Cognitive Deficit in Schizophrenia: From Etiology to Novel Treatments

Schizophrenia is a major mental illness characterized by positive and negative symptoms, and by cognitive deficit. Although cognitive impairment is disabling for patients, it has been largely neglected in the treatment of schizophrenia. There are several reasons for this lack of treatments for cognitive deficit, but the complexity of its etiology—in which neuroanatomic, biochemical and genetic factors concur—has contributed to the lack of effective treatments. In the last few years, there have been several attempts to develop novel drugs for the treatment of cognitive impairment in schizophrenia. Despite these efforts, little progress has been made. The latest findings point to the importance of developing personalized treatments for schizophrenia which enhance neuroplasticity, and of combining pharmacological treatments with non-pharmacological measures.


Introduction
Schizophrenia is a major mental illness characterized by psychosis, apathy, social withdrawal and cognitive impairment [1]. The symptoms of schizophrenia have been placed into three categories: positive symptoms such as delusions and hallucinations [2]; negative symptoms such as affective flattening, alogia and diminished emotional expression; and cognitive deficit [3]. Although schizophrenia was considered an early dementia in the 19th century, the cognitive symptoms of schizophrenia have been largely neglected in the treatment of the disease [4]. However, in the last few years, there has been a growing interest in the study of cognitive impairment [5]. The reasons for this are that cognitive impairment is one of the first symptoms to manifest in schizophrenia, it is disabling for schizophrenic patients, and it contributes to their functional impairment [6]. Cognitive deficit cannot be treated with current antipsychotic drugs, which can only effectively treat positive symptoms of the disease [7]. Therefore, cognitive impairment in schizophrenia is still a clear unmet clinical need.
Cognitive deficit comprises an impairment in several cognitive domains, such as processing speed, attention, working memory, verbal and visual learning, problem solving and social cognition [8]. This impairment is found even in the first episode of schizophrenia [9].
Cognitive deficit is a process involving genomic, neurobiological and neuroanatomic factors that interact with each other in a complex way. The aim of this review is to go over the etiology of neurocognitive impairment in schizophrenia and the status of current research on the treatments for this symptom of schizophrenia.

Hormones
Cognitive deficit in schizophrenic patients has been linked to imbalances in serum levels of cortisol and prolactin.
Cortisol is a hormone that participates in the response to stress and inflammation. Moreover, cortisol easily crosses the blood-brain barrier, binding glucocorticoid receptors in prefrontal cortex, hippocampus and amygdala [29]. Higher cortisol levels are associated with lower hippocampal volume and lower BDNF expression [30,31]. Additionally, higher levels of cortisol and greater blunting of the cortisol awakening response (CAR) are related to poorer performance in cognitive tasks [32,33]. However, the influence of cortisol levels on neuroanatomic findings and in cognitive performance may depend on the sex of the patients and the clinical diagnosis of the patient [34].
Prolactin levels have also been linked to cognitive impairment. However, the study of prolactin levels in schizophrenic patients is difficult because typical antipsychotics, as D 2 receptor antagonists, increase prolactin concentrations in blood because they block D 2 receptors in the tuberoinfundibular pathway [35]. It has been observed that in antipsychoticnaive schizophrenia patients, higher levels of prolactin are related to poorer cognitive performance [36]. The reasons for this poorer performance are not clear, but it was observed that there is a decrease in grey matter volume in patients with hyperprolactinemia [37], and that prolactin levels are related to higher inflammatory markers [38].

Neurotrophic Factors
BDNF is a neuropeptide that enhances brain remodeling and synaptic plasticity. It was observed that reduced levels of BDNF are related to lower cognitive capacities in patients with schizophrenia [39,40]. This reduction in BDNF levels has also been observed in other pathologies related to cognitive deficit. Vasconcelos et al. observed that this reduction in BDNF is related to an increase in oxidative stress in schizophrenic patients, as the administration of alpha-lipoic acid, a drug that counteracts free radicals, increases BDNF and improves cognitive capacities in animal models of schizophrenia [41]. Indeed, clozapine, an antipsychotic drug that improves negative symptoms and cognitive capacities in schizophrenic patients, increases brain BDNF levels [42]. The influence of variants of BDNF in cognitive deficit has also been studied; one example is BDNF Val66Met, which is related to a worse cognitive function [43,44]. However, there is a great deal of variability between different studies because of the heterogeneity among ethnic groups.

