Disability-adjusted life years (DALYs) are an indication of how many years of life are lost because of early deaths or poor health. Unfortunately, psychiatric disorders are amongst the worst conditions in regard to total DALYs (Altinoz and Ince, 2017). Schizophrenia causes a total loss of 16.8 million DALYs and bipolar disorder (BD) 14.4 million. Due to the enormous burden of psychiatric disorders on human wellbeing and global economy, understanding of their novel pathophysiological aetiologies and discovery of new drugs merits a high priority of medical research (Altinoz and Ince, 2017). In this article, we propose that adding acetylsalicylic acid (ASA) and its metabolite gentisic acid (GA) to the conventional psychiatric armamentarium would be beneficial in terms of positive clinical responses. To achieve this goal, we provide – step-by-step – data in regard to prostaglandins (targets of ASA), inflammation in psychiatric disorders, and molecular features of ASA and its quinonoid redox-active metabolite GA. Figure 1 depicts ASA and its metabolic products and Fig. 2 shows neurodetrimental pathways in psychiatric disorders, which can be blocked by ASA and GA.
Prostaglandins and schizophrenia: early hypotheses
The likely associations between neuropsychiatric disorders, inflammation, and prostaglandin metabolism have sparked interest for many years. In late 1970s, Horrobin published several hypothetical articles on this issue (Horrobin, 1977, 1979, 1980). At first, he suggested that schizophrenia may be a result of a relative prostaglandin deficiency as (1) effective antipsychotics induce secretion of prolactin and prolactin stimulates synthesis of prostaglandins; (2) schizophrenics are relatively more resistant to pain and have lesser incidence of rheumatoid arthritis and prostaglandins are major players in pain and rheumatoid arthritis; and (3) high doses of agents acting as prostaglandin antagonists (quinacrine, chloroquine, and steroids) cause schizophrenia-like syndromes (Horrobin, 1977). On the contrary, he also admitted that injections of prostaglandins into animals caused catatonic states and indicated that there might be a bell-shaped dose response curve of prostaglandins in schizophrenia, and hence that prostaglandin excess may also associate with schizophrenia (Horrobin, 1977). He later described further aspects to strengthen his proposal in regard to prostaglandin deficiency in schizophrenia and suggested that (1) schizophrenia may temporarily remit after epileptiform activities or during high fever in which cerebral prostaglandin synthesis is elevated; (2) thrombocytes from schizophrenics fail to synthesise prostaglandin (Pg) E1 when stimulated by ADP and – vice versa – fail to synthesize cyclic AMP in the usual way in response to Pg E1; (3) the very effective antipsychotic clozapine behaves like a prostaglandin-analogue in in vitro studies and in animal behavioural tests (Horrobin, 1979). Nonetheless, Horrobin did not explain why the specific prostaglandin-inhibitory agents, nonsteroidal anti-inflammatory drugs (NSAIDs) including ASA did not induce schizophrenia.
Boullin and Orr (1976) analysed the actions of ASA (900 mg per os) on the responses of thrombocytes to aggregant effects of serotonin, dopamine, or N-dimethyl dopamine in schizophrenics receiving chlorpromazine treatment. Most patients showed increased aggregation responses (secondary phase) to these biological mediators before the ASA exposure. When the aggregation responses were reanalysed 23.5 h after ASA exposure, the secondary phase of the increased aggregation reactions was inhibited. Addition of 0.14–0.3 µM/l of PgE2, 15 s following aggregation-induction, partially restored this secondary phase of aggregation. The authors stated that the secondary aggregation phase may have occurred due to enhanced prostaglandin release or synthesis (Boullin and Orr, 1976). But it should also be emphasized that it may not have been the disease per se, but rather the action of chlorpromazine may have enhanced prostaglandins. Indeed, the enhanced aggregation response of platelets correlated with the efficacy of phenothiazine treatment in schizophrenics (Boullin and Orr, 1976).
In 1998, a carefully conducted study determined overactivity of cyclooxygenase (COX, also called prostaglandin synthase, the enzyme responsible of prostaglandin synthesis) in drug-naive schizophrenics, and its normalization with antipsychotics (Das and Khan, 1998). Platelets exert a chemiluminescent burst during exposure to arachidonic acid (AA), which indicates COX activation as it is inhibited by ASA. Thrombocytes from pharmacologically-naive schizophrenics had significantly enhanced metabolism of AA in comparison to healthy controls, which was normalized with antipsychotic treatment. The authors emphasized that the activity of phospholipase A2 (PLA2), which synthesizes AA from prostaglandins, was increased in the thrombocytes of schizophrenics and can be decreased by antipsychotics (Das and Khan, 1998). Prostaglandins may be involved in schizophrenia in several ways: (1) they are intermediate compounds regulating the postsynaptic signal transduction of neurons harbouring NMDA-type glutamate receptors and (2) they potentiate glutamatergic transmission by blocking glial glutamate reuptake (Laan et al., 2006). Both mechanisms can increase excitotoxic neuronal apoptosis and, vice versa, it would be logical to presume that NSAIDs would decrease this injury via suppressing prostaglandins. Indeed, studies exist which demonstrated reverse association between prolonged use of NSAIDs and the incidence of Alzheimer’s Disease; and basic studies have shown that ASA and its metabolite sodium salicylate protected neurons against the toxicity of excess glutamate in rat neurons and hippocampal slices in vitro (Laan et al., 2006).
Inflammation and psychiatric disorders
Glutamatergic excitotoxic injury and hypothalamic-pituitary-adrenal axis abnormalities intersect with inflammatory pathways in psychiatric disorders
There exist signs of remodelling of dendrites and neural cell atrophy in animal models of depression, and decreases in grey matter volume and loss of glial cells were found in postmortem brain samples from BD patients (Savitz et al., 2012). The neurotrophic effects of lithium, and longitudinal analyses monitoring brain volumetry, suggested that mood disorders may associate with neurotoxic pathways. The final cascades through which neurotoxins exert their effect may involve abnormally high glutamatergic signalling, as supraphysiological induction of glutamatergic NMDA receptors causes atrophy of neural cells and apoptosis of neurons and astrocytic cells (Savitz et al., 2012; Lee et al., 2018; Bauer and Teixeira, 2019; Woelfer et al., 2019). Multiple lines of evidence support this hypothesis. Riluzole, which blocks release of neuronal glutamate, ceftriaxone, which enhances reuptake of glutamate, and antagonists of NMDA receptors, such as ketamine, all attenuate behavioural signs of depression (Savitz et al., 2012). Moreover, genetically stress-sensitive rat strains sjpw differential NMDA receptor-expression, and behavioural signs of depression are attenuated in mice with knockout of a NMDA receptor subunit. In humans, enhanced serum glutamate levels that decrease with antidepressant administration have been demonstrated in major depressive disorder, and similar observations were made in samples of postmortem cerebrospinal fluid (CSF) (Lee et al., 2018; Bauer and Teixeira, 2019; Woelfer et al., 2019).
One potential cause of the glutamatergic dysregulation in BD may be perturbations of immunological pathways. Enhanced levels of inflammatory mediators, including PGE2, chemokine ligand 2, interleukin (IL)-6, IL-1β, interferon-α (IFNα), and tumour necrosis factor α (TNFα) are found in the blood and CSF of subjects suffering from mood disorders, both at basal conditions and following stress (Savitz et al., 2012; Lee et al., 2018; Bauer and Teixeira, 2019; Woelfer et al., 2019). Increased serum levels of positive acute-phase proteins [e.g. haptoglobin, α1-antitrypsin, ceruloplasmin, and C reactive protein (CRP)] but decreased levels of negative acute-phase proteins (e.g. albumin and retinal-binding protein) have also been demonstrated in mood disorders (Savitz et al., 2012; Lee et al., 2018; Bauer and Teixeira, 2019; Woelfer et al., 2019). IFNα treatment of Hepatitis-C triggers major depressive syndrome and/or mania in about 40% of patients, and the actions of conventional antidepressants involve decreases in inflammatory parameters (Savitz et al., 2012). In turn, antidepressants and lithium were found to decrease inflammation and reduce the levels of inflammatory cytokines (Rahola, 2012).
Inflammatory cytokines can change cerebral functions and an increased inflammatory stage exists in mood disorders. The increased activity of the hypothalamic/pituitary/adrenal axis (HPA) in mood disorders may involve inflammatory cascades, since supra-physiological secretion of corticotrophin releasing hormone induces the activation of nuclear factor κB (NF-κB), which stimulates the expression of inflammatory cytokines in immunocytes in the central and peripheral nervous systems (Kim et al., 2016). NF-κB also induces the synthesis and membrane presentation of the major histocompatibility complex (MHC) class-1 and importantly, ASA is an inhibitor of NF-κB (Altinoz et al., 2018c). In general, cortisol inhibits such inflammatory responses, but chronic exposure to stress desensitises the glucocorticoid receptor and subsequently, the anti-inflammatory actions of cortisol (Savitz et al., 2012; Kim et al., 2016; Lee et al., 2018; Bauer and Teixeira, 2019; Woelfer et al., 2019). Cytokines may desensitize the homeostatic system to cortisol; for instance, IL-1 and TNFα block dexamethasone-triggered shuttling of the glucocorticoid receptor from the cytoplasmic compartment to the nucleus. The immunological and glutamatergic excess models of BD are complementary because inflammatory pathways augment excitotoxicity (Kim et al., 2016; Lee et al., 2018; Bauer and Teixeira, 2019; Woelfer et al., 2019). Further, peripheral inflammatory signals activate cerebral microglia, triggering an inflammatory burst of cytokine mediators and free oxygen radicals (FOR) (Savitz et al., 2012; Kim et al., 2016). Cytokines, FOR, and nitrogen radicals impose direct toxicity on oligodendroglia, causing demyelination. Such a process may explain the decrease in oligodendrocytes observed postmortem in the prefrontal cortex of subjects suffering mood disorders (Savitz et al., 2012). The inflammatory micromilieu also impairs glial functioning, causing a decrease of glutamate transporters and inhibition of glial glutamate reuptake, leading to additional activation of excitotoxic and inflammatory cycles (Kim et al., 2016).
