Pharmacoepigenetics in schizophrenia: Predicting drug response

Abstract

Individual differences in treatment response in schizophrenia pose a significant challenge in the management of the disease, due to several biological as well as psychosocial factors, including genetic and epigenetic mechanisms. Pharmacoepigenetics investigates how epigenetic mechanisms affect the variability in effectiveness of treatments and adverse side effects. Antipsychotics such as clozapine (atypical) and haloperidol (typical) directly induce epigenetic changes by altering DNA methyltransferases and histone acetyltransferases, while indirectly affecting neuroinflammatory and stress response pathways. Personalized medicine using epigenetic markers (DNA methylation, non-coding RNAs including microRNAs and long non-coding RNAs) holds great promise for improving the drug response and reducing the side effects of antipsychotic treatment. These developments could revolutionize the treatment of schizophrenia by addressing the complexities involved in responding to treatment. However, ethical and technical barriers to implementing strategies based on epigenetic regulation in clinical practice are fundamental challenges that need to be carefully addressed in this field. This review examined the epigenetic mechanisms involved in the efficacy of antipsychotic drugs.

Key Words: Pharmacoepigenetics; Treatment failures; Personalized medicine; Antipsychotic drugs; Schizophrenia

Core Tip: Epigenetic mechanisms play a critical role in the pathogenesis of schizophrenia and contribute to variability in treatment response. Advances in personalized medicine and epigenetic studies offer promising avenues to improve schizophrenia treatment by tailoring therapeutic strategies to individual genetic and epigenetic profiles. However, further research is needed to develop more specific and effective epigenetic therapies to overcome challenges related to treatment efficacy and duration. Integrating epigenetic data into personalized medicine approaches will ultimately improve patient outcomes and revolutionize schizophrenia management.

INTRODUCTION

Schizophrenia is a severe mental disorder characterized by positive and negative symptoms, leading to significant declines in social and occupational functioning. It is a serious mental health disorder that contributes significantly to the global disease burden. According to the World Health Organization, schizophrenia affects approximately 24 million people worldwide. Among adults, this rate is 1 in 222 people (0.45%). It ranks among the top ten causes of disability in individuals aged 15 to 44[1]. Schizophrenia also imposes a substantial economic burden on societies due to healthcare costs, productivity loss, and the strain on caregivers. In the United States alone, the annual cost associated with schizophrenia is estimated to be around $170 billion[2]. These statistics underscore the significant impact of schizophrenia on individuals and societies, highlighting the need for effective treatment strategies and comprehensive support systems.

Antipsychotics are effective in reducing the severity of positive symptoms, preventing relapses, and lowering hospitalization rates. In addition, several psychosocial therapeutic options targeting negative and cognitive symptom domains are available as adjunctive approaches to antipsychotic treatment. However, about 20%-30% of subjects with schizophrenia do not respond to at least two different trials with antipsychotics of adequate dose and duration and are identified as treatment-resistant subjects.

Heterogeneity is not an exception but rather a defining characteristic of schizophrenia, manifesting in its clinical presentation, environmental and genetic determinants, pathophysiological correlates, and treatment. The outcome rates also exhibit significant heterogeneity. In a seminal 22-year follow-up study, Huber et al[3] examined the course of schizophrenia, identifying 12 distinct illness trajectories, and emphasized the difficulty in predicting outcomes in individual cases. A portion of this variance has been attributed to differences in outcome measures, particularly in response, remission, and recovery, in studies conducted in the late 20^th century[4]. The remission criteria in schizophrenia were formally defined by the Remission in Schizophrenia Working Group in 2005 to establish a standardized approach for assessing symptomatic improvement in clinical and research settings[5]. Nevertheless, heterogeneity persisted in longitudinal studies using the Remission in Schizophrenia Working Group criteria. In a pooled review of treatment outcomes after the first psychotic episode, it was shown that remission rates varied significantly between 17%-78%[6]. In a more recent analysis of remission rates, the overall pooled prevalence remission rate was 56.9% (95% confidence interval: 48.9-64.5)[7]. In a comprehensive review, the World Psychiatric Association Pharmacopsychiatry Section analyzed data from approximately 1600 randomized controlled trials on antipsychotic treatment in schizophrenia and reported significant individual variability in both treatment response and susceptibility to adverse effects highlighting the necessity of personalized medicine in treatment of schizophrenia[8].

PERSONALIZED MEDICINE IN SCHIZOPHRENIA

The treatment of schizophrenia primarily involves the use of antipsychotic medications, which are categorized into two main classes: First-generation antipsychotics and second-generation antipsychotics. First-generation antipsychotics primarily function as dopamine D2 receptor antagonists, effectively reducing positive symptoms such as hallucinations and delusions. Second-generation antipsychotics, on the other hand, operate through a broader mechanism. They act on both dopamine and serotonin receptors, which helps lower the risk of extrapyramidal side effects and aims to prevent the worsening of negative symptoms in schizophrenia[9].

Genome-wide association studies have revealed that schizophrenia is a polygenic disorder with different genetic profiles among patients[10]. This genetic diversity contributes to varying treatment responses and complicates the adoption of a single treatment approach. Approximately 30% of individuals diagnosed with schizophrenia do not respond favorably to conventional antipsychotic therapy, and less than 40% achieve full remission of symptoms[10]. More than 70% of patients with chronic schizophrenia discontinue antipsychotic medication primarily due to poor efficacy[11]. Considerable amount of phenotypical heterogeneity also exists. Therefore, potential treatment approaches tailored to an individual’s genetic, phenotypic, or psychosocial characteristics are necessary. This shift will allow schizophrenia treatment to move away from a trial-and-error approach and toward a more personalized model[12]. Personalized medicine provides numerous benefits, including improved diagnostic accuracy, selection of the most effective treatment for a patient, increased likelihood of successful therapy, reduction of side effects, better disease prevention, cost reduction in healthcare, and promotion of research and innovation[13,14]. Personalized medicine does not aim to develop new drugs for patients but instead categorizes individuals into subpopulations that exhibit different responses to therapeutic agents for specific diseases[13]. There is no universally accepted definition for this concept. The goal is to adapt the right treatment strategy to the right patient at the right time and/or determine disease susceptibility and/or establish a medical model that characterizes an individual’s phenotype and genotype for timely treatment[14].

