands to activate muscarinic receptor subtypes that are preferentially expressed in the brain, such as M1, M4, or M5, should result in a lower risk of peripherally driven effects. The M4 subtype is located at presynaptic sites and may be a heteroreceptor on glutamatergic terminals [144,145].
The M4 receptor is negatively associated with adenyl cyclase activity. It functions as an autoreceptor in the striatum, but it is expressed as a heteroreceptor on glutamatergic axon terminals and regulates glutamate release, predominantly in the cortex and hippocampus [146][147][148][149][150][151]. Patch clamp recordings confirmed its ability to reduce excessive glutamate efflux in the cortex [152]. The expression of the receptor in the structures involved in schizophrenia pathophysiology is shown in Figure 5A and Table 4a,b. Postmortem studies indicate decreased expression of M4 receptors in the hippocampus and parietal cortex of patients with schizophrenia ( Figure 5b and Table 3a).

Postsynaptic Regulation of Neuronal Circuits in Patients with Schizophrenia
The selection of receptors expressed on cell bodies and dendrites deserves attention in schizophrenia drug development. Their activation changes the neuronal potential and signal transduction along the axon terminal, which may affect distant neurons.

Postsynaptic Regulation of Neuronal Circuits in Patients with Schizophrenia
The selection of receptors expressed on cell bodies and dendrites deserves attention in schizophrenia drug development. Their activation changes the neuronal potential and signal transduction along the axon terminal, which may affect distant neurons.

mGlu 5 Receptor
The mGlu 5 receptor is a member of the group I metabotropic glutamate receptor family, and it has three splice variants [153]. In contrast to the group II and group III receptors, this subtype interacts with phosphatase C and stimulates inositol production via Gαq signaling.
The mGlu 5 receptor is expressed near NMDA receptors and is functionally linked via Shank and Homer proteins [154]. Therefore, the stimulation or inhibition of the mGlu 5 receptor influences NMDA-mediated signaling [155][156][157], indicating that the pharmacological manipulation of this receptor represents a high risk. Fortunately, Conn and coworkers identified that the modulation of NMDA currents was not critical for mGlu 5 pharmacology and discovered biased, selective potentiators of mGlu 5 receptors coupled to Gαq-mediated signaling but not mGlu 5 modulation of NMDAR currents or NMDAR-dependent synaptic plasticity in the rat hippocampus [158]. These ligands bind to sites distinct from the orthosteric (or endogenous) ligand, often with improved subtype selectivity and spatiotemporal control over receptor responses, which constitutes a novel therapeutic approach.
The mGlu 5 receptors generally function as postsynaptic receptors on dendritic spines and shafts, but they were also detected presynaptically on axon terminals in the cortex and hippocampus. Electron microscopy and immunocytochemical studies indicated that these neurons may synthesize GABA [159,160]. The receptor is widely distributed across the brain, including structures that are critical in schizophrenia arousal. The most intense labeling was observed in the hippocampus, followed by the cortex, and the lowest expression was observed in the striatum. A schematic of the distribution of this receptor within these structures in the healthy brain is shown in Figure 6A and Table 4A,B. In postmortem studies, the expression of mGlu 5 receptors was decreased in the prefrontal cortex and cerebellum ( Figure 6B and Table 3C). The data from the frontal cortex are inconclusive, as the expression is increased in some regions and decreased in others. The mGlu5 receptor is expressed near NMDA receptors and is functionally linked via Shank and Homer proteins [154]. Therefore, the stimulation or inhibition of the mGlu5 receptor influences NMDA-mediated signaling [155][156][157], indicating that the pharmacological manipulation of this receptor represents a high risk. Fortunately, Conn and coworkers identified that the modulation of NMDA currents was not critical for mGlu5 pharmacology and discovered biased, selective potentiators of mGlu5 receptors coupled to Gαq-mediated signaling but not mGlu5 modulation of NMDAR currents or NMDAR-dependent synaptic plasticity in the rat hippocampus [158]. These ligands bind to sites distinct from the orthosteric (or endogenous) ligand, often with improved subtype selectivity and spatiotemporal control over receptor responses, which constitutes a novel therapeutic approach.
The mGlu5 receptors generally function as postsynaptic receptors on dendritic spines and shafts, but they were also detected presynaptically on axon terminals in the cortex and hippocampus. Electron microscopy and immunocytochemical studies indicated that these neurons may synthesize GABA [159,160]. The receptor is widely distributed across the brain, including structures that are critical in schizophrenia arousal. The most intense labeling was observed in the hippocampus, followed by the cortex, and the lowest expression was observed in the striatum. A schematic of the distribution of this receptor within these structures in the healthy brain is shown in Figure 6A and Table 4a,b. In postmortem studies, the expression of mGlu5 receptors was decreased in the prefrontal cortex and cerebellum (Figure 6b and Table 3c). The data from the frontal cortex are inconclusive, as the expression is increased in some regions and decreased in others.
Stimulation of mGlu5 exerted antipsychotic-like activity in a vast range of animal models [51,53].