Meta-analyses on cytokines in schizophrenia have revealed that some cytokines, including IL-12, TNFα, IFNγ and soluble CD25, are elevated in and constitute trait markers of this disease, whereas IL-1β, IL-6, and TGFβ are acute schizophrenia state markers (Berk et al., 2013). Cytokines such as IL-1, IL-6, and TNFα stimulate indoleamine 2, 3-dioxygenase (IDO), which catalyses the catabolism of tryptophan, the precursor of serotonin and a modifier of T cell function, into kynurenine (Savitz et al., 2012; Lee et al., 2018; Bauer and Teixeira, 2019; Woelfer et al., 2019). Induction of the kynurenine cascade shunts tryptophan away from serotonin production, likely reducing serotonergic transmission. Kynurenine is metabolised into quinolinic acid (Quin), a potent NMDA receptor-agonist, and a mediator of lipid peroxidation, which induces neural cell injury through oxidative stress and supraphysiological activation of NMDA receptors (Savitz et al., 2012; Lee et al., 2018; Bauer and Teixeira, 2019; Woelfer et al., 2019).
Recent research has focused on the role of infective and inflammatory conditions in the etiopathogenesis of schizophrenia. For instance, prenatal bacterial or viral infections during gestation were found to associate with higher schizophrenia risk in the offspring during adulthood. In utero exposure to infection perturbs neural development, potentially enhancing vulnerability to schizophrenia. Moreover, higher maternal levels of IL-8 during human gestation are associated with a higher schizophrenia risk in the offspring (Keller et al., 2013). Also, there are meta-analyses showing perturbed cytokine profiles in people suffering from schizophrenia, compared to healthy subjects (Keller et al., 2013). A meta-analysis comprising 62 studies, including 2298 subjects with schizophrenia and 1858 healthy controls, confirmed the cytokine perturbations in schizophrenia. In vivo, IL-1RA, soluble IL-2 receptor (sIL-2R), and IL-6 are enhanced and IL-2 levels decreased. In another meta-analysis surveying 40 studies, the extent of effects, in comparison to controls, of patients at first episode, and acutely relapsing patients were similar, indicating that abnormally enhanced levels of cytokines in schizophrenia are not due to antipsychotic treatment per se (Keller et al., 2013). Transforming growth factor-β (TGFβ), IL-1β, and IL-6 were significantly enhanced in the patients with first episode or acute relapse and constituted state biomarkers. Conversely, IL-12, IFNσ, TNFα, and sIL-2R were trait markers (Keller et al., 2013).
Hypothetically, stimulated microglia in the CNS may release proinflammatory cytokines leading to neural injury that is involved in the pathogenesis of schizophrenia (Keller et al., 2013). Significant associations were also demonstrated with several markers spanning the MHC loci on chromosome 6p21.3-22.1 supporting the immune/inflammatory pathways behind mental disorders (Keller et al., 2013). Also, the inflammatory cytokines IFN, TNFα, and IL-6 may lead to mood changes by increasing activity of the HPA axis thereby enhancing systemic levels of cortisol causing hypercortisolemia (Rosenblat et al., 2016). Indeed, increased levels of steroids have been found in mania and depression. Increased concentrations of inflammatory cytokines also lower the production, transport, and sensitivity of glucocorticoid receptors in the hypothalamus and hypophysis; and subsequently impair the negative feedback control of steroid synthesis causing for chronic increase of cortisol (Rosenblat et al., 2016).
Inflammation triggers oxidative stress in psychiatric disorders and neurodegeneration
Free radicals are reactive molecular species that harbour one unpaired electron. The cerebral tissue is particularly vulnerable to their detrimental actions, and aberrantly high oxidative stress has been found to associate with the risk of neural disorders (Calderon Guzmán et al., 2007). Free radicals induce oxidative stress, causing injury to various cell components including DNA. They particularly target cell membrane lipids, and the nervous system is highly vulnerable to such oxidative injury as the majority of neural molecules are constituted from lipids (Calderon Guzmán et al., 2007). Inflammation triggers the production of oxygen radicals and increased levels of pro-oxidant injury end products have been found in depression (Rahola, 2012). Oxidative stress causes neurodegenerative diseases by inducing apoptosis, necrosis and injury to DNA damage, and also influences gene expression and triggers protein damage and proteolysis; conversely, antioxidant enzymes [glutathione peroxidase (GSH-Px), superoxide dismutase (SOD), and catalase] are therapeutic in neurodegenerative disease models (Rahola, 2012). Another factor is nitrosative stress, where aberrations in metabolism of nitric oxide (NO) cause aberrant S-nitrosylation of cysteine residues in proteins. Excess NO interacts with superoxide anions formed within the mitochondria leading to the formation of peroxynitrite, a robust pro-oxidant highly detrimental to neurons (Rahola, 2012).
Moreover, Quin, a toxic metabolite of aberrant tryptophan metabolism, also exerts oxidant actions through the production of chelates of ferrous quinolinate, causing lipid peroxidation (Rahola, 2012). There is also evidence of lowered antioxidants in depression such as Coenzyme Q, glutathione, serum vitamins C and E, zinc and albumin (Rahola, 2012; Berk et al., 2013). In comparison to healthy people, patients suffering from major depressive disorders have significantly enhanced levels of oxidative stress intermediates, including products of lipid peroxidation, the marker of oxidative DNA damage 8- hydroxy-2’-deoxyguanosine, and reduced omega-3 fatty acids (which are indicative of oxidative injury to red cell membranes). A higher oxidative stress ratio of total plasma peroxides to total plasma antioxidants is seen in drug-naive patients with major depressive disorder; positive correlations were found between depression severity scores and indices of oxidative stress; and conversely, oxidative parameters normalized with recovery from depression (Berk et al., 2013). Clinically employed antidepressants exert antioxidant properties, and antidepressant improvement of depressive pathology was associated with a prominent decrease in oxidative indices (Berk et al., 2013). Also, in schizophrenia, increased oxidant stress and decreased levels of antioxidants, such as glutathione, was reported as early as 1934, with many subsequent studies revealing reduced levels of antioxidant, including catalase, SOD, and GSH-Px, in schizophrenia (Berk et al., 2013). Lipid peroxidation is demonstrated by the enhanced levels of thiobarbituric acid reactive species (TBARS) and malondialdehyde in schizophrenic patients. In schizophrenia, DNA injury and protein carbonylation is enhanced, leading to induction of apoptosis, further damaging normal cerebral functioning (Berk et al., 2013). Genetic polymorphisms in the glutamate-cysteine ligase, which encodes a protein involving in glutathione synthesis, have been demonstrated in BD and schizophrenia (Berk et al., 2013).
Overall, it is predicted that antioxidant agents would alleviate the oxidant injury associated with the redox imbalances in psychiatric disorders, while anti-inflammatory agents would attenuate neuroinflammation by decreasing intracerebral infiltration and noxious mediator release of immune cells (i.e. hydrogen peroxide, NO, etc). On the contrary, it is conceivable that many mechanisms of inflammation and oxidant injury and conversely, actions of anti-inflammatory agents and antioxidants, overlap. Indeed, notwithstanding that the terms ‘antioxidant’ and ‘anti-inflammatory’ define two different types of biochemical activity, the majority of anti-oxidants are anti-inflammatory and the majority of anti-inflammatory agents are antioxidants (Pekoe et al., 1982; Arulselvan et al., 2016). This is due to the fact that the synthesis of the inflammatory prostaglandins is catalysed by COX (prostaglandin-H-synthases) via inherent mechanisms involving peroxide formation (Kulmacz, 1998).
Antioxidants with benefits in psychiatric disorders
A significant amount of preclinical data suggests that antioxidants may be beneficial in the treatment of psychiatric disorders; and many clinical trials are currently being conducted to evaluate their potential. However, few completed clinical studies have suggested prominent benefits of antioxidants. During our extensive review of the literature, we observed that clinical studies have demonstrated that two agents in particular, N-acetylcysteine (NAC) and ASA, provided benefits. Besides being powerful anti-inflammatory agents, salicylates and ASA could also act as antioxidants via direct effects and via stimulating heme oxygenase (Sagone and Husney, 1987; Kuhn et al., 1995; Stone et al., 2014; Schrör and Rauch, 2015). Glutathione (GSH) is a major antioxidant endogenously present in brain (Gu et al., 2015). Depletion of GSH and deficits in GSH-related enzymes are involved in the pathogenesis of autism, BD, schizophrenia, and Alzheimer’s disease (Gu et al., 2015). Animal models of various neurological disorders and/or samples from humans with neuropsychiatric disorders have revealed altered levels of GSH, oxidized glutathione (GSSG), decreased ratio of GSH/GSSG, and/or impaired expressions or activities of GSH-related enzymes, in the cerebral tissue and blood of these patients (Gu et al., 2015). GSH redox imbalance may be a primary trigger of these disorders and may also be used as a biomarker for their diagnosis. NAC enhances the cerebral concentration of GSH and constitutes a promising candidate for the treatment of neurodegenerative and neuropsychiatric disorders (Gu et al., 2015). NAC alleviates oxidative injury, neuronal apoptosis, dysfunctioning of mitochondria, neuroinflammation, and dysregulation of glutamate and dopamine in neuropsychiatric disorders and neurodegenerative conditions (Deepmala et al., 2015). There is preliminary clinical evidence suggesting that NAC may be beneficial in attention deficit hyperactivity disorder, anxiety, and mild traumatic brain injury (Deepmala et al., 2015).
In the following discussion, we summarize the results of clinical studies that have assessed the efficacy of antioxidant adjunction to the classical treatment strategies in psychiatric disorders.