Personalized medicine involves the development of individualized treatment strategies based on a person’s genetic profile to predict disease, prevent its onset, and make decisions regarding lifestyle and disease management[14,15]. However, genetic profiles can explain the differences between individuals to some extent[16,17]. Recently, there has been an increase in the number of studies suggesting that epigenetic changes as an alternate mechanism may also contribute to interindividual drug response variability by changing expression levels of various genes[18] (Figure 1). The interaction of environmental factors with risk genes involved in the pathophysiology of schizophrenia, as well as the individual or combined use of antipsychotic drugs to treat this disease, may result in alterations to epigenetic homeostasis. Prior to the investigation of epigenetic changes induced by environmental factors and antipsychotic medications, a concise overview of the predominant epigenetic mechanisms will be provided.

Figure 1 Preferred Reporting Items for Systematic Reviews and Meta-Analyses flowchart illustrating the selection process of studies included in this review.

EPIGENETIC MODIFICATIONS

The term “epigenetics” was introduced by Waddington (1939), who proposed the term “epigenetic landscape” to describe the molecular and biological mechanisms that transform a genetic trait into a phenotype[19]. Epigenetic modifications refer to changes in gene expression independent of DNA sequence alterations. That is, epigenetics turns on or off different genes that can make us prone to disease. Epigenetic mechanisms include DNA methylation, RNA methylation, histone acetylation, histone methylation, and RNA regulation via non-coding RNAs (ncRNAs), such as long ncRNAs (lncRNAs) and microRNAs (miRNAs)[20]. Among those, miRNAs, DNA methylation, and histone modification are accepted as three primary alterations[19]. These mechanisms regulate gene expression patterns by modifying DNA accessibility and chromatin structure. Such processes play a crucial role in the normal development and differentiation of various cell lineages in an adult organism[21]. Researchers state that the division and differentiation events of a single cell during embryogenesis are associated with epigenetics[22]. In multicellular organisms, different phenotypes may arise from the same genotype due to the possibility of transferring epigenetic markers formed during development to offspring[23]. As a result, monozygotic twins have the same genetic information, but they may have a different epigenetic profile determined by the environment in which they live and grow, which may cause changes in health and disease phenotype[23].

Changes in the internal and external environments of biological systems, such as oxidative stress and dietary alterations, can lead to epigenetic modifications[15]. These changes may have evolved to provide flexibility in adapting to environmental cues[24]. Understanding how lifestyle factors such as sex, disease onset age and progression, comorbidities, medication use, diet, exercise, and stress influence the epigenome can pave the way for new preventive strategies and personalized medicine approaches[25]. Unlike genetic changes that are difficult to reverse, epigenetic aberrations are pharmaceutically reversible. With the development of drugs that target specific epigenetic mechanisms involved in the regulation of gene expression, the development and use of epigenetic tools is a feasible and effective approach that can be clinically applied to the treatment of various diseases[19].

DNA methylation

DNA methylation is a well-characterized and extensively studied epigenetic modification that emerged with research by Mahler and Griffith in 1969, who suggested that DNA methylation may play an important role in the function of long-term memory[26]. DNA methylation regulates gene expression, X chromosome inactivation, imprinting, and chromosomal organization[19]. It is also considered one of the most critical modifications leading to disease[15,27]. Aberrant methylation in gene regions may contribute to the development of cancer, autoimmune disorders, and neurodegenerative diseases[19,27]. Moreover, DNA methylation can occur in genes encoding drug-metabolizing enzymes, drug transporters, and drug targets, thereby altering pharmacokinetics and pharmacodynamics[27].

DNA methylation occurs through the addition of a methyl group (-CH3) to cytosine residues in DNA. The addition of a methyl group to the intact C-5 position of a cytosine during DNA replication may contribute to the establishment of de novo DNA methylation[28]. This process typically takes place in promoter regions containing CpG dinucleotide-rich sequences, known as CpG islands, which cover approximately 60% of the promoter region. These regions are at least 200 bp and are more than 55% conserved throughout evolution. The original epigenetic status must be maintained for normal development. Disruption of this balance can lead to an abnormal epigenetic landscape depending on time and location. DNA located in some promoter regions can cause heritable transcriptional silencing when methylated[19].

The enzymes responsible for DNA methylation belong to the DNA methyltransferase (DNMT) family, classified into five groups[15]. Three major DNMTs have been identified: DNMT1, DNMT3a, and DNMT3b. In addition to these, DNMT3 L is also present, although it lacks catalytic activity[20]. Active demethylation is mediated by ten-eleven translocation family proteins, which facilitate the conversion of methylated cytosines into unmethylated cytosines through thymine-DNA glycosylase-mediated base excision repair[20].

Histone modifications

Histone complexes consist of two unstable dimers H2A and H2B. A tetramer of H3 and H4 wrapped with 147 bp of DNA to form the nucleosome. The histone complex facilitates the condensation of genomic DNA and is involved in post-transcriptional modification[19]. More than 70 different histone amino acid modifications, including methylation, acetylation, phosphorylation, ubiquitination, and SUMOylation, have been identified[19,20]. Acetylation occurs through the transfer of an acetyl group from acetyl-CoA to lysine residues on the N-terminal tails of histone proteins, a process mediated by histone acetyltransferases. This modification can be reversed by histone deacetylases (HDACs)[15,20]. These enzymes cause DNA to wrap tightly around histone proteins, leading to reduced gene transcription. Acetylation usually allows active transcription, whereas hypoacetylation leads to transcriptional inactivity. Histone methylation can indicate both active and inactive transcription, and monomethylation, dimethylation, and trimethylation status have different effects[19,20]. The balance between these two dynamic processes plays a crucial role in chromatin architecture, gene transcription, cell cycle progression, growth, apoptosis, differentiation, DNA replication, and repair mechanisms[20]. Histone proteins regulate the spatial and temporal fine-tuning and coordination of gene expression during neurogenesis. Mutations in histone-related enzymes can lead to the development of diseases such as cancer, autoimmune diseases, endocrine diseases, and psychological disorders[25].