Muscarinic M1 Receptor
The M1 receptor is expressed in the cerebral cortex, hippocampus, thalamus, and striatum ( Figure 7A and Table 4a,b) [161][162][163][164], and it activates phospholipase C and MAPK in the cerebral cortex in mice [165]. The M1 receptor colocalizes with NMDA receptors in hippocampal pyramidal neurons, and the simultaneous activation of the M1 and NMDA receptors increases NMDA currents [166]. Deletion of the M1 receptor results in a partial impairment of long-term potentiation in the

Muscarinic M 1 Receptor
The M 1 receptor is expressed in the cerebral cortex, hippocampus, thalamus, and striatum ( Figure 7A and Table 4A,B) [161][162][163][164], and it activates phospholipase C and MAPK in the cerebral cortex in mice [165]. The M 1 receptor colocalizes with NMDA receptors in hippocampal pyramidal neurons, and the simultaneous activation of the M 1 and NMDA receptors increases NMDA currents [166]. Deletion of the M 1 receptor results in a partial impairment of long-term potentiation in the hippocampus [166], which is also reflected in behavior [166,167]. Despite the presence of intact hippocampus-dependent memory, M 1 -/-mice show a deficit in consolidation over time during contextual fear conditioning, as well as impairments in win-shift and social discrimination learning, which suggests a role for the M 1 receptor in cortex-dependent memory or hippocampal-cortical interaction [166]. M 1 receptor deletion leads to elevated basal striatal dopamine release and locomotor activity, which is further enhanced by amphetamine challenge [167,168]. interaction [166]. M1 receptor deletion leads to elevated basal striatal dopamine release and locomotor activity, which is further enhanced by amphetamine challenge [167,168]. The antipsychotic activity of M1 receptor ligands has not been extensively tested in preclinical studies. Our studies are some of the first to show activity in animal models of schizophrenia [169]. However, M1 ligand activity was observed in models of positive and cognitive, but not negative, symptoms of the disease [169,170].
Postmortem studies revealed decreased expression of M1 receptors in various regions of the cerebral cortex in patients with schizophrenia ( Figure 7b and Table 3a).

Muscarinic M5 Receptor
The M5 receptor accounts for approximately 2% of all muscarinic receptors in the brain [164], and it is the least studied muscarinic receptor. It is expressed in the hippocampus, hypothalamus, cerebral cortex, striatum, substantia nigra pars compacta and ventral tegmental area ( Figure 8 and Table 4a,b) [162,163,171]. It is also found on blood vessels in the brain [172,173]. The location of M5 receptors suggests a role in the regulation of dopamine release [174]. These receptors colocalize with D2 dopamine receptors in the substantia nigra pars compacta [171]. Due to the lack of selective M5 receptor ligands, the first preclinical studies were performed in mice lacking this receptor. The M5-/mice showed no changes in motor coordination or basal locomotor activity, and no significant changes in locomotor activity were observed after amphetamine administration [175]. Deletion of the M5 receptor did not affect animal social interactions but weakened sensory motor gating processes [172,176]. M5-/-mice also showed a memory impairment in the new object recognition test and the Y maze [172]. The memory impairment may be partially explained by morphological (reduced number of dendritic spines) and physiological (reduced expression of NMDA, AMPA, and kainate receptor subunits, reduced frequency of spontaneous postsynaptic potentials, reduced LTP, and neurotransmitter release disturbances) changes within the hippocampal formation [172]. As shown in our previous studies, a PAM of the M5 receptor exerted antipsychotic-like effects on models of positive and cognitive, but not negative, symptoms of schizophrenia [169,170]. The antipsychotic activity of M 1 receptor ligands has not been extensively tested in preclinical studies. Our studies are some of the first to show activity in animal models of schizophrenia [169]. However, M 1 ligand activity was observed in models of positive and cognitive, but not negative, symptoms of the disease [169,170].
Postmortem studies revealed decreased expression of M 1 receptors in various regions of the cerebral cortex in patients with schizophrenia ( Figure 7B and Table 3A).