Berk et al. (2008b) conducted a multicentre, randomized, double-blind, placebo-controlled study to assess the safety and efficacy of adjuvant oral NAC (1 g per os twice daily) in chronic schizophrenia. They randomized 140 chronic schizophrenic patients and 84 of these completed the scheduled 24-week treatment. Measures included the Positive and Negative Symptoms Scale (PANSS) and its subscales, Clinical Global Impression (CGI) Severity and Improvement scales, extrapyramidal rating scales and general functioning. Patients receiving NAC treatment improved more than placebo-treated patients in PANSS total, PANSS, PANSS general, CGI-Severity (CGI-S), and CGI-Improvement (CGI-I) scores. NAC therapy also significantly attenuated akathisia. Effect sizes at the end point were consistent with moderate benefits. The investigators concluded that adjunctive NAC is a safe and effective augmentation treatment for chronic schizophrenia.
Farokhnia et al. (2013) conducted a double-blind, randomized, placebo-controlled study on 42 patients with chronic schizophrenia. Patients with a score of 20 or more on the negative subscale of PANSS were included and equally randomized to be treated with NAC (up to 2 g/day) or placebo, adjunctive to risperidone (up to 6 mg/day) for 8 weeks. At the end of the study, NAC-treated patients improved significantly more in terms of the PANSS total and negative subscale scores than that in the placebo group.
Rapado-Castro et al. (2015) conducted a placebo controlled study in 121 schizophrenic patients who were randomised for 24 weeks of treatment (placebo = 62; NAC = 59). Baseline duration of the illness was classified as <10 years, 10 to <20 years, and >20 years. Mixed model repeated measures analysis was employed to evaluate the influence of the disease duration on response to NAC treatment. Adjunctive NAC exerted an advantage over placebo on positive symptom reduction and functioning in patients with chronic schizophrenia. A significant interaction between the disease duration and response to NAC-treatment was consistently revealed for functional variables and positive symptoms. This mediator effect of disease duration in response to NAC-treatment was more prominent in patients with 20 years or more of disease duration. In a further study, Rapado-Castro et al. (2017) investigated effects of NAC treatment on cognitive functioning in psychosis. The analyses were conducted in a pooled subgroup of psychotic patients who underwent neuropsychological assessment in two different trials of schizophrenia and BD. A sample of 58 participants was randomized to receive either 2 g/day of NAC or placebo for 24 weeks. Working memory, attention, and executive function domains were measured. Patients treated with NAC had significantly better working memory performance in comparison to with placebo at the end of the trial.
Sepehrmanesh et al. (2018) conducted a randomized, double-blind, placebo-controlled, 12-week trial to evaluate the effectivity of 1200 mg NAC as an adjunctive treatment in 84 patients suffering from chronic schizophrenia. In addition to the PANSS, the patients were evaluated with a standard neuropsychological screening test and the Mini-Mental State Examination (MMSE). NAC-treated patients demonstrated significant improvement in the positive and negative PANSS subscales, as well as in short-term and working memory, attention, speed of processing, and executive functioning.
Berk et al. (2008a) reported results of a multicentre, randomized, double-blind, placebo-controlled study of 75 patients with BD treated with NAC (1 g per os twice daily) adjunctive to their orthodox medication over 24 weeks. Time to a mood episode and scores on the Montgomery Asberg Depression Rating Scale (MADRS) were assessed, as well as Bipolar Depression Rating Scale and 11 other ratings of quality of life, functioning and clinical status. NAC therapy significantly improved MADRS scores and most secondary scales, which was evident by 8 weeks on the Social and Occupational Functioning Assessment Scale and the Global Assessment of Functioning Scale, and at 20 weeks on the MADRS. It was concluded that NAC is a safe and efficient adjunct treatment for the depressive component in BD.
Fernandes et al. (2016) surveyed the literature for randomized, double-blind, placebo-controlled trials using NAC for depressive symptoms regardless of the main psychiatric disorder. They performed a full review and meta-analysis on data obtained from 574 participants, of whom 291 were randomized to be treated with NAC and 283 with placebo with a follow-up that varied from 12 to 24 weeks. NAC treatment significantly improved depressive symptoms as assessed by Hamilton Depression Rating Scale and MADRS; and patients receiving NAC also had better depressive symptoms scores on the Clinical Global Impressions-Severity of Illness scale. Moreover, global functionality improved more in the NAC group than in patients receiving placebo.
Hasebe et al. (2017) explored effects of adjunctive NAC treatment (2 g/day compared with placebo) in unipolar depression in a 12-week trial. NAC therapy significantly improved depressive symptoms on the Montgomery-Asberg Depression Rating Scale (MADRS) in a manner.
Bauer et al. (2018) conducted a randomized, double blind, placebo-controlled study of NAC and aspirin as adjunctive agents for the treatment of bipolar depression. Their sample population included 24 patients with BD who had a MADRS score ≥ 20 and who were randomly assigned to be treated with either NAC (1 g), aspirin (1 g), combined aspirin and NAC (1 g each), or placebo. Participants completed questionnaires for global functioning and mood. Following 16 weeks of treatment, NAC + aspirin was associated with higher probability of treatment-response (67%) in comparison to NAC (57%) or aspirin (33%) alone.
Minarini et al. (2017) emphasized that potential benefits were found in clinical trials testing the adjunctive NAC in schizophrenia, BD, cannabis use disorder, and skin-picking (excoriation) disorder. Skvarc et al. (2017) performed a systematic literature review of the effects of NAC treatment on human cognition by analysing 12 suitable articles, including four examining Alzheimer’s disorder; three examining healthy subjects; two examining physical trauma; one examining BD; one examining schizophrenia; and one examining ketamine-induced psychosis. The data revealed significant cognitive improvements following NAC therapy.
Most recently, Çakici et al. (2019) systematically searched Embase, PubMed, the National Institutes of Health website (http://www.clinicaltrials.gov), and the Cochrane Database of Systematic Reviews for randomized clinical studies that investigated the efficacy of anti-inflammatory agents in schizophrenia. Their analysis yielded 56 studies that provided data on the efficacy of aspirin, celecoxib, statins, pioglitazone, davunetide, dextromethorphan, pregnenolone, estrogens, bexarotene, fatty acids, melatonin, varenicline, minocycline, NAC, piracetam and Withania somnifera extract. Among all these agents, the results of only estrogens, minocycline, aspirin, and NAC were significant in the meta-analysis of at least two studies.
Inhibition of COX-1 may be a more relevant strategy than inhibiting COX-2 to alleviate neuroinflammation in psychiatric disorders
In knock-out models, COX-1 and COX-2 differentially influenced leucocyte attraction during neuroinflammatory process and it was demonstrated that inhibition of COX-1 is neuroprotective, whereas a decrease in COX-2 further enhances neural injury (Choi et al., 2009, 2010; Savitz et al., 2012). COX-1 is predominantly localized in microglia, which exerts major roles in neuroinflammation, while COX-2 mainly localizes in pyramidal neurons (Choi et al., 2009, 2010; Savitz et al., 2012). The effects of COX-1 or COX-2 blockade on cerebroventricular lipopolysaccharide (LPS)-triggered neural inflammation were evaluated by comparisons of COX-1 (−/−) and COX-2 (−/−) knockout mice with wild-type (WT) (+/+) controls. After LPS exposure, accumulation of leukocytes and inflammatory processes were relieved in the COX-1 (−/−) knockout mice but enhanced in the COX-2 (−/−) knockout mice, in comparison to WT controls. Moreover, in COX-1 (−/−) knockout mice, the β-amyloid Aβ1-42-induced inflammatory pathways and accompanying neural injury were relieved in comparison to WT mice (Choi et al., 2009; Choi et al., 2010; Savitz et al., 2012). Corroborating these findings, epidemiological studies revealed that COX-1-inhibitors but not COX-2 selective NSAIDs decrease risk of AD (Choi et al., 2009; Choi et al., 2010; Savitz et al., 2012). Moreover, COX-2 exerts anti-inflammatory and protective efficacies, and may regulate the intactness of the blood-brain barrier, long-term potentiation, and synaptic transmission (Berk et al., 2013). These data indicate that the blockade of COX-1 activity may be a more relevant strategy to alleviate the neuroinflammation and neurodegeneration in psychiatric disorders (Savitz et al., 2012).
Acetylsalicylic acid is not a general nonsteroidal anti-inflammatory drugs and exerts unique pharmacomolecular actions
ASA is a NSAID and the most widely employed drug for pain relief (Calderón Guzmán et al., 2007). ASA is rapidly hydrolysed to salicylates, produced in vivo via the influence of the hydroxyl radical (OH_), and to 2,3 and 2,5-dihydroxybenzoic acid (also known as GA) (Calderón Guzmán et al., 2007; Altinoz et al., 2018a, 2018b). One of the actions of salicylates is their extinction of free radicals, providing neural protection (Calderón Guzmán et al., 2007). ASA inhibits COX-1 activity much more prominently than COX-2 and enhances lipoxygenase-derived eicosanoids, including the anti-inflammatory lipoxin A4. ASA also acetylates COX-2 protein to a remodelled enzyme that converts unesterified AA to inflammation-relieving mediators including 15-epi-lipoxin A4 (Stolk et al., 2010). The acylated enzyme can also convert docosahexaenoic acid (DHA, 22:6n-3) to 17-(R)-OH-DHA, which, similar to its metabolites tri(R)-OH-DHA [resolvin (R) D1] and di(R)-OH-DHA [neuroprotectin (R) D1], is significantly anti-inflammatory (Stolk et al., 2010).
During the initiation of inflammatory process, eicosanoids such as prostaglandins and leukotrienes function as mediators of inflammation (Berk et al., 2013). The second stage of resolution is the biosynthesis of lipid analogues that brake inflammatory cascades and induce resolution. These resolving molecules include lipoxins (LXs) and their ASA-triggered carbon-15 epimers, as well as protectins and resolvins, which originate from ω3 fatty acids (Berk et al., 2013). LXs and ASA-triggered LXs (ATLs) alleviate inflammation. Cells that synthesize COX-2 (microglia and macrophages) synthesize ATLs in the presence of ASA (Berk et al., 2013). Particularly, in a cytokine-modified micromilieu, COX-2 acetylation by ASA shifts the enzyme’s catalytic activity to an R-lipoxygenase, leading to production of 15R-hydroxyeicosatetraenoic acid, which is rapidly converted to 15-epimeric-LXA4 or 15-epimeric-LXB4 by 5-lipoxygenase (Berk et al., 2013).