NcRNAs

NcRNAs, such as miRNAs, lncRNAs, and small interfering RNAs, play a critical role in regulating gene expression through various mechanisms at both the transcriptional and post-transcriptional levels. MiRNA is a class of small ncRNAs, approximately 22 nucleotides long. About two-thirds of the mRNA in the human body is regulated by miRNAs, and the proportion of genes that are regulated by miRNAs is more than 60%. MiRNAs cannot be translated into proteins. Instead, they can regulate gene expression without altering the genetic code. Their main function is to downregulate gene expression in different ways, including mRNA splicing, translational inactivation, and deadenylation to produce a mitotically heritable outcome[19]. The miRNAs of the human genome are localized on chromosomes, either individually or in clusters. MiRNAs are classified by location as intronic, exonic, intergenic, and intragenic. A single miRNA can bind to up to 200 target genes, each with a different function, and can regulate hundreds of functions in a cell. Recent studies have shown that miRNAs are highly stable in extracellular fluids, including saliva, plasma, bronchial lavage, pleural fluid, cerebrospinal fluid, and peritoneal fluid. They are potential candidates for biomarker discovery due to their extracellular stability and accessibility. MiRNAs are potential regulators of fundamental cellular processes through a various mechanisms. These regulatory processes are essential for maintaining cellular homeostasis and are often disrupted in diseases, particularly cancer and neurodegenerative disorders[19,25].

NcRNAs are of significant importance in controlling gene expression through epigenetic regulation, particularly in human development and disease. These ncRNAs can significantly influence chromatin structure, transcription, and post-transcriptional gene expression. For instance, lncRNAs can alter the chromatin state by recruiting chromatin-modifying complexes to specific genomic regions, thus activating or repressing the transcription of target genes. Research has shown that ncRNA epigenetics impacts nearly all aspects of RNA metabolism, controlling ncRNA stability, the processing of miRNAs, and interactions between lncRNAs and RNA-binding proteins[25]. Studies in this field provide insights into how drugs affect ncRNA genes, exploring mechanisms of action and potential therapeutic interventions.

ENVIRONMENTAL FACTORS INFLUENCING SCHIZOPHRENIA PROGNOSIS: AN EPIGENETIC PERSPECTIVE

Many factors contribute to individual variability in treatment outcomes in schizophrenia. Some of the factors associated with poor treatment response in schizophrenia describe a ‘state’ rather than direct ‘exposure.’ For example, being male, early age of onset, insidious onset, the non-affective nature of psychosis, low premorbid intellectual functioning, and treatment non-adherence fall into this category. In contrast, some risk factors linked to the course of schizophrenia define an actual ‘exposure’ rather than a state. These include cultural influences, cannabis use, childhood trauma, and the duration of untreated psychosis (DuP). Epigenetic research investigates the ultimate effects of an environmental exposure. Therefore, in this section, exposure-based factors that have been associated with a poor prognosis and/or low treatment response in schizophrenia provide a framework for investigating epigenetic effects that will briefly be discussed.

Among these exposure-based factors, the ‘global outcome discrepancy’ perhaps operates on a broader scale. This phenomenon was first identified in the Determinants of Outcomes of Severe Mental Disorders study conducted by the World Health Organization in the early 1980s, which suggested that schizophrenia outcomes tend to be more favorable in developing countries compared to developed ones[29]. The study found that recovery rates were higher in developing countries, with an average of 37% compared to 15.5% in developed countries. Moreover, patients in developing countries experienced longer periods of intact social functioning despite a significantly lower proportion receiving continuous medication. This paradoxical gradient may be associated with differences in the socio-cultural fabric of developing and developed countries, which may foster stronger family bonds and greater social capital, ultimately leading to better support. Although controversial views exist[30], this finding is particularly interesting as it exemplifies the role of culture in the context of gene-environment interactions that shape the course of schizophrenia.

DuP refers to the time elapsed between the onset of the first psychotic symptom and the initiation of antipsychotic therapy. A well-established finding in the literature is the association between longer DuP and poorer treatment outcomes. Retrospective studies have linked prolonged DuP to increased symptom severity and cognitive dysfunction, while prospective research has identified its association with reduced remission rates[31-33]. Although there is no study directly investigating the relationship between longer DuP and epigenetic effects, the chronic stress caused by prolonged untreated psychosis and/or related environmental factors may lead to epigenetic modifications, which in turn could influence the course of the disease.

Comorbid substance use disorders particularly cannabis and alcohol use, lead to prominent negative outcomes in subjects with schizophrenia. Substance use disorder occurring alongside schizophrenia has been associated with worsening clinical symptoms, reduced treatment adherence, impaired global functioning, heightened risk of violence and suicide, and an increased likelihood of relapse and re-hospitalization[34]. It has been shown that cannabis use, in particular, may influence the severity and long-term trajectory of schizophrenia via epigenetic mechanisms. Regular cannabis use in diagnosed patients with schizophrenia has been associated with epigenetic modifications in cannabinoid receptor 1 and DA receptor subtype 2 (DRD2) genes[35], which may contribute to the worsening of clinical symptoms. Epigenetic modifications in catechol-O-methyltransferase, glucocorticoid receptor gene (NR3C1), and brain-derived neurotrophic factor (BDNF) genes have been linked to altered antipsychotic treatment response in patients with schizophrenia who use cannabis[36,37]. DNA methylation changes include hypermethylation of catechol-O-methyltransferase, which affects dopamine metabolism[35], regional hypomethylation of discs-large associated protein 2, which involved in synapse organization, neuronal signaling[38], and the increase in cannabinoid receptor 2 and methylating enzymes DNMT3a and DNMT1 mRNAs[39]. Histone modifications such as increased trimethylation of histone H3 at lysine 9 (H3K9me3) marks in DRD2 alter dopamine receptor expression[35], while histone acetylation changes in BDNF may impair neuroplasticity, potentially reducing antipsychotic efficacy. Furthermore, miRNA regulation is also implicated, with altered expression of miRNA profiles detected in the entorhinal cortex of adolescent rats exposed to cannabis[40], an area that is affected even during the first psychotic episode in schizophrenia[41].