Neurotransmitters
Cognitive deficit in schizophrenia has been linked to imbalances in neurotransmitters such as glutamate, GABA, dopamine, acetylcholine and histamine [45]. Glutamate and GABA play a major and well-known role in cognitive deficit in schizophrenia. An imbalance in both neurotransmitters was observed in DLPFC [46]. This area is associated with working memory-that is, the ability to manipulate information to guide behavior or thought [47]. The performance of working memory tasks implies the activation of the area showing an increase in gamma oscillatory activity (30-80 Hz) of glutamatergic pyramidal neurons, whose synchrony is regulated by GABA and glutamate [48]. In patients with schizophrenia-related cognitive deficit, there is no increase in gamma oscillatory activity in response to working memory demands [49].
It has been described that in patients suffering from schizophrenia-induced cognitive deficit there is a reduction in the glutamatergic stimulation of NMDA receptors in pyramidal neurons [50]. The activation of NMDA receptors depends upon the binding of co-agonists such as D-serine and glycine [51]. In schizophrenia-induced cognitive deficit there is a higher blockage of glycine sites in NMDA receptors due to higher levels of kynurenic acid in dorsal prefrontal cortex [52]. Inflammatory cytokines have also been reported to increase kynurenic acid levels, linking the observed increase in inflammatory cytokines with imbalances in glutamate signaling in schizophrenia subjects [53]. This is in agreement with the fact that most animal models of schizophrenia implicate the administration of drugs that block NMDA receptors to the animals, such as MK-801 (dizocilpine), ketamine or PCP (phencyclidine) [54].
However, the reduced sensitivity of NMDA receptors in pyramidal neurons in schizophrenia patients does not implicate a reduction in glutamate in DLPFC [55]. In fact, it has been described that glutaminase, the enzyme that participates in glutamate synthesis, is highly overexpressed in schizophrenia subjects [56]. Furthermore, there is an increase in the expression of metabotropic glutamate receptors in DLPFC of schizophrenia patients [57]. Those receptors exert a presynaptic regulation of glutamatergic signaling [58].
These imbalances in glutamatergic signaling are accompanied by changes in GABAergic inputs to pyramidal neurons, which are also crucial for pyramidal neurons' synchronicity [59]. It was observed that in schizophrenia subjects there is a reduction in the expression of the 67 kDa isoform of glutamate decarboxylase (GAD67) [60]. This enzyme synthetizes GABA, and its deficit suggests a reduction in GABAergic signaling in DLPFC. This is consistent with the observed reduction in parvalbumin (PV) in DLPFC [61]. PV is a marker of GABAergic interneurons on DLPFC that buffers calcium facilitating GABA release [62], which modulates the activity of pyramidal neurons [63]. The reduction in PV suggests that there is a reduction in GABA-mediated synapsis.
There is also a reduction in other subpopulations of GABAergic neurons, such as somatostatin (SST)-expressing GABAergic neurons. A global reduction in SST expression in DLPFC in schizophrenia patients has been reported [64].
Although not fully described, changes in other neurotransmitters such as acetylcholine, histamine or dopamine have been reported in schizophrenia-induced cognitive deficit patients [46]. It has been suggested that there is a relationship between a decrease in striatal dopamine and the observed disruption in cortico-cerebellar-thalamic-cortical circuits in cognitive deficits in schizophrenia [65,66].
Recently, it has been demonstrated that the acetylcholinesterase blocker galantamine improves cognitive deficit in schizophrenia, suggesting that a reduction in acetylcholine signaling may play a role in cognitive deficit [67]. It has also been observed that, in schizophrenic patients, there is lower M 1 receptor binding in the hippocampus, which results in impaired learning [68].
The role of histamine in cognitive deficit in schizophrenic patients is unclear, but Jin et al. described an increase in the expression of histamine H 3 receptors in prefrontal cortex neurons of schizophrenic patients. These data suggest that the H 3 receptor may play a role in cognitive decline in those patients [69].