Muscarinic M 5 Receptor
The M 5 receptor accounts for approximately 2% of all muscarinic receptors in the brain [164], and it is the least studied muscarinic receptor. It is expressed in the hippocampus, hypothalamus, cerebral cortex, striatum, substantia nigra pars compacta and ventral tegmental area ( Figure 8 and Table 4A,B) [162,163,171]. It is also found on blood vessels in the brain [172,173]. The location of M 5 receptors suggests a role in the regulation of dopamine release [174]. These receptors colocalize with D 2 dopamine receptors in the substantia nigra pars compacta [171]. Due to the lack of selective M 5 receptor ligands, the first preclinical studies were performed in mice lacking this receptor. The M 5 -/-mice showed no changes in motor coordination or basal locomotor activity, and no significant changes in locomotor activity were observed after amphetamine administration [175]. Deletion of the M 5 receptor did not affect animal social interactions but weakened sensory motor gating processes [172,176]. M 5 -/mice also showed a memory impairment in the new object recognition test and the Y maze [172]. The memory impairment may be partially explained by morphological (reduced number of dendritic spines) and physiological (reduced expression of NMDA, AMPA, and kainate receptor subunits, reduced frequency of spontaneous postsynaptic potentials, reduced LTP, and neurotransmitter release disturbances) changes within the hippocampal formation [172]. As shown in our previous studies, a PAM of the M 5 receptor exerted antipsychotic-like effects on models of positive and cognitive, but not negative, symptoms of schizophrenia [169,170].  Table 4a,b summarizes the available data on the expression of particular receptors in rodents and humans. Studies of protein expression were performed using immunohistochemistry, Western blotting and immunoprecipitation, and mRNA expression was investigated using in situ hybridization, PCR, or Northern blotting. All investigated receptors were widely expressed in structures that are important in schizophrenia arousal (e.g., cortex, hippocampus, and striatum).  Table 4A,B summarizes the available data on the expression of particular receptors in rodents and humans. Studies of protein expression were performed using immunohistochemistry, Western blotting and immunoprecipitation, and mRNA expression was investigated using in situ hybridization, PCR, or Northern blotting. All investigated receptors were widely expressed in structures that are important in schizophrenia arousal (e.g., cortex, hippocampus, and striatum). Table 4. The expression of muscarinic (M 1 , M 4 , and M 5 ), GABA (GABA B ), and metabotropic glutamate (mGlu 2 , mGlu 5 , mGlu 4 , mGlu 7 , and mGlu 8 ) receptors in the rodent (A) or human brain (B). Protein expression was determined using immunohistochemistry, Western blotting, and immunoprecipitation. The mRNA levels were assessed using in situ hybridization, PCR, or Northern blotting.
Comparisons of the intensity of receptor expression with the antipsychotic efficacy of ligands activating these receptors clearly show that the activity of the ligands does not necessarily correspond with the intensity of receptor expression in relevant structures. Therefore, orthosteric agonists or PAMs of mGlu 4 receptors exhibit excellent activity in animal models of schizophrenia [130,236], but these receptors are expressed at the lowest levels in the cortex and hippocampus compared to other brain areas [209,232]. Instead, the high expression of mGlu 4 receptors in the globus pallidus, where it is a heteroreceptor on GABAergic terminals, makes it a good target for anti-Parkinson drugs [237]. However, stimulation of these receptors may increase the risk of adverse effects on non-Parkinson patients. Much lower doses of mGlu 4 PAMs/orthosteric agonists were active in animal models of schizophrenia than in models of Parkinson's disease [237]. Therefore, the risk of inducing adverse effects during antipsychotic treatment appears to be relatively low.
The extensive expression of GABA B and mGlu 5 receptors in cortical structures and the hippocampal formation [187,190] and their lower expression in deeper brain structures positively correlate with the activity of their ligands in animal models of schizophrenia and exclusively support the use of these receptors as targets for antipsychotic drugs. The functional connection of mGlu 5 with NMDA receptors increases the risk of inducing adverse effects with activation of mGlu 5 receptors, but biased ligands may be a solution [53].
Despite the initial hopes for mGlu 2 receptors as antipsychotic drug targets, their expression in the cortex and hippocampus is relatively low [204].
The high expression of muscarinic receptors in structures related to schizophrenia arousal makes them excellent antipsychotic drug targets [163,238], and the efficacy of compounds activating these receptors was confirmed in animal models [152,169]. Of the three analyzed receptors, M 1 was expressed at the highest levels.
The direct stimulation of post-or presynaptic sites results in the regulation of a particular neuron, which subsequently affects the neurons it innervates. The mechanisms engaged in the stabilization of inhibitory-excitatory balance in the CNS that are responsible for the antipsychotic effects of compounds are schematically shown in Figure 9. The aim of successive psychotropic treatment is to maintain homeostatic balance in the brain. Due to the extraordinary complexity of the central nervous system and its sensitivity to external factors, the precision and sensitivity of pharmacological manipulations must be considered to avoid adverse effects due to the unnecessary effects on the neuronal pathways responsible for other brain activities and functions.