Childhood trauma has been associated with poorer treatment outcomes in schizophrenia, including lower remission rates, increased symptom severity, and higher relapse risk. Patients with a history of severe childhood trauma require higher doses of antipsychotic medication yet exhibit lower remission rates compared to those with minimal trauma exposure[42]. Additionally, patients with first-episode psychosis with a history of childhood trauma display more severe depressive symptoms and agitation during the first year of treatment, negatively affecting their response to antipsychotic therapy[43]. The strongest finding that links childhood trauma to epigenetic mechanisms is the association between DNA methylation in the exon 1F region of the glucocorticoid receptor NR3C1 gene[44]. Changes in NR3C1 methylation have been consistently observed in individuals exposed to childhood trauma. These epigenetic modifications have been suggested to be associated with dysregulated stress response in the hypothalamic-pituitary-adrenal axis. At least one study reported an association of NR3C1 polymorphism and attempted suicide in subjects with schizophrenia[45].

A recent meta-analysis on predictors of schizophrenia outcomes found that greater baseline symptom severity, poorer initial global functioning, longer DuP, and a higher number of previous hospital admissions were all associated with worse prognosis in schizophrenia[46]. Collectively, these factors suggest a potential link between the amount of experience of psychosis and its trajectory. Direct evidence on the epigenetic effects of experiencing psychosis itself remains limited. However, shortened telomere length is a well-replicated finding in schizophrenia, and it has been associated with greater severity of psychotic symptoms[47], poor clinical outcomes[48], and reduced response to antipsychotic therapy[49]. Telomere length has been proposed as an ‘epigenetic trait’ influenced not only by telomerase expression but also by environmental factors. Early-life exposure to stress has been suggested to precondition adult telomere length shortening, which in turn may increase susceptibility to disease[50]. Thus, telomere shortening may mediate the impact of psychosis on its own long-term outcome.

Some exposure-based factors associated with schizophrenia comprise elements for which there is no consensus on their definition or agreement on their operational criteria. These include family climate-related factors, such as high levels of ‘expressed emotion,’ as well as complex social factors like social defeat, social capital, and urban exposure. Regarding the epigenetic effects of parenting, a study found that retrospectively reported levels of maternal care were associated with lower DNA methylation status in exon III of the oxytocin receptor gene and BDNF in peripheral mononuclear blood cells in adults[51]. However, investigating the epigenetic effects of such complex social factors presents a far more challenging process, requiring more nuanced methodologies and advanced research approaches.

EFFECTS OF ANTIPSYCHOTICS ON SCHIZOPHRENIA EPIGENETICS

Reverse pharmacoepigenetic is a term used to describe the process of pharmacological agents triggering epigenetic alterations and causing changes in gene expression patterns. This concept underscores the potential of drugs as epigenetic modulators, thereby transforming the epigenetic landscape of cells and biological outcomes[52,53]. It emphasizes the bidirectional relationship between pharmacological action and epigenetic changes, a concept that can help design personalized and more effective therapeutic strategies[54]. The main epigenetic mechanisms influenced by antipsychotics include DNA methylation, histone modifications, and ncRNA regulation (Table 1)[55,56].

Table 1 Effects of antipsychotic drugs and environmental factors on the epigenetics of schizophrenia.