Genetic Findings
Schizophrenia is a multifactorial disease whose etiology involves the interaction between environmental and genetic factors (see [70] for a review). Several genes related to cognitive deficit in schizophrenia have been identified.
Most of them encode proteins that participate in neurotransmission, such as the glutamate receptor-encoding GRIN2B and GRIN2A genes, the serotonin receptor-encoding HTR2A gene or the COMT gene [70]. COMT encodes catechol-O-methyltransferase, the enzyme that degrades catecholamines such as noradrenaline and dopamine [71]. Val158Met COMT polymorphism has been related to social cognitive deficits [72], resulting from a reduction in dopaminergic neurotransmission [73]. Nevertheless, more research is needed to identify the exact role of COMT polymorphisms in schizophrenia. Another gene involved in cognitive deficit in schizophrenia is AKT1, which encodes a serine threonine kinase activated by dopamine type 2 receptor agonists [74]. It also plays a crucial role in the neuregulin signaling pathway, and is related with social recognition [75].
Another subset of genes involved in schizophrenia are those related to brain development, for example DTNBP1, encoding dysbindin. This protein is involved in hippocampal formation [76], highlighting its role in cognitive processes critical to schizophrenia [77]. In recent years, several studies have associated the presence of single-nucleotide polymorphisms (SNPs) with cognitive deficit in schizophrenia [78,79]. Another gene related to neurodevelopment is DISC1, which participates in the acquisition of neuronal phenotypic features including axonal growth and dendritic spine formation, and in neuronal intracellular transport [80]. It was observed that some variants of the DISC1 gene are associated with alterations in the expression of genes that participate in brain development, leading to intellectual disabilities [81].

Treatment of Cognitive Deficit in Schizophrenia
The treatment of schizophrenia currently relies on antipsychotics. There are two different subgroups of antipsychotics: typical and atypical [82]. Typical antipsychotics such as haloperidol or chlorpromazine are D 2 receptor antagonists [83]. They are effective against positive symptoms, but they are less effective and even ineffective against negative symptoms and cognitive deficit [83]. Atypical antipsychotics like clozapine, olanzapine and risperidone block both D 2 and 5-HT 2A receptors, modulating both glutamatergic and dopaminergic neurotransmission in prefrontal cortex [84].
Several meta-analyses have reflected that atypical antipsychotics elicited a slight improvement in some cognitive functions, while none of them were found to have a positive profile in all cognitive functions [85,86]. In the most recent study, Baldez et al. analyzed 54 randomized double-blinded studies to create a ranking of the effect of antipsychotics on neuropsychological tests [87]. They found that olanzapine was ranked first in motor performance and visuoconstruction, amisulpride came first in attention and verbal memory, ziprasidone in working memory, sertindole in processing speed and perphenazine in executive functions, while lurasidone occupied the first position in a composite score. They also observed that clozapine and typical antipsychotics occupied the last positions [87].
De la Fuente-Revenga et al. reported that chronic treatments with atypical antipsychotics such as clozapine could induce the transcription of the HDAC2 gene, with a detrimental effect on cognitive functions in cortical neurons [88]. This effect seems to be mediated by 5-HT 2A receptors [89], suggesting that compensatory mechanisms may confer atypical antipsychotics both positive and negative effects on cognitive function, with a high variability between individuals. Therefore, because the efficacy of antipsychotics in treating cognitive deficit is very poor, several strategies have been developed to address the cognitive symptoms of schizophrenia (Table 1). Table 1. Summary of the drugs mentioned in the review assayed in clinical trials with their mechanisms of action, the quality of the evidence and the observed effect. Neuroprotective steroid Dehydroepiandrosterone Randomized double-blind trial Slight improvement in attention and visual and movement skills [120] Neuroprotective steroid Dehydroepiandrosterone Randomized double-blind trial No benefit [121] Estrogen agonist in brain Raloxifene Randomized double-blind trials Improvement in verbal memory and other cognitive domains [122,123] Raloxifene Randomized double-blind trial No benefit [124] Progesterone precursor Pregnenolone Randomized double-blind trial Improvement in memory and working attention [121] Randomized double-blind trials No benefit [125,126]