Treatment of healthy subjects with low doses of ASA markedly enhances plasma ATL levels, with simultaneous blockage of the biosynthesis of thromboxanes, indicating that ATLs may contribute to the beneficial biological effects of ASA (Berk et al., 2013). The actions of ATLs as anti-inflammatory mediators are established, with their efficacies including the blockage of attraction and activation of neutrophils and eosinophils. LXs and ATLs also induce genes (including NAB1) mediating resolution of inflammation, and modify NF-κB activity, as well as inducing the engulfment of apoptotic cell debris by human macrophages in a noninflammatory manner and converting cytokine-induced macrophages from an inflammatory state to an anti-inflammatory phenotype (Berk et al., 2013). Varying ASA doses (100–300 mg/day) reduce concentrations of proinflammatory mediators including CRP, IL-6, and TNFα in plasma of patients with cardiovascular disease. ASA reduces inflammatory cytokines, such as IL-8 and TNFα, but not anti-inflammatory cytokines, such as IL-4 and IL-10 (Berk et al., 2013). ASA treatment in various cells, including HeLa cells and fibroblasts, significantly suppressed IL-1 and TNFα-induced induction of NF-κB. ASA also reduces Th17 responses in mouse models of LPS-triggered pulmonary inflammation and inhibits the IL-1β-stimulated formation of NO and the inducible NO synthase (Berk et al., 2013).
Use of acetylsalicylic acid in psychiatric disorders
Acetylsalicylic acid and schizophrenia
In animal models, it was found that chronic exposure to inflammatory cytokines leads to behavioural changes that mimic schizophrenia (Nawa and Yamada, 2012). Laan et al. (2010) conducted a placebo-controlled, double-blind randomized study, in which 70 antipsychotic-treated schizophrenia patients were recruited. They randomized the patients to placebo or adjuvant ASA treatment (1000 mg/day) and over 3 months monitored psychiatric symptoms with the Positive and Negative Syndrome Scale (PANSS). The primary outcome was the alteration in the total PANSS score. Secondary outcomes were alterations in the PANSS subscales and cognitive tests. The PANSS is widely employed to study the effectiveness of antipsychotics and is accepted as a ‘gold standard’ that all evaluations of psychotic disorders should use (Opler et al., 2017). It measures two different types of symptoms in psychotic disorders: positive symptoms, which are distortions or excess of normal functioning (such as hallucinations and delusions); and negative symptoms, which represent a decrease or loss of normal functioning, including normal thinking, actions, and the ability to express emotions properly. A larger reduction of PANSS score was witnessed in the ASA group in comparison to the placebo group, with similar yet insignificant results for scores on the PANSS subscales. The effect on total PANSS score was higher in patients with more perturbed immune functioning. ASA did not alter cognitive functioning and significant untoward effects were not observed (Laan et al., 2010).
Webb et al. (2013) published an interesting case of childhood schizophrenia, which dramatically responded to ASA as an adjunct to aripiprazole. A 13-year-old male presented to a psychiatric unit after increasingly bizarre behaviour (Webb et al., 2013). For several weeks before admission, his symptoms markedly worsened. His PANSS score on admission was 79. For treatment, he was started on aripiprazole and showed partial improvement, appearing less frightened of his hallucinations, but responding to internal stimuli with prominent negative symptoms. After 3 weeks of treatment on aripiprazole as monotherapy, his PANSS score was 59. Given the lack of further clinical progress, he was started on ASA 325 mg twice daily as an adjunctive treatment. Within 2 days, he had a notable reduction in hallucinatory behaviour, appearing much less distracted by internal stimuli during the interview. He also was more fluent in his speech, showed more signs of facial expression, and began to spontaneously interact with other patients on the unit and his PANSS score decreased to 45, with improvements in both the negative and positive subscales.
Sommer et al. (2014) performed a meta-analysis that included only double-blind, placebo-controlled randomized studies. ASA, estrogens, and N-acetyl-cysteine (NAC) exerted significant positive effects. Addition of ASA to antipsychotic treatment was encouraging, as well as the adjuvant application of NAC and estrogens. They also reported that ASA and celecoxib differed in their actions, with ASA providing higher benefits. Köhler et al. (2016) investigated whether adjunctive use of acetaminophen or NSAIDs influenced 2-year relapse of schizophrenia in a population-based setting. In 16 235 patients with schizophrenia treated with antipsychotics, 1480 (9.1%) also consumed NSAIDs and 767 (4.7%) acetaminophen. Unexpectedly, usage of NSAIDs was found to associate with a higher relapse risk, particularly in users of diclofenac and ASA. They attributed these findings to a potential influence of underlying somatic comorbidity.
Acetylsalicylic acid and bipolar disorder
Horrobin and Lieb (1981) supposed that the recurring mood episodes in BD may be caused by fluctuating immuno-inflammatory processes and hypothesized that lithium may be providing its stabilizing actions on mood via modifying the immune system. Since then, several groups have shown mutual interactions between BD and inflammatory conditions. Peripheral cytokines may pass through the blood-brain barrier via the choroid plexus (Weller et al., 1996). Moreover, a relatively recent investigation found functional lymphatic vessels coating dural sinuses, indicating another mechanism whereby peripheral cytokines influence the central nervous system (CNS) (Louveau et al., 2015). Several analyses demonstrated associations between increased peripheral inflammatory cytokines during depressive and manic episodes and euthymia, indicating a mild yet sustained inflammatory state (Rosenblat et al., 2016). Serum levels of inflammatory mediators IL-4, TNFα, sIL-2R, IL-1β, IL-6, soluble receptor of TNFα type 1 (STNFR1), and CRP are increased in patients with BD in comparison to healthy subjects (Rosenblat et al., 2016). During euthymia, sTNFR1 is a constantly enhanced inflammatory mediator. During mania, serum concentrations of IL-6, TNFα, sTNFR1, IL-RA, CXCL10, CXCL11, and IL-4 are increased; while during depression, serum concentrations of sTNFR1 and CXCL10 are increased (Rosenblat et al., 2016).
Monoamine levels may be changed by proinflammatory cytokines IL-2, IL-6, and TNFα (Capuron et al., 2003). IFN and IL-2 also directly enhance IDO enzyme activity, thus enhancing the catabolism of tryptophan to depression-inducing mediators (TRYCATs) (Dunn et al., 2005). Reduction of tryptophan leads to decrease in production and release of serotonin (5-HT), which associates with affective and cognitive dysfunctioning (Rosenblat et al., 2016). Serotonin levels may also be influenced via the IL-6 and TNFα stimulated catabolism of serotonin to 5-hydroxyindoleacetic acid (5-HIAA) (Rosenblat et al., 2016). Thereafter, microglia may be supraphysiologically activated, acting detrimentally on neural paths in cerebral regions with key role controlling cognition and mood including the anterior cingulate cortex, prefrontal cortex, hippocampus, amygdala and insula in patients with BD (Rosenblat et al., 2016). Action pathways of classical and novel antidepressants are partly attributed to their influences on several immune pathways, including (1) alleviating inflammation accompanied by enhanced levels of TNFα, IL-1, and IL-6; (2) decreasing cellular immunostimulation and Th1 and Th17 responses while enhancing regulatory T lymphocytes; (3) reducing oxidative and nitrosative pathways and enhancing the total antioxidative status; and (4) protecting mitochondria and mitochondrial DNA against noxious events (Berk et al., 2013).
Since rats treated with mood stabilizers had lower cerebral AA (20:4n-6) metabolism, including expression of COX and PLA2, Stolk et al. (2010) presumed that NSAIDs and glucocorticoids, which also block the brain AA pathways, may also reduce symptoms of BD. They obtained medication histories on patients treated with lithium from the Netherlands PHARMO Record Linkage System (which connects records of more than 2 million individuals) and stratified data in regard to drug types that block PLA2 and/or COX enzymes, and duration of usage. They employed incidence medication event density (increase of dosage or medication change) as an indicator for worsening of clinical symptoms. They found that low-dose (≤80 mg/day) ASA significantly decreased the relative risk of worsening in patients under lithium treatment, while glucocorticoids or other NSAIDs did not (Stolk et al., 2010).
Savitz et al. (2018) enrolled 99 depressed outpatients with BD in a 6-week, double-blind, placebo-controlled trial and randomized the patients to one of the four groups: minocycline (100 mg b.i.d.) + ASA (81 mg b.i.d.) (M + A); minocycline + placebo ASA (M + P); placebo-minocycline + ASA (A + P); and placebo-minocycline + placebo ASA (P + P). They found a significantly higher permanent clinical response in the M + A group (44%) vs. the P + P group (21%).
Kessing et al. (2019) investigated Danish nation-wide population-based registers to assess whether continued use of low dose aspirin, high dose aspirin, other NSAIDs, statins, allopurinol, and angiotensin agents lower the rate of incident mania/BD (Kessing et al., 2019). Their analysis included a nation-wide longitudinal study using Poisson regression analyses including all persons in Denmark who were treated with the medication of interest, and a random sample of 30% of the Danish population. A total of 1 605 365 participants, with a median age 57 years, were treated with the one of the six drugs of interest during the period from 2005 to 2015. Continued use of low dose aspirin, angiotensin agents, and statins were associated with lowered rates of mania/BD, while continued uses of high dose of aspirin and other NSAIDs were associated with an increased rate of BD. The higher rates of BD in association with higher dose aspirin usage were attributed to the coincident chronic pain and cardiovascular disorders frequent in these patients.
Haarman et al. (2019) found that the benefit of aspirin in BD may also associate with stimulation of remyelination, as myelin damage is demonstrated in this disorder. They also pointed out that both aspirin and lithium inhibit glycogen-synthase kinase-3β (GSK-3β) and synergize to reduce cerebral prostaglandin levels. Very recently, Zhang et al. (2019) recapitulated manic-like symptoms of BD with intracerebroventricular administration of ouabain, which induced many mania-like features including increased stereotypic counts, traveling distance in open-field test, and lowered expression of brain-derived neurotrophic factor (BDNF), IFN-γ, and toll-like receptor 3, which are frequently encountered in patients with BD. ASA partially reversed some of these features including the increased stereotypic counts, and also increased BDNF levels.