Mechanism	Drugs/environmental factors	Result/summary	Ref.
DNA methylation	Clozapine	Clozapine increased methylation and gene expression	[69]
	Olanzapine	Olanzapine increased expression of histone deacetylase genes
	Risperidone	Risperidone activated some genes while suppressing others
	Haloperidol	Haloperidol caused the most substantial increase in gene expression and significant effects on histone modification
	Risperidone, olanzapine, clozapine	DNA methylation was highest in non-responders. Methylation levels may tend to decrease in treatment responders	[56]
	Clozapine	Methylation sites associated with schizophrenia were found to be hypermethylated, and dynamic epigenetic regulation was identified	[59]
	Various antipsychotics	Clozapine and olanzapine induced global hypermethylation, whereas risperidone and quetiapine led to hypomethylation in specific regions	[64]
	Risperidone	Good responders showed hypermethylation in genes linked to neuroplasticity and cognition. Poor responders exhibited hypermethylation in genes involved in drug metabolism and immune response	[67]
	Childhood trauma	Childhood trauma was shown to induce epigenetic changes in stress-related genes, potentially contributing to schizophrenia	[44]
	Cannabis	Cannabis use was found to increase DNMT enzymes that suppress GABA synthesis, potentially raising the risk of schizophrenia through epigenetic mechanisms	[39]
	Risperidone	Risperidone treatment was found to shift DNA methylation patterns in patients with schizophrenia closer to those of healthy individuals	[66]
	Various antipsychotics	Clozapine, haloperidol, and atypical antipsychotics increased methylation, whereas risperidone caused decreased CYP2D6 methylation, possibly affecting drug metabolism	[65]
	Haloperidol, risperidone	Haloperidol and risperidone were found to induce significant hypermethylation in CpG regions	[62]
	Various antipsychotics	While no significant changes in DNA methylation were detected, widespread differences in gene expression levels were observed	[70]
	Clozapine, various antipsychotics	The majority of schizophrenia-associated differentially methylated positions were hypermethylated. Significant clozapine-associated DMPs were identified, all showing hypermethylation	[60]
	Cannabis	Early exposure to cannabis disrupted dopaminergic and endocannabinoid systems through decreased DNA methylation in the DRD2 gene and increased expression of the CNR1 gene, leading to schizophrenia-like behaviors	[35]
	Cannabis	Cannabis use was found to increase DLGAP2 gene expression through hypomethylation, potentially raising the risk of neurodevelopmental disorders such as autism and schizophrenia	[38]
	Marijuana	Marijuana was shown to increase synaptic plasticity by increasing BDNF levels through DNA hypomethylation, leading to an increased risk of disease in individuals genetically predisposed to schizophrenia	[36]
	Aripiprazole	Methylation levels of the ANKK1 gene were identified as a potential biomarker for predicting aripiprazole treatment response	[68]
	Clozapine	While global methylation levels generally decreased, the specific increase in methylation of certain genes contributed to the therapeutic effects of clozapine in treatment-resistant schizophrenia	[58]
	Haloperidol, clozapine, olanzapine	These drugs caused increased global DNA methylation levels and expression of epigenetic genes, including DNA methyltransferases, methyl CpG binding proteins, and DNA demethylases	[18]
	Risperidone	Elevated methylation rate within the promoter regions of the CYP3A4 and CYP2D6 genes was observed	[91]
DNA methylation/miRNA	Haloperidol, clozapine, olanzapine	Haloperidol had the strongest effect on methylation, followed by clozapine and olanzapine, with miR-29b downregulation and increased expression of DNMT1 and DNMT3A contributing to the hypermethylation	[57]
DNA methylation/histone modification	Cannabis exocannabinoid	Cannabis caused DNA hypomethylation in DRD2, COMT, DLGAP2, STAT3, PENK genes and hypermethylation in DNMT3a/b, NCAM1, and AKT1 genes. It also increased histone acetylation and methylation	[37]
Histone-chromatin modification	Various antipsychotics	Haloperidol increased histone acetylation and modulated DRD2 gene methylation and expression in the striatum. Olanzapine increased histone acetylation at BDNF. Clozapine increased histone methylation at GAD1. Risperidone reduced H3K27ac levels in the striatum, impacting genes like DRD2 and inflammatory pathways. Quetiapine modified global H3K9me2 levels in the prefrontal cortex	[64]
	Various antipsychotics	In individuals receiving antipsychotic treatment, HDAC activity decreased, and histone acetylation increased, leading to an increase in gene expression	[74]
	Clozapine, haloperidol, risperidone	Clozapine reduced histone acetylation in the GRM2 gene and increased histone acetylation and methylation in the GRM3 gene	[77]
	Various antipsychotics	Antipsychotic use reversed schizophrenia-related epigenetic changes in some genes and caused new changes in others. Antipsychotics altered histone acetylation and methylation, especially in dopamine-related and glutamate-related genes	[78]
	Clozapine, risperidone, haloperidol	Clozapine altered histone acetylation and methylation at ADRA2A and ADRA2C promoters. By contrast, risperidone and haloperidol caused no significant changes in epigenetic markers or gene expression	[75]
	Clozapine	Clozapine caused decreased H3 acetylation and increased HDAC2 binding at the mGlu2 promoter, which led to suppressed gene expression through enhanced histone deacetylation. These effects were absent in HDAC2 knockout mice	[76]
ncRNAs (lncRNAs)	Atypical antipsychotics	Risperidone regulated dopamine signaling by increasing MIAT levels. Atypical antipsychotics generally normalized NEAT1 levels. LncRNAs such as MIAT and NEAT1 played central roles in the pathogenesis of schizophrenia	[79]
	Various antipsychotics	Antipsychotics decreased MEG3, PINT, and GAS5 levels. Antipsychotics can modulate the immune system by decreasing the level of MEG3. Risperidone may regulate the immune response by altering heterochromatin-related genes	[89]
ncRNAs (miRNAs)	Clozapine (used only for TRS patients)	MiR-675-3p levels were elevated in TRS patients, and clozapine further increased miR-675-3p levels in TRS patients. Clozapine also increased miR-675-3p levels in cell culture	[88]
	Haloperidol	Haloperidol caused downregulation of miR-137-3p and increased Nr3c1 mRNA expression	[86]
	Olanzapine	Olanzapine significantly reduced miR-195 levels in treatment-responsive patients, which was positively correlated with the total score of PANSS	[83]
	Clozapine	Clozapine counteracted NMDA blockade, improved cognitive function, normalized some miRNA levels - especially decreasing miR-184 - and modulated the estrogen signaling pathway, potentially relevant for treating schizophrenia in women	[87]
	Olanzapine, haloperidol	Olanzapine and haloperidol reduced the disease-related increase in miR-223 levels in neurons, reducing pressure on glutamate receptors and indirectly contributing to regulating glutamate receptor levels	[84]
	Haloperidol, clozapine	Haloperidol and clozapine increased miRNA-143 levels and suppressed NRG1 by blocking the D2 receptor, altering neuroplasticity and neurotransmission	[85]
	Haloperidol, clozapine, olanzapine	In vitro studies, the drugs downregulated miR-27a and increased the expression of ABCB1 and CYP3A4 genes. CYP2D6 was upregulated only by haloperidol. The downregulation of miR-27a and increased expression of ABCB1 were also observed in the clinical setting	[82]
Oxidative stress	Various antipsychotics	Long-term antipsychotic use may increase dopamine turnover and trigger free radical production and oxidative stress. Low DNA methylation may increase CYP2E1 gene expression, which may trigger the development of tardive dyskinesia by raising the level of oxidative stress	[92]

ABCB1: ATP-binding cassette sub-family B member 1; ADRA2C: Adrenergic alpha 2C receptor; ADRD2A: Adrenergic receptor alpha 2; AKT: Protein kinase B; ANKK1: Ankyrin repeat and kinase domain containing 1; BDNF: Brain-derived neurotrophic factor; CNR1: Cannabinoid receptor 1; COMT: Catechol-O-methyltransferase; CYP2D6: Cytochrome P450 2D6; DLGAP2: Discs-large associated protein 2; DMPs: Differentially methylated probes; DNMT: DNA methyltransferase; DRD2: DA receptor subtype 2; GABA: Gamma-aminobutyric acid; GAD1: Glutamate decarboxylase 1; GAS5: Growth arrest-specific transcript 5; GRM: Glutamate receptor; H3K27ac: H3 acetylation of lysine 27; HDAC: Histone deacetylase; MEG3: Maternally expressed gene 3; MIAT: Myocardial infarction associated transcript; miRNA: MicroRNA; NMDA: N-methyl D-aspartate; NCAM1: Neural cell adhesion molecule 1; NEAT1: Nuclear paraspeckle assembly transcript 1; NMDA: N-methyl D-aspartate; Nr3c1: Glucocorticoid receptor gene; NRG1: Neuregulin 1; PANSS: Positive and Negative Syndrome Scale; PENK: Proenkephalin; PINT: p53-induced transcript; STAT3: Signal transducer and activator of transcription 3; TRS: Treatment-resistant.