Antioxidant Compounds
A multitude of drugs with antioxidant effects have been tested. In clinical trials, they appeared to be effective in alleviating cognitive deficit in schizophrenia, although most trials included a low number of patients [127].
Some antioxidant compounds have been reported to induce the transcription of genes encoding neurogenesis-related proteins [128]. Thus, alpha-lipoic acid increases BDNF expression in animal models of schizophrenia, with an improvement in memory impairment [41]. In a small clinical trial, it was observed that dietary supplementation with n-3 polyunsaturated fatty acids (PUFAs) counteracted cortical thickness observed in schizophrenic patients [90]. Another effect of antioxidant compounds is the enhancement of the nitrergic activity in the central nervous system [41]. Nitric oxide has been demonstrated to be effective in activating NMDA receptors in animal models of schizophrenia [129], but its efficacy has not yet been confirmed in clinical trials [130].
Another antioxidant compound, N-acetylcysteine, has been described as slightly beneficial for cognitive deficit in schizophrenic patients, improving cognitive speed [91]. This may be due to its role as precursor of glutamate, enhancing glutamatergic transmission in DLPFC and complementing its antioxidant effect [131,132].
Tetracyclines are a group of antibiotic drugs with an antioxidant effect that have been suggested to be beneficial for cognitive deficit in schizophrenia [133]. In clinical trials, minocycline has demonstrated a beneficial effect in alleviating cognitive deficit in schizophrenia, improving information processing speed [92,93].
In summary, it has been observed that drugs with an antioxidant effect may help to relieve cognitive deficit in schizophrenia by a combination of neurogenic, neuroprotective and nitrergic mechanisms, but their role in the pharmacotherapy of schizophrenic patients has not yet been properly established.

Modulation of Serotonergic Neurotransmission
As described above, atypical antipsychotic drugs interact with serotonin 5-HT 2A receptors, with both positive and negative effects on cognitive function. The negative effects of chronic antipsychotic treatments on cognitive function mediated by 5-HT 2A receptors could be avoided employing HDAC2 inhibitors [134].
In addition to the role of 5-HT 2A receptors in the treatment of cognitive deficit in schizophrenic patients, the role of other serotonin receptors has also been studied. Some atypical antipsychotic drugs, such as quetiapine or aripiprazole, stimulate 5-HT 1A receptors [135,136], providing an inhibitory feedback control of serotonin release in several areas of the brain [137]. This effect is related with an improvement in some cognitive functions such as verbal memory in schizophrenic patients, as seen with the 5-HT 1A agonist tandospirone [138]. Nevertheless, activation of 5-HT 1A receptors is also associated with hallucinations and nightmares, suggesting that 5-HT 1A receptor agonists may exacerbate positive symptoms of schizophrenia [139].
Antagonists of 5-HT 3 receptors, such as the anti-emetic drug ondansetron, were shown to be mildly effective against cognitive deficits in schizophrenia in clinical trials, although these trials were conducted with very few patients [95]. Tropisetron, another 5-HT 3 receptor antagonist, improved cognition in schizophrenic patients [96]. Tropisetron is also an agonist of α 7 nicotinic acetylcholine receptors, demonstrating that the synergy between acetylcholine receptor agonism and 5-HT 3 receptor antagonism is beneficial for the treatment of cognitive deficit in schizophrenia [140].
The 5-HT 6 receptor is another serotonin receptor involved in cognitive function (see [141] for a review). It was demonstrated that AVN-492, as an antagonist of 5-HT 6 receptors, counteracted cognitive impairment in animal models of schizophrenia [142]. Nevertheless, findings in clinical trials with schizophrenic patients were contradictory. Only some of them showed a scarce beneficial effect of the 5-HT 6 antagonist AVN-211 in cognitive function [97,98]. This may be due to the fact that most of the compounds assayed as cognitive enhancers were not specific 5-HT 6 receptor antagonists, suggesting the existence of synergistic interactions with other receptors.
The 5-HT 7 receptor has also been proposed as a target for novel drugs improving cognitive enhancement (see [143] for a review). It has been suggested that the activation of the 5-HT 7 receptor reduces neuronal excitability, and it has been proposed that 5-HT 7 antagonists could exert a beneficial effect in cognition and memory [144]. In animal models of schizophrenia, it has been demonstrated that the inhibition of 5-HT 7 receptors exerts a synergistic effect with the inhibition of other receptors such as sigma receptors or 5-HT 1 receptors, both by combining them with an atypical antipsychotic or by using a multitarget strategy [145][146][147].
Although there is not much information about 5-HT 5A receptor function, it has been demonstrated that its inhibition exerts a procognitive effect in animal models of schizophrenia [148,149]. A recent work by Yamazaki et al. demonstrated that this effect is due to the activation of dopaminergic and GABAergic neurons in the prefrontal cortex as a result of the inhibition of the serotonin receptor [150].
Some antidepressants are also modulators of serotonergic neurotransmission, as serotonin reuptake inhibitors. However, recent evidence indicates that their effect may be related to an enhancement of neuroplasticity [151].