Acetylsalicylic acid treatment may reduce cardiac disease and malignancy-associated mortality in patients with psychiatric disorders
Patients with schizophrenia have a 79% increased risk to die of cardiovascular disorders. By inhibiting platelet aggregation, COX-1 inhibition decreases the risk of cardiovascular events. Cancer is another source of increased mortality in schizophrenic patients as these patients generally smoke heavily. Long-term treatment with ASA reduces cancer mortality, even at relative low dose. Therefore, it is hypothesized that ASA adjunction might reduce morbidity and mortality in schizophrenia patients (Sommer et al., 2012). BD also associates with nearly a two-fold cardiovascular mortality risk when compared with general population estimates. Putative factors underlining the link between BD and metabolic syndrome include behavioural features, shared metabolic derangements, and adverse effects of antipsychotics. Valproate and olanzapine are widely used as mood stabilizers in BDs, which could trigger weight gain and diabetes, conditions that enhance risk of mortal cardiovascular events (Fond et al., 2014).
Hypothesis to combine acetylsalicylic acid with its metabolite gentisic acid in treatment for psychiatric diseases to increase antioxidant and anti-inflammatory effects
The formation of gentisic acid
ASA causes the production of considerable amounts of GA in human. Hepatic ASA metabolism via cytochromes yields the production of GA as a secondary quinonoid. The term ‘quinonoid’ determines the similarity to the quinone by possession of a molecular structure defined by a benzene nucleus harbouring two double bonds inside the nucleus and two external double bonds attaching the nucleus at para or ortho positions. ASA is first metabolized to salicylate, which has a markedly prolonged half-life and concentrates in the body; salicylic acid is then catabolized by pathways also involving the production of GA; thus, production of GA after salicylate treatment is used as a means to define the in vivo synthesis of OH radicals (Grootveld and Halliwell, 1986). In patients treated with 6–8 ASA tablets (650 mg daily corresponding to 3.9–5.2 g), plasma levels of GA are between 5 and 25 μM (Cleland et al., 1985). In human plasma, salicylate levels of 2 mM can be achieved with ASA treatment, which leads to GA concentrations reaching 20 μM (Exner et al., 2000). We have recently published an article that discusses molecular actions of GA (Altinoz et al., 2018b) and below, we summarize some of these actions pertinent to neuropsychiatric disorders.
Blood levels of gentisic acid during its direct use as a medicine in rheumatic diseases and its high biosafety
The minimum lethal dose of GA in rats is 8.5–9 g/kg, which correspond to about 600 g for a 70 kg normal person, which are very large dosages suggesting its high biosafety. No detrimental actions on growing rats occurred from feeding gentisate in amounts up to 20 g daily and the prothrombin time was also not effected by feeding gentisate at the 30 and 40 g levels for 18 weeks (Gorsuch, 1950). Investigations with 14C GA injection in rats revealed a high organismal dissemination, but prolonged feeding with 2% GA in rats led to negligible storage in brain, hepatic tissue, and omental fat, indicating that sustained exposure to GA would be needed to assess its therapeutic actions (Astill et al., 1964). Mongrel dogs fed with 50 mg/kg gentisate three times a day for the first 5 days, and twice on the 6th day of each week for 7 weeks showed no pathological changes in aorta, cardiac, pulmonary, hepatic, renal, splenic, gastric and intestinal tissues or in pancreas, thyroid, adrenal, and bladder tissues (Rosenberg et al., 1952). GA exerts 8-to-10 times lesser toxicity than salicylates in animals and also saves the plasma alkali reserves (Kleinsorge and Pohl, 1953). Salicylate inhibits the in vitro activity of hyaluronidase at very high concentrations, whereas GA demonstrates a similar action at a few μg/ml (Meyer and Ragan, 1948; Astill et al., 1964). Since hyaluronidase contributes to cartilage injury, GA has been employed as an antirheumatic medicament since 1948 (Meyer and Ragan, 1948). Unexpectedly, GA, at comparable doses to salicylate, was analgesic and antiedematous, and relieved joint heat with accompanying lowering of the sedimentation and body temperature. Moreover, patients who are intolerant to ASA because of duodenal ulcers did not show stomach irritation while consuming GA (Meyer and Ragan, 1948; Astill et al., 1964).
Another feature of GA is that it does not change thrombocyte activity even at 500 μM, suggesting lack of bleeding risk with GA treatment (O’Brien, 1968). GA use at very high dosages does not cause confusion or delirium, which may be seen with high doses of salicylates (Kuhn and Bergmann, 1954). In regard to embryonic toxicity, while GA doses up to 1.9 mM was not toxic to rat embryos, salicylate caused marked lethality in embryos indicating again higher biosafety of GA (Greenaway et al., 1984). While salicylate triggers dose-dependent swelling and uncoupling of mitochondria, GA does not cause mitochondrial swelling or uncoupling (Gutknecht, 1992). Much more noteworthy (and important for our hypothesis), when GA is co-applied with salicylate, it does not attenuate its clinical effects or increase its toxicity (Boyd et al., 1950). ASA intoxication happens when the total salicylate concentrations exceed 3 mM with effects including increased consumption of oxygen and body hyperthermia (symptoms similar to Reye Syndrome), indicating that oxidative phosphorylation is the main factor in salicylate poisoning (Gutknecht, 1992). Unlike salicylate and ASA, GA levels up to 20 mM do not block transfer of protons through phospholipid bilayers; thus, it is also safer in regard to general metabolic effects (Gutknecht, 1992).
Gentisic acid as a strong antioxidant
GA has a more profound antioxidant efficacy than trolox (a water-soluble vitamin E analogue) (Yeh et al., 2004) and forms stable chelates with important cations in biological systems including Fe3+, Cu2+, and Al3+ (Pecci and Foye, 1960). Myeloperoxidase (MPO)-driven tyrosine radicals can induce low density lipoprotein (LDL)-oxidation, which is involved in the initiation of atherosclerosis (Hermann et al., 1999). At an easily achievable level (3 μM), GA blocked tyrosyl radical-induced LDL oxidation, under biochemical conditions where even salicylate acts as a pro-oxidant. Furthermore, even when oxidization of LDL was maximally stimulated by salicylate, this oxidation was blocked in salicylate/GA ratios that could be reached in the blood of patients under ASA treatment (Hermann et al. 1999). When actions of GA, ASA, salicylate, and o-anisic acid were assessed in the DPPH (diphenylpicryl hydrazyl) free radical (100 μM)-scavenging test, only GA (50 μM) demonstrated scavenging efficacy, while all other salicylic acids (50 μM) were inefficient (Hermann et al., 1999). GA produces a more stable phenoxyl radical than salicylate and hence, it may not be involved in a chain reaction of free radicals with polyunsaturated fatty acids. Moreover, phorbol-myristate-acetate induction of neutrophils led a roughly 30-fold enhancement in the production of dienes, which was further doubled when neutrophils were treated with salicylate; while admixture of GA (200 μM, 30.8 μg/ml) with this salicylate solution hindered this oxidation reaction and reduced diene formation by about 50% (Hermann et al., 1999).
Glucose-driven radicals may oxidize LDL, which may be involved in the initiation of diabetes-provoked atherosclerosis (Exner et al., 2000). Glucose auto-oxidation leads to production of versatile reactive molecule species including superoxide, hydrogen peroxide, hydroxyl, hydroxyalkyl, and peroxylradicals. ASA, salicylate, and their metabolites 2,3-dihydroxybenzoic acid (2,3-DHBA) and GA were studied to determine whether they could inhibit glucose-induced LDL oxidation (Exner et al., 2000). Only 2,3-DHBA and GA inhibited LDL oxidation and the increase of endothelial tissue factor synthesis stimulated by glucose-oxidised LDL. LDL oxidation by glucose was hindered by achievable levels of GA (20–50 μM) where ASA failed to inhibit this reaction (Exner et al., 2000). Cholesterol ester hydroperoxides were formed by treating human plasma with Cu2+ to assess the action of GA on free radical-induced plasma lipid damage. GA reduced LDL oxidation in achievable concentrations and in a dose-dependent manner (0.1–1 μM). GA also inhibited the production of plasma cholesterol ester hydroperoxides (20 μM), demonstrating a free radical quenching potential in the DPPH assay in a dose-dependent manner (20–50 μM). Cholesterol ester hydroperoxides arose following absolute consumption of ascorbate (Vitamin C) suggesting that ascorbate is the first defense mediator against plasma lipid peroxidation (Ashidate et al., 2005). The consumption of GA begins after total consumption of both ascorbate and CoQ10; while the concentrations of other plasma antioxidant molecules (α-tocopherol, uric acid, and bilirubin) remained unchanged, indicating that GA is more efficient than these endogenous antioxidants in blocking lipid peroxidation in plasma (Ashidate et al., 2005). GA effects on rat liver microsomal lipid peroxidation and hydroxyl radical (OH) formation were assessed in NADPH-dependent, 50 μM Fe+2–500 mM ascorbate (Fe+2–AA) or 50 μM Fe+2 systems, respectively (Ozgová et al., 2003). GA blocked lipid peroxidation in NADPH-dependent and Fe+2 –AA systems with IC50 values above 30 μM and 80 μM, respectively. GA also scavenged.OH radicals in NADPH-dependent and Fe+2 –AA systems with IC50 values above 40 μM; while it was inefficient to chelate Fe+2 and underwent oxidation with Fe+3 (Ozgová et al., 2003).