Antipsychotics have been shown to affect DNA methylation patterns, which can have significant implications for gene expression and, consequently, the clinical manifestations of schizophrenia[52,53,55]. Swathy and Banerjee[18] reported that antipsychotic drugs increased global DNA methylation, which is associated with elevated expression of epigenetic genes, including DNMTs, methyl CpG binding proteins, and DNA demethylases. Swathy et al[57] demonstrated that haloperidol, clozapine, and olanzapine caused global DNA hypermethylation in both in vitro models and patients with schizophrenia. Atypical antipsychotics exhibited lower global methylation compared to typical antipsychotics. Clozapine induces hypermethylation at specific CpG sites, while hypomethylation occurs at the genome-wide level. While global methylation levels tend to decline, the specific increase in methylation of certain genes appears to play a role in the therapeutic effects of clozapine in treatment-resistant schizophrenia[58]. Moreover, the majority of schizophrenia-related differentially methylated positions were found to be hypermethylated in clozapine treatment[59,60]. A survey with monozygotic twins showed that genes related to neuronal and synaptic functions were observed at a higher frequency among genes with altered DNA methylation and expression in the clozapine-responsive sibling than in the clozapine-unresponsive sibling[61]. However, another study suggested that methylation levels may tend to decrease in treatment responders, showing the highest level of DNA methylation in non-treatment responders[56].

Both haloperidol and risperidone were reported to induce hypermethylation at CpG sites located away from the promoter, indicating a complex pattern of methylation changes that could influence gene expression related to schizophrenia[62]. Risperidone displayed gene-specific methylation alterations. For instance, it reduces cytochrome P450 2D6 (CYP2D6) methylation and is associated with DRD2 hypermethylation, which may impact drug metabolism[63-65]. Risperidone has also been associated with changes in DNA methylation in peripheral blood, which parallels neuroimaging findings and cognitive phenotypes, suggesting that these epigenetic modifications may reflect underlying neurobiological changes. A high percentage of the changes in DNA methylation contributed positively to treatment, as methylation levels approached those of healthy individuals[66]. In addition, risperidone responders showed hypermethylation in genes linked to neuroplasticity and cognition[67].

A study with aripiprazole showed that responders had higher pre-treatment ankyrin repeat and kinase domain containing 1 methylation levels compared to non-responders. It suggests that ankyrin repeat and kinase domain containing 1 methylation is a predictor of treatment response rather than being a direct consequence of aripiprazole administration[68]. A methylome-wide analysis was conducted that hypermethylation was associated with treatment response, while hypomethylation was linked to treatment resistance. Antipsychotic treatment could reverse pathological DNA methylation patterns, influencing gene expression, and potentially improving treatment outcomes[69]. By contrast, Perzel Mandell et al[70] demonstrated altered expression of the genes in synaptic signaling and neurogenesis pathways despite the absence of significant DNA methylation changes that were not detected. This finding suggests that antipsychotics may influence gene expression through mechanisms apart from DNA methylation.

Histone modifications, such as acetylation and methylation, are essential in regulating gene expression and chromatin structure. Antipsychotic medications have been shown to significantly change histone modification in various brain regions associated with schizophrenia[55,71]. Osuna-Luque et al[72] demonstrated that atypical antipsychotics, such as risperidone and aripiprazole, cause histone modifications, including acetylation, in Caenorhabditis elegans, which can be transmitted to subsequent generations. Feiner et al[73] also reported that risperidone induced histone modifications through methylation and phosphorylation changes and relaxed heterochromatin.

Haloperidol increases histone acetylation and modulates DRD2 gene methylation in the striatum, while risperidone reduces H3 acetylation of lysine 27 (H3K27ac) levels in the striatum, thereby impacting genes such as DRD2 and inflammatory pathways. Quetiapine modifies global H3 dimethylation of lysine 9 levels in the prefrontal cortex, suggesting a role in chromatin remodeling and transcriptional regulation[63]. Martínez-Peula et al[74] compared histone posttranslational modifications and HDAC activity in individuals taking antipsychotic medication to non-users by using autopsy dorsolateral prefrontal cortex samples of patients with schizophrenia. They demonstrated an increased expression of H3 dimethylation of lysine 9, H3K4me3, and H3K9ac levels and a decreased HDAC activity, resulting in increased histone acetylation and suggesting a more open chromatin structure.

Clozapine remains the first-line treatment for treatment-resistant schizophrenia, despite other antipsychotics being approximately equally effective. Even 65 years after its discovery, clozapine continues to hold its place as the preferred option for treatment-refractory psychosis. However, the exact mechanism underlying its unique efficacy remains unclear, and epigenetic effects may indeed play a role in its distinct therapeutic profile. It has been shown that clozapine impacts the synapse-related genes adrenergic receptor alpha 2 and adrenergic alpha 2C receptor by regulating both histone modifications[75], causing suppression of the glutamatergic system receptor (mGlu2R) gene through increased histone deacetylation[76], and reducing histone acetylation in the mGlu2R gene[77]. In a study examining the frontal cortex of patients with schizophrenia, it was suggested that antipsychotics may exert their effects by altering histone modifications in dopamine and glutamate-related genes such as DRD3, glutamate receptor 3, and glutamate ionotropic receptor NMDA type subunit 2A[78]. However, haloperidol and risperidone did not alter significant histone modifications or changes in the expression of these genes[77].

Antipsychotics normalize abnormal gene expression levels by acting on ncRNAs in the context of disease pathophysiology[79-81]. Antipsychotics affect neurotransmission and synaptic plasticity by altering miRNAs such as miR-124-3p, miR-520c-3p, miR-132, miR-181b, and miR-30e, which are the most important candidates[81]. Antipsychotics downregulate miR-29b, increasing methyltransferase expression and global DNA methylation levels in patients with schizophrenia[57]. In addition, Swathy et al[82] showed that miR-27a and miR-128a expression was significantly downregulated in antipsychotic drug-treated cells, leading to the upregulation of drug-metabolizing genes in in vitro studies. The upregulation of miR-27a and its effects on these genes have been validated in patients with schizophrenia in the clinical setting.