Regulation of GABAergic Neurotransmission
As described above, there is a reduction in the GABAergic transmission in DLPFC in schizophrenic patients, leading to a desynchronization of the depolarization of pyramidal neurons. Nevertheless, benzodiazepines, as non-selective agonists of GABA receptors, exert a deleterious effect in cognitive function in schizophrenic patients-specifically in attention and working memory [99]. This may be due to the fact that stimulation of GABA receptors exerts contradictory actions in neurons. The stimulation of synaptic GABA A receptors induces a phasic inhibition of the neuron, and the stimulation of extrasynaptic GABA A receptors induces a tonic inhibition of the neuron [7]. Attempts to develop selective modulators of GABA A receptors did not yield compounds with a procognitive effect [152].
On the other hand, one GABA B receptor agonist, baclofen, showed an improvement of cognitive function in an animal model of schizophrenia [153]. However, this result contradicts those published in other studies [7].
Another strategy to regulate GABAergic neurotransmission is the employment of GABA prodrugs such as BL-1020, an ester between the atypical antipsychotic perphenazine and GABA [154]. The use of this prodrug demonstrated a beneficial effect on cognitive function in animal models of schizophrenia and in phase 2 clinical trials [100,155]. Unfortunately, the results of phase 3 assays were negative [101].

Potentiation of Histaminergic Neurotransmission
An increase in the expression of histamine H 3 receptors in the prefrontal cortex has been observed in schizophrenic patients. In fact, it has been observed that H 3 receptor antagonists exerted a procognitive role in preclinical cognitive models [156]. However, in a phase 2 clinical trial, the H 3 receptor antagonist ABT-288 failed to demonstrate any benefit in comparison to placebo in schizophrenic patients [102]. Recent research has described a procognitive effect of samelisant, an inverse agonist of H 3 receptors, in animal models of schizophrenia [157]. Nevertheless, more studies are required.

Potentiation of Cholinergic Neurotransmission
Smoking has long been observed to be more common in schizophrenic patients than in the general population [158]. It has been proposed that this could be due to the fact that nicotine improves cognition in schizophrenic patients [159,160]. Indeed, it has been widely described that α 7 nicotinic acetylcholine receptor activators have a beneficial procognitive effect both in animal models of schizophrenia [161,162] and in patients, although modulation of nicotinic receptors in humans entails a high risk of adverse events [163]. The reason for the beneficial effect of α 7 nicotinic acetylcholine receptor activators on cognitive deficit is complex. On the one hand, they enhance theta activity and synaptic plasticity in hippocampal neurons, potentiating memory [164]. On the other hand, a nicotinic acetylcholine receptor agonist was shown to enhance the release of dopamine, glutamate and acetylcholine in the cerebral cortex and in the nucleus accumbens, which are reduced in schizophrenic patients with cognitive deficit [165].
One of the assayed α 7 nicotinic acetylcholine receptor activators is varenicline, which is widely used for smoking cessation [166]. Although a beneficial effect was observed in animal models [167,168], there is no evidence from clinical trials of its efficacy in alleviating cognitive deficit in schizophrenic patients [103]. Other compounds activating α 7 nicotinic acetylcholine receptors such as bradanicline, nelonicline and encenicline did not show a beneficial effect in clinical trials [104][105][106].
A positive effect, namely, the alleviation of cognitive deficit in in vivo models of schizophrenia, was also described resulting from activation of muscarinic acetylcholine receptors (specifically M 1 and M 4 receptors) present in the central nervous system [169,170]. Xanomeline, as an M 1 and M 4 receptor agonist, showed a beneficial effect in cognitive deficit in a clinical assay with 20 subjects [107]. Recently, it was shown that this drug induced beneficial effects in neuronal connectivity in animal models of schizophrenia [171].
The potentiation of cholinergic neurotransmission has not only been achieved by the direct activation of acetylcholine receptors. Another strategy to treat cognitive deficit in schizophrenia is the use of acetylcholinesterase inhibitors such as galantamine or donepezil, which prevent the degradation of acetylcholine in the synaptic cleft [172,173]. Galantamine has been assayed both in animal models of schizophrenia and in clinical trials in combination with memantine, an antagonist of NMDA glutamate receptors with procognitive effects [174]. Unfortunately, although the combination of both drugs seemed to be beneficial for cognitive improvement in animal models of schizophrenia [174], no robust benefits were observed in clinical trials with schizophrenic patients [108,175].