Joshi et al. (2012) analysed actions of GA on several in vitro systems, including the rat hepatic mitochondria (RHM) and human erythrocytes (Joshi et al., 2012). GA quenched hydroxyl radicals and subsequently inhibited the production of the oxidizing phenoxyl radical (~24%) and reduced adduct radical (~76%). GA also quenched organohaloperoxyl radical, and demonstrated almost absolute protection against γ- radiation in RHM as assessed with production of TBARS and peroxide at a concentration of 10 μM, without leading to pro-oxidant actions up to 200 μM. GA also inhibited the RHM-protein damage caused by γ-radiation in a dose-dependent manner up to 100 μM, which was determined by the levels of protein carbonyls and decrease of protein thiols (Joshi et al., 2012). GA efficiently protected the antioxidant SOD enzyme against γ-radiation and also inhibited the γ-radiation-induced red blood cell lysis in a dose-dependent manner up to100 μM (~75% inhibition of γ-radiation-induced lysis of human red blood cells at 100 μM). The authors also indicated that (1) the free-radical scavenging and radiation-protective features of GA happened due to its phenoxyl group; (2) both phenolic groups of GA were involved in its antioxidant efficacy; (3) GA exerted a reducing efficacy on Fe+3 (Joshi et al., 2012). GA protected against lipid injury via quenching the lipid radicals (L•/LOO•) and subsequently inhibited the oxidizing chain reactions. Its radiation-protective action for lipids was higher than for proteins in aqueous phase and was saturated at about 10 μM (Joshi et al., 2012). The reactions of •OH radical with proteins causes the production of protein carbonyls, oxidation-induced amino acid loss (tryptophan, tyrosine, and cysteine), and structural damage to enzymes. GA could quench protein radicals and •OH to protect proteins and the 3D-conformal structure of antioxidant enzymes including SOD (Joshi et al., 2012). GA and its isomers 2,3-DHBA and 3,4-DHBA demonstrated increased peroxyl radical scavenging efficacy in comparison to other free radical-scavengers, including thioacrolein, allicin, melatonin, dopamine, and caffeine. Thus, at physiological pH in aqueous solution, these three DHBA analogues, including GA, constitute the most potent peroxyl radical scavengers known (Perez-Gonzalez et al., 2014). GA inhibition of COX (also known as prostaglandin endoperoxide synthase) may also associate with its free radical-scavenger activity (Borges and Castle, 2015). The ionization potential of GA is lower than that of ASA and salicylic acid, making it a more potent donor of single electrons than these two molecules. Thus, it was proposed that GA could block COX via scavenging the tyrosyl radical at the COX active site, which catalyses the cyclization and oxygenation of AA to yield PGG2 (Borges and Castle, 2015).
Yeh and Yen (2006) treated rats orally with GA (100 mg/kg) for 2 weeks, which enhanced both the mRNA levels and activities of SOD, GPx, and catalase in hepatic and intestinal tissues. GA also decreased oxidized glutathione and enhanced reduced glutathione and the antioxidant capacity within the hepatic parenchyma. GA also increased the absolute level of Nrf2 [Nuclear factor (erythroid-derived 2)-like 2, NFE2L2], a transcription factor regulating the antioxidant defenses. Nrf2 is a crucial protein, which regulates the transcription of free radical-scavenging enzymes protecting against oxidative injury and inflammation. Thus, various compounds that stimulate Nrf2 cascades are highly being investigated for diseases that associate with oxidant injury. GA is a plausible candidate to induce Nrf2 besides its various benefits in regard to scavenging free radicals and alleviating inflammation (Yeh and Yen, 2006).
Gentisic acid inhibits cyclooxygenase and 12-lipoxygenase enzymes
The action of the NSAIDs is attributed mainly to the inhibition of the COX, which catalyses the first step in the conversion of AA to thromboxanes and prostaglandins (Hinz et al., 2000). ASA acts via specific acetylation of a serine hydroxyl in the COX enzyme, which juxtaposes the active COX site; yet this model cannot explain the blockage of COX that is seen with nonacylating NSAIDs. The synthesis of GA from salicylate might occur in an inflammatory milieu as a consequence of ‘co-oxidation’ associated with prostaglandin-H-Synthase (PGH-synthase)-peroxidase activity, which is seen with various phenolic molecules (Holmes et al., 1985). GA in this oxidative milieu might undergo further oxidation to reactive quinonoid molecules, which could inhibit COX activity irreversibly via covalent bonds. The oxidized derivative of GA, which most likely leads this electrochemical inactivation, is carboxybenzoquinone (Holmes et al., 1985).
The proposal that blockade of prostaglandin biosynthesis mediates the pharmacological actions of NSAIDs is challenged when the actions of salicylate and ASA are compared (Hinz et al., 2000). Unlike its acetylated derivative acetylsalicylate, salicylate does not inhibit COX-1 and COX-2 in vitro but demonstrates comparable pain-relief and anti-inflammatory actions as acetylsalicylate (Hinz et al., 2000). Acetylsalicylate (IC50 of 5.35 mM) inhibits LPS-stimulated and COX-2-mediated production of PGE2 in macrophages, whereas no significant blockage happened after exposure to sodium salicylate and salicyluric acid at levels up to 100 μM (Hinz et al., 2000). Nonetheless, GA (10–100 μM) and salicyl-CoA (100 μM), the intermediate molecule in the production of salicyluric acid from salicylic acid, markedly blocked LPS-induced PGE2 production (Hinz et al., 2000). GA significantly inhibited (by around 35%) PGE2 production at a level of 10 μM, reaching 62% inhibition at 100 μM (Hinz et al., 2000). In contrast, γ-resorcylic acid (2,6-dihydroxybenzoic acid) did not change prostaglandin synthesis, indicating that the para-substitution of hydroxy groups is crucial for COX-2 inhibition by GA (Hinz et al., 2000). None of the salicylate analogues changed the expression of COX-2 as determined by real-time RT-PCR (Hinz et al., 2000).
Gentisic acid inhibits neutrophil aggregation and superoxide anion release
GA inhibits aggregation of neutrophils and release of superoxide anion following challenge with stimuli including calcium ionophore or AA, and this blockage happens much markedly than provided by ASA or salicylate (Lorico et al., 1986). Release of superoxide anion is blocked by 97% at 1 mM of GA, while equimolar levels of ASA or salicylate were efficient at around 40% (Lorico et al., 1986). Taken together, strong antioxidant and anti-inflammatory features of GA may potentiate the neuroprotective effects of its parental compound, ASA. Furthermore, it also has the potential to alleviate – rather than to increase – side effects of ASA. We strongly believe that GA highly merits to be studied as an ASA-adjunct to combine with antipsychotic drugs in treatment of psychiatric disorders. Lastly, we provide evidence that systemically administered GA may cross the blood-brain barrier and exert neuroprotective actions in vivo.
Evidence indicating that gentisic acid crosses the blood-brain barrier and demonstrates neuroprotective efficacies
It was demonstrated that GA is endogenously formed in rats exposed to hyperoxia and exogenous ASA and Vitamin A combination synergistically enhanced GA levels, indicating a potential role of endogenous GA as a defense molecule against oxidative injury (Calderon Guzmán et al., 2007). Indeed, GA is produced as an endogenous siderophore in humans (Altinoz et al., 2018a, 2018b). Kabra et al. (2014) analysed the neuroprotective activity of GA in mouse and rat models of Parkinson’s disease. Three behavioural models were employed, including haloperidol-induced catalepsy and reserpine antagonism in Swiss Albino mice and orofacial dyskinesia in Wistar rats. Various parameters (cataleptic behaviour, rearing and grooming frequencies, horizontal movements) and dyskinetic behaviour such as vacuous chewing and tongue protrusion were analysed to assess the anti-Parkinson’s activity of GA. GA (80 mg/kg) lowered the frequency of vacuous chewing and tongue protrusion and exerted a significant (P < 0.01) decrease in the duration of cataleptic behaviour in a dose-dependent manner in comparison to haloperidol control.
Haloperidol neurotoxicity may occur due to free radicals and lipoperoxidation products increasing the turnover of dopamine, which remains available for catabolism either by auto-oxidation or by oxidative deamination catalysed by monoamine oxidase (MAO). Subsequently, it may lead to an increase in dopamine metabolites along with a decrease in dopamine receptor activation. During this process, H2O2 is formed and becomes an important source of oxidative injury in catecholaminergic neurons (Kabra et al., 2014). Indeed, certain studies revealed that haloperidol neurotoxicity occurs due to the inhibition of mitochondrial electron transfer with an increase of O2 and H2O2 production. In particular, mitochondrial complex-1 perturbations play important roles in the Parkinson’s pathogenesis. Oxidative injury and subsequent cell death happen in the substantia nigra pars compacta when there is (a) an enhanced dopamine turnover, causing abnormally high peroxide formation; (b) a decline in glutathione (GSH) content, attenuating the brains capacity to scavenge H2O2; or (c) an enhancement in reactive iron species, which lead to the formation of hydroxyl radicals (Kabra et al., 2014). After GA treatment, significant alterations occurred in Parkinson’s affected rodents in regard to lipid peroxidation and antioxidant concentrations in concert with its strong antioxidant potentials. Reduction of oxidative injury may be one of the likely mechanisms for the anti-Parkinson effects of GA (Kabra et al., 2014). As yet, no further studies exist in regard to other behavioural effects of GA.
As suggested above, psychiatric disorders exert a huge burden on the global human health and economy. Moreover, the efficacies of the currently available drugs in the psychiatric armamentarium are suboptimal, especially when schizophrenia is considered. Research focusing on the discovery of new molecules targeting neurotransmitter systems is undoubtedly important and essential. But repurposing of old and cheap drugs with a relatively high clinical safety may also provide clinical benefits in management of psychiatric disorders. Such a strategy makes particular sense when considering that the neurotransmitter hypotheses do not completely explain the pathogenesis of psychiatric disorders. It is increasingly recognized that inflammatory pathways and free radical stress/oxidant injury play crucial roles in the etiopathogenesis of psychiatric conditions. Hence, a simple and universally available drug, ASA, with its anti-inflammatory and anti-oxidant features may augment treatment efficacy in psychiatric disorders; and indeed, there exist some clinical data, which support this hypothesis. In this article, we also hypothesized that GA, a redox-active ASA metabolite, may further potentiate neuroprotective and antipsychotic-augmenting effects of ASA, as it is a much more potent antioxidant than its parental compound, and has a much lower toxicity profile than ASA and salicylates. Currently, there is no FDA approval for the usage of aspirin and GA in the treatment of psychiatric disorders. Nonetheless, we strongly believe that ASA and its combination with GA merit to be studied in animal models of psychiatric disorders to test whether they could alleviate neuroinflammation and augment the efficacy of antipsychotics. If positive results were obtained, clinical trials could be conducted more easily than for trials investigating novel molecules, as both ASA and GA have a long history of clinical application and their safety limits are well established.