According to a study, olanzapine significantly reduced miR-195 levels in treatment-responsive patients. These decreases were positively correlated with the total score of the Positive and Negative Syndrome Scale and the rate of decrease in the general psychopathological subscale score. This suggests that miR-195 may predict olanzapine response in patients, and it should be considered a potential therapeutic target for antipsychotics[83]. Olanzapine and haloperidol were shown to reduce the disease-related increase in miR-223 levels in neurons, reducing pressure on glutamate receptors and indirectly contributing to regulating glutamate receptor levels[84]. Haloperidol and clozapine reversed the schizophrenia-like locomotor hyperactivity caused by phencyclidine, blocking the D2 receptors, increasing miRNA-143 levels, and decreasing neuregulin 1 mRNA levels[85]. Yoshino et al[86] observed that haloperidol caused downregulation of miR-137-3p. The downregulation increased Nr3c1 mRNA expression, which is associated with inflammatory responses, cellular proliferation, and differentiation in target tissues.

Huang et al[87] blocked N-methyl D-aspartate receptors by using dizocilpine (MK-801) and caused schizophrenia-like symptoms in rats. RNAs such as miR-184 and miR-138-5p were mainly observed to increase. After administration of clozapine, it was reported that the effect of N-methyl D-aspartate blockade was attenuated, miR-184 levels decreased, and neurotransmitter systems became more balanced. The estrogen signaling pathway was also found to be regulated by clozapine, which may indicate a role for estrogen in the treatment of schizophrenia in women. Another study examined miR-675-3p, which is associated with synaptic function, neuroinflammation, neuronal development, and cell death mechanisms. In schizophrenia, miR-675-3p levels were elevated in treatment-resistant patients and decreased in non-resistant patients. Clozapine further increased the levels in treatment-resistant, similar to in vitro experiments. This finding suggests that miR-675-3p may be an important biomarker in schizophrenia and play a critical role in understanding the mechanism of action of clozapine[88]. As research continues to elucidate the intricate relationships between antipsychotic treatment and epigenetic mechanisms such as DNA methylation, histone modifications, and ncRNA regulation, there is potential for developing innovative therapeutic strategies that leverage these insights to improve the management of schizophrenia.

Antipsychotics act on lncRNA and the immune system in schizophrenia. Sudhalkar et al[89] found that maternally expressed gene 3 (MEG3) lncRNA levels increased, while p53-induced transcript (PINT) and growth arrest-specific transcript 5 (GAS5) levels decreased in patients with psychosis, suggesting a potential link between changes in MEG3, GAS5, and PINT levels and antipsychotic drug efficacy. They also showed that antipsychotic drugs such as risperidone affect the immune system by increasing MEG3 and GAS5 expression while decreasing PINT expression, highlighting the complex interplay among psychosis, antipsychotic drugs, and the immune system. Risperidone may regulate the immune response via heterochromatin-related genes. Furthermore, it has been reported that risperidone regulates dopamine signaling by increasing myocardial infarction associated transcript, a nuclear lncRNA, and olanzapine upregulates dopamine receptors by increasing levels of NONHSAT089447. Atypical antipsychotics generally normalize nuclear paraspeckle assembly transcript 1 levels and can be decisive in treatment response[79]. In addition to lncRNAs, antipsychotic drugs also influence drug metabolism and efficacy through other epigenetic mechanisms, such as DNA methylation and miRNA regulation.

Epigenetic mechanisms and pharmacokinetic genes

Epigenetic changes, such as aberrant DNA methylation and altered miRNA expression, may contribute to individual differences in antipsychotic drug response by altering the expression of various genes[18]. Some of these genes are mentioned in the previous section. This section discusses how epigenetic changes caused by antipsychotic drugs used to treat schizophrenia may alter the expression of pharmacogenetic genes in particular.

As with many other drugs, antipsychotics are biotransformed in the liver and subsequently excreted from the body. In phase I, antipsychotics are mainly converted into less active metabolites by cytochrome P450 (CYP450) enzymes, while in phase II, more polar metabolites are formed through conjugation reactions and eliminated from the body. In phase II, which is the final stage in the biotransformation of antipsychotics, phase II products are transported out of the cell through cell membranes by carrier proteins including ATP-binding cassette sub-family B member 1 (ABCB1)[90]. Genes that encode enzymes and proteins involved in phase I, phase II, and phase III of drug biotransformation are designated pharmacogenetic genes. Alterations in the expression of these genes, attributable to genetic or epigenetic factors, can significantly impact the efficacy of treatment.

Swathy et al[82] showed that antipsychotic medications alone or in combination alter the expression of pharmacogenetic genes such as CYP2D6, CYP3A4, and ABCB1 in different in vitro studies using the HepG2 human liver cell line. A notable finding from the study by Swathy et al[82] was the observation that the periods of peak ABCB1 gene expression were subject to variation following treatment with different pharmaceutical agents (24 hours after treatment for olanzapine and haloperidol, 18-24 hours for clozapine). ABCB1 encodes P-glycoprotein, which regulates the permeability of antipsychotic drugs across the blood-brain barrier, thereby affecting drug efficacy. Another noteworthy finding of the study was the observation that the downregulation of miR-27a resulted in increased expression of the ABCB1 gene following drug treatment, but not through methylation events. These findings were confirmed by analyzing patient samples. CYP450 enzymes, including CYP1A2, CYP2D6, and CYP3A4, play a pivotal role in the biotransformation of clozapine, olanzapine, and haloperidol, which are employed in the treatment of schizophrenia. Swathy et al[82] measured the post-treatment expression levels of the genes encoding these enzymes in their aforementioned study. In addition to miR-27a, which increases the expression of the ABCB1 gene, they determined that miR-28a is also a target miRNA in vitro. However, they were unable to confirm this finding in clinical samples, a probable consequence of CYP450 enzymes being induced/inhibited by co-administered drugs. In their study, Shi et al[91] investigated the relationship between the treatment efficacy of risperidone, another medication used to treat schizophrenia, and methylation in the promoter region of some pharmacogenetic genes (ABCB1, CYP2D6, and CYP3A4) in the Henan and Shanghai populations. They hypothesized that an elevated methylation rate within the promoter region of the CYP3A4 gene might contribute to the diminished efficacy of risperidone treatment.