Potentiation of Glutamatergic Neurotransmission
As described above, schizophrenic patients with cognitive deficit show a reduction in NMDA receptor stimulation in DLPFC with imbalances in glutamate synthesis. Thus, several strategies have been proposed to tackle glutamatergic neurotransmission in order to relieve cognitive deficit in schizophrenic patients. One of them is the direct activation of NMDA receptors by employing CIQ isomers [(R) and (S)-(3-chlorophenyl) (6,7-dimethoxy-1-((4-methoxyphe-noxy)methyl)-3,4-dihydroisoquinolin-2(1H)-yl)meth-anone] [176], but because most animal models of schizophrenia employ NMDA antagonists to induce schizophrenia-like symptoms, it is not clear how translational those findings are. In animal models of schizophrenia it was observed that NMDA receptor antagonists such as meman-tine exert a beneficial effect for cognitive deficit [177], but this effect was not observed in clinical trials [109].
Another strategy to activate NMDA receptors is the enhancement of the function of co-agonists of the receptor such as glycine or serine. Inhibitors of glycine transporter-1 (GlyT1) have been widely studied as a treatment of cognitive deficit in schizophrenia because they elevate synaptic glycine levels [178]. However, studies assessing the clinical efficacy of the GlyT1 inhibitors have been contradictory: bitopertin failed to show a robust benefit in alleviating cognitive deficit [110], while BI 425809 showed a slight improvement in cognitive functions [111]. The different effect could be due to the small number of patients included in clinical trials, to the heterogeneity of clinical scales to assess cognitive deficit or to the short follow-up time in clinical trials [110,111,179]. On the other hand, cycloserine has been assayed as an activator of NMDA receptors, but the evidence of a positive effect on cognitive deficit in schizophrenia is not strong [180].
The activation of glutamate metabotropic receptors has been studied as a novel mechanism to improve cognitive deficit in schizophrenia because the activation of metabotropic receptors reduces the release of glutamate in cortical neurons, which, as exposed above, is paradoxically augmented in schizophrenia [181]. Results in mice are contradictory because a clear benefit was observed in some studies, while other studies failed to show a benefit of activating metabotropic receptors on cognitive deficit [182,183]. This lack of efficacy was also observed in clinical trials with LY2140023 [112]. Regardless, animal studies suggest that there are several strategies that appear to improve the effect of activating metabotropic receptors, such as administering them in adolescence, before the onset of schizophrenia symptoms [184,185]. Some studies have also shown a synergistic effect between the activation of glutamate metabotropic receptors and M 4 acetylcholine receptors [186]. Shen et al. described that the activation of M 4 receptors enhanced brain neuroplasticity [187], suggesting that the activation of metabotropic glutamate receptors requires an increase in neuronal plasticity to exert a beneficial effect on cognitive deficit.
The modulation of AMPA receptors has also been studied as an approach to enhance cognition in schizophrenic patients because the interplay between NMDA and AMPA receptors is critical for neuroplasticity [188]. The activation of AMPA receptors participates in cognitive processes such as learning and memory; however, AMPA agonists tend to induce AMPA receptor desensitization [189]. To avoid this effect, ampakines (allosteric potentiators of AMPA receptors) have been proposed to alleviate cognitive deficit, having been shown to improve cognitive functions in animal models of schizophrenia [190]. In clinical trials, the ampakine CX-516 has been shown to improve memory and attention in patients treated with clozapine [113], although it showed no clear beneficial effects in monotherapy [114].