M.A.A. wrote the article. A.O. reviewed the article and made relevant criticisms. All authors read and approved the article.
Conflicts of interest
There are no conflicts of interest.
Altinoz MA, Ince B. Hemoglobins emerging roles in mental disorders. Metabolical, genetical and immunological aspects. Int J Dev Neurosci. 2017; 61:73–85
Altinoz MA, Elmaci I, Cengiz S, Emekli-Alturfan E, Ozpinar A. From epidemiology to treatment: aspirin’s prevention of brain and breast-cancer and cardioprotection may associate with its metabolite gentisic acid
. Chem Biol Interact. 2018; 291:29–39
Altinoz MA, Elmaci I, Ozpinar A. Gentisic acid
, a quinonoid aspirin metabolite in cancer prevention and treatment. New horizons in management of brain tumors and systemic cancers. J Cancer Res Oncobiol. 2018b; 1:1–19
Altinoz MA, Ince B, Tek C, Srihari VH, Guloksuz S. The NF-κb signaling pathway: an important therapeutic target in psychiatric disorders. Mol Psychiatry. 2018; 23:490–491
Arulselvan P, Fard MT, Tan WS, Gothai S, Fakurazi S, Norhaizan ME, Kumar SS. Role of antioxidants and natural products in inflammation. Oxid Med Cell Longev. 2016; 2016:5276130
Ashidate K, Kawamura M, Mimura D, Tohda H, Miyazaki S, Teramoto T, et al. Gentisic acid
, an aspirin metabolite, inhibits oxidation of low-density lipoprotein and the formation of cholesterol ester hydroperoxides in human plasma. Eur J Pharmacol. 2005; 513:173–179
Astill BD, Fassett DW, Roudabush RL. The metabolism of phenolic antioxidants. 4. The metabolites of gentisic acid
in the dog. Biochem J. 1964; 90:194–201
Bauer ME, Teixeira AL. Inflammation in psychiatric disorders: what comes first? Ann N Y Acad Sci. 2019; 1437:57–67
Bauer IE, Green C, Colpo GD, Teixeira AL, Selvaraj S, Durkin K, et al. A double-blind, randomized, placebo-controlled study of aspirin and N-acetylcysteine as adjunctive treatments for bipolar depression. J. 2018; 80:18m12200
Berk M, Copolov DL, Dean O, Lu K, Jeavons S, Schapkaitz I, et al. N-acetyl cysteine for depressive symptoms in bipolar disorder
–a double-blind randomized placebo-controlled trial. Biol Psychiatry. 2008; 64:468–475
Berk M, Copolov DL, Dean O, Lu K, Jeavons S, Schapkaitz I, et al. N-acetyl cysteine as a glutathione precursor for schizophrenia
–a double-blind randomized placebo-controlled trial. Biol Psychiatry. 2008; 64:361–368
Berk M, Dean O, Drexhage H, McNeil JJ, Moylan S, O’Neil A, et al. Aspirin: a review of its neurobiological properties and therapeutic potential for mental illness. BMC Med. 2013; 11:74
Boullin DJ, Orr MW. Effects of aspirin and prostaglandin E2 on secondary phase aggregation responses of schizophrenic patients treated with chlorpromazine. Br J Clin Pharmacol. 1976; 3:929–933
Borges RS, Castle SL. The antioxidant properties of salicylate derivatives: a possible new mechanism of anti-inflammatory activity. Bioorg Med Chem Lett. 2015; 25:4808–4811
Boyd LJ, Lombardi AA, Svigals CS. Sodium gentisate in the treatment of chronic arthritis. Bulletin NY Med Coll. 1950; 13:91–96
Calderón Guzmán D, Trujillo Jiménez F, Hernández García E, Juárez Olguín H. Assessment of antioxidant effect of 2,5-dihydroxybenzoic acid and vitamin a in brains of rats with induced hyperoxia. Neurochem Res. 2007; 32:1036–1040
Capuron L, Neurauter G, Musselman DL, Lawson DH, Nemeroff CB, Fuchs D, Miller AH. Interferon-alpha-induced changes in tryptophan metabolism. Relationship to depression and paroxetine treatment. Biol Psychiatry. 2003; 54:906–914
Chen YH, Lin SJ, Chen YL, Liu PL, Chen JW. Anti-inflammatory effects of different drugs/agents with antioxidant property on endothelial expression of adhesion molecules. Cardiovasc Hematol Disord Drug Targets. 2006; 6:279–304
Choi SH, Aid S, Bosetti F. The distinct roles of cyclooxygenase-1 and -2 in neuroinflammation: implications for translational research. Trends Pharmacol Sci. 2009; 30:174–181
Choi SH, Aid S, Choi U, Bosetti F. Cyclooxygenases-1 and -2 differentially modulate leukocyte recruitment into the inflamed brain. Pharmacogenomics J. 2010; 10:448–457
Cleland LG, Lowthian PJ, Imhoff D, Bochner F, Betts WH, O’Callaghan J. Plasma and synovial fluid gentisate in patients receiving salicylate therapy. J Rheumatol. 1985; 12:136–139
Çakici N, van Beveren NJM, Judge-Hundal G, Koola MM, Sommer IEC. An update on the efficacy of anti-inflammatory agents for patients with schizophrenia
: a meta-analysis. Psychol Med. 2019; 49:2307–2319
Das I, Khan NS. Increased arachidonic acid induced platelet chemiluminescence indicates cyclooxygenase overactivity in schizophrenic subjects. Prostaglandins Leukot Essent Fatty Acids. 1998; 58:165–168
Deepmala, Slattery J, Kumar N, Delhey L, Berk M, Dean O, et al. Clinical trials of N-acetylcysteine in psychiatry and neurology: a systematic review. Neurosci Biobehav Rev. 2015; 55:294–321
Dunn AJ, Swiergiel AH, de Beaurepaire R. Cytokines as mediators of depression: what can we learn from animal studies? Neurosci Biobehav Rev. 2005; 29:891–909
Exner M, Hermann M, Hofbauer R, Kapiotis S, Speiser W, Held I, et al. The salicylate metabolite gentisic acid
, but not the parent drug, inhibits glucose autoxidation-mediated atherogenic modification of low density lipoprotein. FEBS Lett. 2000; 470:47–50
Farokhnia M, Azarkolah A, Adinehfar F, Khodaie-Ardakani MR, Hosseini SM, Yekehtaz H, et al. N-acetylcysteine as an adjunct to risperidone for treatment of negative symptoms in patients with chronic schizophrenia
: a randomized, double-blind, placebo-controlled study. Clin Neuropharmacol. 2013; 36:185–192
Fernandes BS, Dean OM, Dodd S, Malhi GS, Berk M. N-acetylcysteine in depressive symptoms and functionality: a systematic review and meta-analysis. J Clin Psychiatry. 2016; 77:e457–e466
Fond G, Leboyer M, Macgregor A, Girerd N. Aspirin for prevention of cardiovascular events in bipolar disorders. Psychiatry Res. 2014; 219:238–239
Gorsuch MT. Clinical and laboratory investigation of sodium gentisate as an anti-rheumatic treatment. Med Womans J. 1950; 57:9–23; passim
Greenaway JC, Bark DH, Juchau MR. Embryotoxic effects of salicylates: role of biotransformation. Toxicol Appl Pharmacol. 1984; 74:141–149
Grootveld M, Halliwell B. Aromatic hydroxylation as a potential measure of hydroxyl-radical formation in vivo. Identification of hydroxylated derivatives of salicylate in human body fluids. Biochem J. 1986; 237:499–504
Gu F, Chauhan V, Chauhan A. Glutathione redox imbalance in brain disorders. Curr Opin Clin Nutr Metab Care. 2015; 18:89–95
Gutknecht J. Aspirin, acetaminophen and proton transport through phospholipid bilayers and mitochondrial membranes. Mol Cell Biochem. 1992; 114:3–8
Haarman BC, de Boer MK, van der Gaag CM, Burger H, Nolen WA, Schoevers RA. Aspirin for recurrence prevention in bipolar disorder
-promising, yet clinically understudied? Bipolar Disord. 2019; 21:13–15
Hasebe K, Gray L, Bortolasci C, Panizzutti B, Mohebbi M, Kidnapillai S, et al. Adjunctive N-acetylcysteine in depression: exploration of interleukin-6, C-reactive protein and brain-derived neurotrophic factor. Acta Neuropsychiatr. 2017; 29:337–346
Hermann M, Kapiotis S, Hofbauer R, Seelos C, Held I, Gmeiner B. Salicylate promotes myeloperoxidase-initiated LDL oxidation: antagonization by its metabolite gentisic acid
. Free Radic Biol Med. 1999; 26:1253–1260
Hinz B, Kraus V, Pahl A, Brune K. Salicylate metabolites inhibit cyclooxygenase-2-dependent prostaglandin E(2) synthesis in murine macrophages. Biochem Biophys Res Commun. 2000; 274:197–202
Holmes TJ Jr, Vennerstrom JL, John V. Inhibition of cyclooxygenase mediated by electrochemical oxidation of gentisic acid
. J Biol Chem. 1985; 260:14092–14095
Horrobin DF. Schizophrenia
as a prostaglandin deficiency disease. Lancet. 1977; 1:936–937
Horrobin DF. Schizophrenia
: reconciliation of the dopamine, prostaglandin, and opioid concepts and the role of the pineal. Lancet. 1979; 1:529–531
Horrobin D. Prostaglandins and schizophrenia
. Lancet. 1980; 1:706–707
Horrobin DF, Lieb J. A biochemical basis for the actions of lithium on behaviour and on immunity: relapsing and remitting disorders of inflammation and immunity such as multiple sclerosis or recurrent herpes as manic-depression of the immune system. Med Hypotheses. 1981; 7:891–905