Strategies Based on Bidirectional Inhibition of Glutamate Release
The individual differences between subjects, the complexity of microcircuits that regulate basic processes and the expression of receptors within these microcircuits have not been fully recognized in patients with schizophrenia and may determine the effectiveness and safety of treatment.
Although several studies and clinical trials have been conducted, the treatment of negative and The aim of successive psychotropic treatment is to maintain homeostatic balance in the brain. Due to the extraordinary complexity of the central nervous system and its sensitivity to external factors, the precision and sensitivity of pharmacological manipulations must be considered to avoid adverse effects due to the unnecessary effects on the neuronal pathways responsible for other brain activities and functions.

Strategies Based on Bidirectional Inhibition of Glutamate Release
The individual differences between subjects, the complexity of microcircuits that regulate basic processes and the expression of receptors within these microcircuits have not been fully recognized in patients with schizophrenia and may determine the effectiveness and safety of treatment. Although several studies and clinical trials have been conducted, the treatment of negative and cognitive symptoms of schizophrenia remains unsatisfactory. Extensive research has been performed to develop new solutions, but spectacular success is lacking.
Exclusive stimulation of the receptors expressed in neuronal circuits involved in the pathophysiology of schizophrenia, without effects on dopaminergic neurotransmission and/or NMDA receptor-mediated signaling, should minimize the risk of adverse effects and improve the effectiveness of therapy. Our recent studies proposed a treatment based on the simultaneous stimulation of two receptors that are crucial for regulation of glutamatergic networks, and the results have been published [138,152,169,170,236,239]. In these studies, select combinations activating mGlu 2 /M 1 , mGlu 2 /M 5 , and mGlu 4 /M 4 were not shown to alter prolactin levels or locomotor activity [152,170], prompting us to speculate that the use of sub-effective doses of at least two ligands may be safer than the highest dose of each compound alone or in combination with D 2 -based drugs [169,170].
The studies were performed using ligands that activate the receptors described in the first part of this review, e.g., muscarinic M 1 , M 4 and M 5 , GABA B and metabotropic glutamate receptors (mGlu 2 , mGlu 4 and mGlu 5 receptors). Different combinations of ligands were used, and their efficacies were investigated by performing a vast range of tests in rodents that reflected the positive, negative, and cognitive symptoms of schizophrenia (Table 6).

Simultaneous Administration of Ligands Activating Receptors Associated with Adenyl Cyclase Activity
The investigated combinations of ligands and their efficacies in animal models are shown in Table 7. The best working pair of compounds with evident efficacy in models of the positive, negative, and cognitive symptoms of schizophrenia were ligands that activated mGlu 4 /M 4 receptors and mGlu 2 /M 4 receptors (although these drugs were not tested in the models of positive symptoms) [152,239]. The simultaneous activation of GABA B receptors with mGlu 4 or M 4 receptors was not effective in models of negative symptoms and/or cognitive decline [169,236], and thus these combinations are less attractive for the reversal of negative and cognitive symptoms. However, the simultaneous activation of GABA B /M 4 or mGlu 4 receptors may be safer and more effective in patients with positive symptoms because the treatment of positive symptoms using current neuroleptic drugs carries a high risk of adverse effects. Table 7. Efficacy of the investigated combinations of ligands in tests assessing antipsychotic activity in rodents: "+"-compounds reversed the induced disruptions, "−/+"-compounds showed a trend toward reversing the induced disruptions, and "−"-compounds had no effect on the induced disruptions. The synergistic effects of ligands with affinity for two different presynaptically located receptors may result from several factors:

Synaptic Localization
The receptors are localized on one axon terminal, putatively a glutamatergic terminal. The concomitant stimulation results in the inhibition of glutamate release, and the ligands may complement the action of the other ligand. The receptors may act separately or through heterodimer formation (for a detailed description, see Section 4.1.1) The receptors are localized on different nerve endings that innervate one brain area and/or several different structures. The receptors may complement the action of the other in that area, as shown in Figure 9.