Potentiation of Dopaminergic Neurotransmission
As described above, cognitive deficit in schizophrenia is related to a decrease in dopaminergic neurotransmission in DLPFC. Most dopamine receptors in DLPFC neurons are D 1 receptors, which co-localize in dendritic spines with hyperpolarization-activated cyclic nucleotide-gated potassium channels (KCNQ1), whose dysfunction seem to be related to schizophrenia [191]. Because the relationship between dopaminergic neurotransmission and cognition follows an inverted U-shaped curve, both D 1 agonists and antagonists have been assayed in animal models of schizophrenia as cognitive enhancers [192,193]. As dopaminergic neurotransmission refines the synaptic signals that reach the dendrites, the effect of dopaminergic drugs depends on the baseline levels of dopamine in the prefrontal cortex [194]. In the last few years, there has been a significant effort to develop D 1 receptor positive allosteric modulators (PAMs) for cognitive deficit in schizophrenia (for a review see [195]). In phase 1 clinical trials, D 1 receptor PAMs have been shown to be safe and tolerable [196]. However, one phase 2 clinical trial evaluating the procognitive efficacy of the D 1 receptor PAM ASP4345 was stopped because the primary endpoint of the assay was not reached [197]. D 3 receptor antagonism has been proposed as a mechanism for novel drugs against cognitive deficit in schizophrenia. The D 3 receptor plays a crucial role in the regulation of dopamine release. Blockage of the D 3 receptor enhances dopaminergic neurotransmission, with a beneficial effect on cognitive deficit [198]. On the other hand, the D 3 receptor interacts with the nicotinic α 7 receptor, boosting neuroplasticity (see [199] for a review). D 3 and D 2 receptor homology is very high, making it very difficult to develop drugs to selectively inhibit D 3 receptors [200]. However, it has been proposed that the procognitive effects of several atypical antipsychotics may be due to their effects on D 3 receptors (see [201] for a review).
Another strategy to increase the dopaminergic neurotransmission in DLPFC is the employment of the dopamine reuptake inhibitor modafinil. Modafinil has long been employed as a cognitive enhancer in both healthy individuals and in patients with neurodegenerative diseases and psychiatric pathologies [202]. Although it has been classified as a dopamine reuptake inhibitor, modafinil also modulates norepinephrine and serotonin transport, enhances glutamatergic neurotransmission and blocks GABAergic signaling [203]. It was suggested that modafinil improved cognitive functions in animal models of schizophrenia [204], but a recent systematic review by Ortiz-Orendain et al. concludes that the evidence for the efficacy of modafinil in alleviating cognitive deficit in schizophrenic patients is weak [116].

Antidepressant Drugs
Antidepressant drugs have been tested for the treatment of cognitive deficit in animal models of schizophrenia. Some of them, such as reboxetine or escitalopram, have been shown to be beneficial, especially in combination with different atypical antipsychotic drugs [205][206][207]. However, a meta-analysis highlighted that there was a high variability between assays, leading to the conclusion that there were no clinically relevant effects on cognition in schizophrenic patients [117]. These results take on new meaning in light of the recent study by Casarotto et al., in which the authors demonstrated that the mechanism of action of both typical and long-acting antidepressants is to enhance neuronal plasticity by binding to the TrkB BDNF receptor [151].

Inhibition of Phosphodiesterases
Phosphodiesterase inhibition is a promising mechanism of action for improving cognitive function in schizophrenia because cAMP and cGMP are second messengers of many receptors whose hypofunctions are involved in cognitive deficit in schizophrenia, such as dopamine or glutamate [208]. Preventing the degradation of cAMP and cGMP is expected to potentiate central neurotransmitters' activity. From the eleven members of the phosphodiesterase family, phosphodiesterases 1, 2, 4, 5, 9, 10 and 11 are widely expressed in central nervous system (see [209] and [210] for a review). We will focus on phosphodiesterases 1, 4 and 10, as they are the most studied in schizophrenia-induced cognitive deficit.
Phosphodiesterase 1 (PDE1) is an enzyme related to oxidative stress that colocalizes with dopamine receptors [211]. Its inhibition in animal models of schizophrenia appeared to be useful in alleviating cognitive deficit as an adjuvant of antipsychotic therapies with anti-inflammatory effect [212,213].
Phosphodiesterase 4 (PDE4) is one of the most widely studied phosphodiesterases in schizophrenia. It interacts with DISC1 which, as exposed above, is involved in neurogenesis and whose malfunction is related to schizophrenia [214]. Roflumilast, an inhibitor of phosphodiesterase 4, and has been tested for the alleviation of cognitive deficit in schizophrenic patients, showing little improvement in either electrophysiological abnormalities or cognitive impairments elicited by schizophrenia [118,215]. Studies with more patients are needed in order to draw meaningful conclusions about the efficacy of roflumilast in schizophrenia-induced cognitive deficit. Phosphodiesterase 10 (PDE10) has been described as a promising target for the treatment of many neurodegenerative and psychiatric diseases (see [216] for a review). Its blockage in animal models of schizophrenia using rodents and apes induces a beneficial effect in cognitive deficit [217,218]. TAK-063, a PDE10 inhibitor, has reached clinical trials, and although it was shown to be safe in phase 1 [219], it did not show a significant improvement in cognitive abilities in phase 2 studies with schizophrenic patients [119].