Joshi R, Gangabhagirathi R, Venu S, Adhikari S, Mukherjee T. Antioxidant activity and free radical scavenging reactions of gentisic acid
: in-vitro and pulse radiolysis studies. Free Radic Res. 2012; 46:11–20
Kabra MP, Bhandri SS, Sharma A, Gupta RB. Evaluation of anti-parkinson’s activity of gentisic acid
in different animal models. J Acute Dis. 2014141–144
Keller WR, Kum LM, Wehring HJ, Koola MM, Buchanan RW, Kelly DL. A review of anti-inflammatory agents for symptoms of schizophrenia
. J Psychopharmacol. 2013; 27:337–342
Kessing LV, Rytgaard HC, Gerds TA, Berk M, Ekstrøm CT, Andersen PK. New drug candidates for bipolar disorder
-a nation-wide population-based study. Bipolar Disord. 2019; 21:410–418
Kim YK, Na KS, Myint AM, Leonard BE. The role of pro-inflammatory cytokines in neuroinflammation, neurogenesis and the neuroendocrine system in major depression
. Prog Neuropsychopharmacol Biol Psychiatry. 2016; 64:277–284
Kleinsorge H, Pohl W. Gentisic acid
in therapy of rheumatic polyarthritis. Med Klin. 1953; 48:1038–1040
Köhler O, Petersen L, Benros ME, Mors O, Gasse C. Concomitant NSAID use during antipsychotic treatment and risk of 2-year relapse - a population-based study of 16,253 incident patients with schizophrenia
. Expert Opin Pharmacother. 2016; 17:1055–1062
Kuhn H, Bergmann E. New medicinal treatment of rheumatism. Med Klin. 1954; 49:76–78
Kuhn W, Müller T, Bttner T, Gerlach M. Aspirin as a free radical scavenger: consequences for therapy of cerebrovascular ischemia. Stroke. 1995; 26:1959–1960
Kulmacz RJ. Cellular regulation of prostaglandin H synthase catalysis. FEBS Lett. 1998; 430:154–157
Laan W, Selten JP, Kahn RS, Huisman AM, Heijnen CJ, Grobbee DE, Burger H. Acetylsalicylic acid
as an adjuvant therapy for schizophrenia
. Trials. 2006; 7:31
Laan W, Grobbee DE, Selten JP, Heijnen CJ, Kahn RS, Burger H. Adjuvant aspirin therapy reduces symptoms of schizophrenia
spectrum disorders: results from a randomized, double-blind, placebo-controlled trial. J Clin Psychiatry. 2010; 71:520–527
Lee Y, Subramaniapillai M, Brietzke E, Mansur RB, Ho RC, Yim SJ, McIntyre RS. Anti-cytokine agents for anhedonia: targeting inflammation and the immune system to treat dimensional disturbances in depression. Ther Adv Psychopharmacol. 2018; 8:337–348
Lorico A, Masturzo P, Villa S, Salmona M, Semeraro N, de Gaetano G. Gentisic acid
: an aspirin metabolite with multiple effects on human blood polymorphonuclear leukocytes. Biochem Pharmacol. 1986; 35:2443–2445
Louveau A, Smirnov I, Keyes TJ, Eccles JD, Rouhani SJ, Peske JD, et al. Structural and functional features of central nervous system lymphatic vessels. Nature. 2015; 523:337–341
Meyer K, Ragan C. The antirheumatic effect of sodium gentisate. Science. 1948; 108:281
Minarini A, Ferrari S, Galletti M, Giambalvo N, Perrone D, Rioli G, Galeazzi GM. N-acetylcysteine in the treatment of psychiatric disorders: current status and future prospects. Expert Opin Drug Metab Toxicol. 2017; 13:279–292
Nawa H, Yamada K. Experimental schizophrenia
models in rodents established with inflammatory agents and cytokines. Methods Mol Biol. 2012; 829:445–451
O’Brien JR. Effect of anti-inflammatory agents on platelets. Lancet. 1968; 1:894–895
Opler MGA, Yavorsky C, Daniel DG. Positive and negative syndrome scale (PANSS) training: challenges, solutions, and future directions. Innov Clin Neurosci. 2017; 14:77–81
Ozgová S, Hermánek J, Gut I. Different antioxidant effects of polyphenols on lipid peroxidation and hydroxyl radicals in the NADPH-, fe-ascorbate- and fe-microsomal systems. Biochem Pharmacol. 2003; 66:1127–1137
Pecci J, Foye WO. The avidity of salicylic, gentisic, and salicyluric acids for heavy metal cations. J Am Pharm Assoc Am Pharm Assoc. 1960; 49:411–414
Pekoe G, Van Dyke K, Mengoli H, Peden D, English D. Comparison of the effects of antioxidant non-steroidal anti-inflammatory drugs against myeloperoxidase and hypochlorous acid luminol-enhanced chemiluminescence. Agents Actions. 1982; 12:232–238
Perez-Gonzalez A, Galano A, Alvarez-Idaboy A. Dihydroxybenzoic acids as free radical scavengers: mechanisms, kinetics, and trends in activity. New J Chem. 2014; 38:2639–2652
Rahola JG. Somatic drugs for psychiatric diseases: aspirin or simvastatin for depression? Curr Neuropharmacol. 2012; 10:139–158
Rapado-Castro M, Berk M, Venugopal K, Bush AI, Dodd S, Dean OM. Towards stage specific treatments: effects of duration of illness on therapeutic response to adjunctive treatment with N-acetyl cysteine in schizophrenia
. Prog Neuropsychopharmacol Biol Psychiatry. 2015; 57:69–75
Rapado-Castro M, Dodd S, Bush AI, Malhi GS, Skvarc DR, On ZX, et al. Cognitive effects of adjunctive N-acetyl cysteine in psychosis. Psychol Med. 2017; 47:866–876
Rosenberg EF, Krevsky DA, Kagan BM. Laboratory and clinical experience with sodium gentisate in rheumatic disease. Ann Intern Med. 1952; 36:1513–1519
Rosenblat JD, Gregory JM, McIntyre RS. Pharmacologic implications of inflammatory comorbidity in bipolar disorder
. Curr Opin Pharmacol. 2016; 29:63–69
Sagone AL Jr, Husney RM. Oxidation of salicylates by stimulated granulocytes: evidence that these drugs act as free radical scavengers in biological systems. J Immunol. 1987; 138:2177–2183
Savitz J, Preskorn S, Teague TK, Drevets D, Yates W, Drevets W. Minocycline and aspirin in the treatment of bipolar depression: a protocol for a proof-of-concept, randomised, double-blind, placebo-controlled, 2x2 clinical trial. BMJ Open. 2012; 2:e000643
Savitz JB, Teague TK, Misaki M, Macaluso M, Wurfel BE, Meyer M, et al. Treatment of bipolar depression with minocycline and/or aspirin: an adaptive, 2×2 double-blind, randomized, placebo-controlled, phase IIA clinical trial. Transl Psychiatry. 2018; 8:27
Schrör K, Rauch BH. Aspirin and lipid mediators in the cardiovascular system. Prostaglandins Other Lipid Mediat. 2015; 121:17–23
Sepehrmanesh Z, Heidary M, Akasheh N, Akbari H, Heidary M. Therapeutic effect of adjunctive N-acetyl cysteine (NAC) on symptoms of chronic schizophrenia
: a double-blind, randomized clinical trial. Prog Neuropsychopharmacol Biol Psychiatry. 2018; 82:289–296
Skvarc DR, Dean OM, Byrne LK, Gray L, Lane S, Lewis M, et al. The effect of N-acetylcysteine (NAC) on human cognition - a systematic review. Neurosci Biobehav Rev. 2017; 78:44–56
Sommer IE, de Witte L, Begemann M, Kahn RS. Nonsteroidal anti-inflammatory drugs in schizophrenia
: ready for practice or a good start? A meta-analysis. J Clin Psychiatry. 2012; 73:414–419
Sommer IE, van Westrhenen R, Begemann MJ, de Witte LD, Leucht S, Kahn RS. Efficacy of anti-inflammatory agents to improve symptoms in patients with schizophrenia
: an update. Schizophr Bull. 2014; 40:181–191
Stolk P, Souverein PC, Wilting I, Leufkens HG, Klein DF, Rapoport SI, Heerdink ER. Is aspirin useful in patients on lithium? A pharmacoepidemiological study related to bipolar disorder
. Prostaglandins Leukot Essent Fatty Acids. 2010; 82:9–14
Stone WL, Krishnan K, Campbell SE, Palau VE. The role of antioxidants and pro-oxidants in colon cancer. World J Gastrointest Oncol. 2014; 6:55–66
Webb JR, Webb JE, Stubbe DE, Poncin YB. Aspirin as an adjunctive treatment for childhood onset schizophrenia
. J Child Adolesc Psychopharmacol. 2013; 23:585–586
Weller RO, Engelhardt B, Phillips MJ. Lymphocyte targeting of the central nervous system: a review of afferent and efferent CNS-immune pathways. Brain Pathol. 1996; 6:275–288
Woelfer M, Kasties V, Kahlfuss S, Walter M. The role of depressive subtypes within the neuroinflammation hypothesis of major depressive disorder. Neuroscience. 2019; 403:93–110
Yeh CT, Shih PH, Yen GC. Synergistic effect of antioxidant phenolic acids on human phenolsulfotransferase activity. J Agric Food Chem. 2004; 52:4139–4143
Yeh CT, Yen GC. Induction of hepatic antioxidant enzymes by phenolic acids in rats is accompanied by increased levels of multidrug resistance-associated protein 3 mRNA expression. J Nutr. 2006; 136:11–15
Zhang L, An LT, Qiu Y, Shan XX, Zhao WL, Zhao JP, et al. Effects of aspirin in rats with ouabain intracerebral treatment-possible involvement of inflammatory modulation? Front Psychiatry. 2019; 10:497