Tardive dyskinesia is one of the side effects of long-term antipsychotic drugs. Despite the absence of a comprehensive understanding of the underlying mechanisms, one postulation suggests that oxidative stress, induced by neurotoxic free radicals that escalate in conjunction with antipsychotic medication, may underpin the occurrence of dyskinesia[92]. CYP2E1 is involved in oxidative metabolism and generates reactive oxygen species. Its abnormal activation may exacerbate antipsychotic-induced neurotoxicity, contributing to tardive dyskinesia development. In their study, Zhang et al[92] investigated the effect of CYP2E1 in tardive dyskinesia because it is a gene associated with oxidative stress. Their research revealed that promoter DNA methylation at the third CpG site in the CYP2E1 gene was significantly lower in patients with schizophrenia with dyskinesia compared to patients with schizophrenia without dyskinesia and healthy individuals.

Epigenetic interventions in clinical practice

Several studies have shown that epigenetic changes due to the interaction between the genetic background and the environmental risk factors have a crucial role in the development of schizophrenia. Because epigenetic patterns are dynamic and reversible, it is thought that schizophrenia could be treated by restoring them with the use of inhibitors including DNMT inhibitors, HDAC inhibitors and histone methyltransferase inhibitors. DNMT inhibitors such as doxorubicin and zebularine have been shown to regulate the expression of genes associated with schizophrenia, such as reelin and glutamate decarboxylase 1[93]. DNMT inhibitors are not currently a treatment option for schizophrenia because of their low blood-brain barrier permeability[94]. A few studies have shown that histone methyltransferase inhibitors including enhancer of zeste homolog 1, SET domain containing 1A, histone lysine methyltransferase (SETD1A), and SETDB1, may also help treat schizophrenia-like phenotypes such as altered working memory and impaired sociability[95,96]. On the other hand, drugs targeting specific epigenetic modifications, such as HDAC inhibitors, are currently being investigated for their therapeutic potential in cancer and psychiatric disorders[25]. Valproic acid, sodium butyrate, vorinostat, and MS-275 are the inhibitors of class I and II HDACs and their positive effects on schizophrenia-like abnormalities have been demonstrated in animals. In addition, the beneficial effects of valproic acid on schizophrenia treatment were shown in clinical practice, but are still in the trial phase[97].

CONCLUSION

Antipsychotic drug response and adverse effects vary between individuals[10]. Epigenetic changes can contribute to the therapeutic efficacy of antipsychotic medications by altering the gene expression patterns of various genes. This review summarizes how antipsychotic drugs influence drug efficacy and adverse effects by regulating lncRNAs (e.g., MEG3, GAS5, PINT), DNA methylation (e.g., CYP3A4 methylation), and miRNAs (e.g., miR-27a, miR-28a). These studies with small sample sizes or animal models provide valuable insights into the underlying mechanisms. However, their findings cannot be directly extrapolated to human populations. Conclusions from such studies should be considered preliminary and speculative until support is available from larger clinical trials. Furthermore, the variability in study designs, populations, and methodologies can lead to heterogeneity in the studies. A more in-depth analysis of these inconsistencies is essential for more reliable conclusions and guidance for clinical practice. Therefore, the number of more robust, well-controlled studies, and well-powered clinical trials conducted in different populations should be increased to confirm these preliminary findings and assess their generalizability, as in genetic polymorphism studies, which have started to be investigated with great hopes and whose effect has been proven in some medications. In addition, a meta-analysis or systematic review of existing data could also help understand the underlying patterns.

To date, transgenerational epigenetic inheritance has been demonstrated in fly models[98], nematodes[99], and mice[100]. To investigate epigenetic inheritance in humans, however, further studies are needed. If epigenetic marks are shown to be transmitted transgenerationally in humans, they may be associated with a previous generation’s treatment history, which may impact an individual’s response to treatment. In contrast to genetics, the epigenetic pattern is dynamic. It changes according to the course of the disease and the medicines used, and is reversible. By taking advantage of this property, time-dependent epigenetic dynamics and transgenerational inheritance can be used as potential biomarkers[13,14]. On the other hand, there are technical bottlenecks and ethical challenges. First, the multifactorial etiology and variable phenotypic presentation of schizophrenia make it difficult to identify consistent epigenetic biomarkers that predict treatment response[3,4,8]. In addition, the tissue specificity of epigenetic marks, particularly DNA methylation and histone modifications, limits the interpretability of findings[66,78]. Epigenetic modifications are influenced by the pathophysiology of the disease, the duration of treatment, the biochemical status of the patient, genetic factors and environmental factors. Reliance on post-mortem brain samples to assess epigenetic changes has significant limitations[74,77]. Finally, the lack of antipsychotic-naive samples and longitudinal studies makes it difficult to clearly observe changes in the epigenome.

From an ethical point of view, the integration of pharmacoepigenetics into the treatment of schizophrenia raises important concerns. Individuals’ epigenetic profiles not only serve as biological risk indicators but may also contain sensitive information about their social history. This information enables the development of personalized health strategies. However, this process also requires a re-evaluation of fundamental ethical principles such as privacy, discrimination and justice. Epigenetic signatures linked to an individual’s mental health history may be used discriminatorily and inaccurately by employers, insurance companies or legal institutions, which poses ethical risks. Such scenarios not only violate individual privacy but also compromise the principle of social justice. Furthermore, the possibility of epigenetic changes being heritable suggests that these data may have implications not only for the individual but also for future generations[101,102]. In this context, as emphasized by Santaló and Berdasco[101], strict adherence to ethical principles is crucial for the use of epigenetic data in personalized medicine. In disorders such as schizophrenia, how these data are collected, interpreted, and shared is critical to both individual rights and public health.

Future studies should further explore epigenetic biomarkers in antipsychotic drug selection to optimize personalized treatment strategies and reduce the trial-and-error approach in prescribing[13,14]. To integrate personalized medicine into healthcare systems, better strategies must be developed and implemented to educate and train healthcare professionals[103]. However, this field also presents challenges, such as incomplete understanding of the molecular mechanisms of some diseases like schizophrenia, high costs, and complexity.