Steroids
Both cortical and sexual hormones are linked to schizophrenia. Chronic activation of the hypothalamic-pituitary-adrenal (HPA) axis has been suggested to play a role in the pathogenesis of schizophrenia [220]. As described above, both increased cortisol levels and blunted cortisol awakening response have been associated with a worse cognitive functioning in patients with schizophrenia. However, few assays have been performed to evaluate the efficacy of drugs counteracting the deleterious effect of cortical hormones on cognitive deficit in schizophrenic patients. The neuroprotective steroid dehydroepiandrosterone exerts a clear beneficial effect on cognitive function in animal models of schizophrenia [221], even though this effect is not always observed in clinical trials with schizophrenic patients, with only slight increases in attention and movement and visual skills [120,121]. It has been suggested that its beneficial effect may be achieved only in the early stages of the disease [222].
The role of sexual hormones, especially estrogens and progesterone, in the pathogenesis of schizophrenia is better known. Estrogens participate in neuronal development and regulation and exert a neuroprotective role [223]. It has been described that estrogen replacement therapy as an adjunctive to antipsychotics may exert a slight beneficial effect in cognitive functions in some schizophrenic patients [224]. However, estrogen replacement therapy increases the risk of uterine and breast cancer [225]. Raloxifene is a drug with estrogen-agonistic properties in the brain and estrogen-antagonistic properties in breast and uterus, reducing the risk of uterine and breast cancer. The effect of raloxifene in cognitive deficit is inconsistent, as it has been shown to have a beneficial effect in some trials [122,123], while no benefit was seen in others [124]. These differences could be explained by the heterogeneity of the patients included in each assay, because the effect of raloxifene may depend on the severity of schizophrenia and whether the patients are postor premenopausal women.
Pregnenolone is a precursor of progesterone and other steroid hormones that has also been tested as a drug for cognitive impairment in schizophrenic patients due to its beneficial effect as a neurogenesis enhancer, anti-apoptotic agent, HPA axis modulator, enhancer of myelination and regulator of GABAergic and glutamatergic neurotransmission [226]. As with raloxifene, the effects of pregnenolone are inconsistent between different trials, with a benefit seen in several trials [121], while no benefit of pregnenolone was produced in others [125,126]. The reason for these differences between trials may be due to different baseline serum pregnenolone levels. High serum pregnenolone is correlated with lower improvements in cognitive function.

Conclusions and Future Perspectives
Despite the large amount of work done investigating the etiologies and treatments of cognitive deficit in schizophrenia, little progress has been made. There are several reasons for this slow development. One of them is the low translationality of the animal models of schizophrenia. Most rely on giving mice a drug such as MK-801 or scopolamine that induces schizophrenia-like symptoms. Schizophrenia is more complex than that and, as explained in the Introduction, neuroanatomical abnormalities, biochemical imbalances and genetic alterations concur in its etiology. To overcome this drawback, novel models of schizophrenia must be developed, including new animal models such as genetically modified mice [227] and in vitro models using immortalized neuronal cells relevant to schizophrenia [228], which could be helpful for early drug discovery.
Another reason for the high attrition rate is the holistic complexity of schizophrenia. Schizophrenia is a syndrome characterized by a series of symptoms and signs, and the alterations behind them vary between patients. In addition, even the physiological status of each patient influences these symptoms and signs. Therefore, a more personalized treatment should be prescribed for schizophrenic patients. The determination of biomarkers and the advances in pharmacogenomics could make it possible to identify the most appropriate treatments for each patient [229].
Lack of adherence to the treatment in patients with schizophrenia also hampers the discovery of novel efficient drugs for cognitive deficit in schizophrenia because it introduces biases in the measurement of drug efficacy. To overcome this drawback, a more personalized therapy with motivational interviews to emphasize the benefits of the treatment and to identify the problems associated with the therapy could improve patient adherence to treatment [230].
Several successful treatments used for cognitive deficit in schizophrenia exert their action by enhancing neuroplasticity. This suggests that the combination of those pharmacological treatments with non-pharmacological therapies such as cognitive training could represent an advancement in the treatment of cognitive deficit in schizophrenic patients.