Heterodimerization
As mentioned above, G protein-coupled receptors are known to form homo-and heteromeric structures. In the physiological state, mGlu receptors function as homodimers composed of two identical subunits, and each subunit may both bind the ligand and activate G-protein signaling (for a review see: Wieronska et al., 2016 [51]). The GABA B receptor functions as a heterodimer composed of two subunits, GABA B1 and GABA B2 . The subunits depend on each other, i.e., GABA B1 binds the ligand and GABA B2 activates the signal transduction pathway [240].

Simultaneous Administration of Ligands Activating Receptors Associated with Adenyl Cyclase and the Inositol Phosphate Signaling Pathway
As shown in Table 8, the activity of the combined administration of sub-effective doses of an allosteric agonist of M 1 or PAM of M 5 receptors with sub-effective doses of PAMs of mGlu 2 or GABA B receptors was observed in models of the cognitive symptoms of schizophrenia, but not in the models of positive symptoms [170]. No activity of the allosteric ligands of M 1 or M 5 receptors was observed in models of negative symptoms of schizophrenia [170]. Therefore, their combinations with ligands activating mGlu 2 or GABA B receptors were not tested.
The costimulation of GABA B -mGlu 5 receptors exhibited clear and evident efficacy in models of the positive, negative and cognitive symptoms of schizophrenia, which were comparable to the effects of the active dose of each ligand administered alone [138].
The expression of the receptors supports different mechanisms of the synergistic effects than the presynaptically expressed receptors.
As indicated above, the activation of mGlu 2 or GABA B receptors inhibits glutamate release. Therefore, the dual action involves an increase in the inhibition on the one hand and the inhibition of excitation on the other hand, which restores brain homeostasis. Table 8. Efficacy of investigated combinations of ligands in tests assessing antipsychotic activity in rodents: "+"-compounds reversed the induced disruptions and "−"-compounds had no effect on the induced disruptions.

Conclusions
The figures shown below (Figures 10 and 11) schematically illustrate the coexistence of particular types of receptors in select structures.
The benefits and advantages of the combined activation of two selected receptors are sufficient to support the use of this approach in the treatment of schizophrenia.
Neither of the proposed treatments are based on the inhibition of dopaminergic receptors. Therefore, it may be speculated that the treatments are less burdened with the induction of adverse effects such as motor coordination and prolactin levels that are typical for presently used typical and second-generation neuroleptics. Preliminary experimental results supporting such conclusions can be found in Cieslik et al. 2018, Cieslik et al. 2019, and Cieslik et al. 2020 [152,169,170].
The results presented in the studies by Cieslik et al. 2018; 2020 indicate that the combined administration of the highest doses of the compounds or the administration of the highest dose of one compound with a subactive dose of the other does not produce additive effects [152,170]. Thus, the dosage does not need to be increased, and subsequently, the risk of unnecessary exposure to a treatment to obtain a therapeutic effect is relatively low. This finding might indicate the limited risk of unexpected events or toxic effects due to the accidental administration of a double dose of medications, which is particularly important for the mGlu 4 or GABA B receptor . As stated above, the mGlu 4 receptor, which is expressed in striatopallidal pathways, is considered an antiparkinsonian target [46,260]. The overstimulation of the receptor in these brain areas may result in undesired effects that counteract the putative antipsychotic efficacy. On the other hand, overstimulation of the GABA B receptor may exert adverse effects, such as sedation [139,140,261].
Analyses of the figures show overlap in the expression of particular receptors in select brain areas. The activation of receptors that are expressed at lower levels, such as mGlu 2 or mGlu 4, together with other types of receptors that are expressed at higher levels may complement the efficacy of the other receptor.
Overall conclusions obtained from the results discussed above and the consequences of the simultaneous administration of two compounds are as follows: − the dose of each compound may be reduced and the antipsychotic-like efficacy is the same as the highest dose of each compound administered alone (this approach may potentially allow us to avoid putative adverse effects or unnecessary exposure of the prodrug to patients, as shown previously for mGlu 2/3 agonists); − the action of the combined treatment might be selective in specific areas and thus may target a specific group of symptoms; − the ligands administered in combinations may complement the action of the other ligand and compensate for possible receptor dysfunctions, activating both homodimers and heterodimers/heterocomplexes.

Conclusions
The figures shown below (Figures 10 and 11) schematically illustrate the coexistence of particular types of receptors in select structures.   The benefits and advantages of the combined activation of two selected receptors are sufficient to support the use of this approach in the treatment of schizophrenia.
Neither of the proposed treatments are based on the inhibition of dopaminergic receptors. Therefore, it may be speculated that the treatments are less burdened with the induction of adverse