Genetics of psychotropic medication induced side ...

J Neural Transm DOI 10.1007/s00702-014-1290-3

PSYCHIATRY AND PRECLINICAL PSYCHIATRIC STUDIES - ORIGINAL ARTICLE

Genetics of psychotropic medication induced side effects in two independent samples of bipolar patients Chiara Fabbri • Daniel Souery • Raffaella Calati • Concetta Crisafulli • Armando Chierchia Diego Albani • Gianluigi Forloni • Alberto Chiesa • Rosalba Martines • Othman Sentissi • Julien Mendlewicz • Giovanni De Girolamo • Alessandro Serretti

•

Received: 16 June 2014 / Accepted: 2 August 2014 Ó Springer-Verlag Wien 2014

Abstract The treatment of bipolar disorder (BD) usually requires combination therapies, with the critical issue of the emergence of adverse drug reactions (ADRs) and the possibility of low treatment adherence. Genetic polymorphisms are hypothesized to modulate the pharmacodynamics of psychotropic drugs, representing potential biological markers of ADRs. This study investigated genes involved in the regulation of neuroplasticity (BDNF, ST8SIA2), second messenger cascades (GSK3B, MAPK1, and CREB1), circadian rhythms (RORA), transcription (SP4, ZNF804A), and monoaminergic system (HTR2A and COMT) in the risk of neurological, psychic, autonomic, and other ADRs. Two independent samples of BD patients naturalistically treated were included (COPE-BD n = 147;

Electronic supplementary material The online version of this article (doi:10.1007/s00702-014-1290-3) contains supplementary material, which is available to authorized users. C. Fabbri A. Chiesa R. Martines A. Serretti (&) Department of Biomedical and NeuroMotor Sciences, University of Bologna, Viale Carlo Pepoli 5, 40123 Bologna, Italy e-mail: [email protected] D. Souery Laboratoire de Psychologie Medicale, Universite´ Libre de Bruxelles and Psy Pluriel, Brussels, Belgium R. Calati IRCCS Centro S. Giovanni di Dio, Fatebenefratelli, Brescia, Italy

STEP-BD n = 659). In the COPE-BD 34 SNPs were genotyped, while in the STEP-BD polymorphisms in the selected genes were extracted from the genome-wide dataset. Each ADRs group was categorized as absent-mild or moderate-severe and logistic regression with appropriate covariates was applied to identify possible risk genotypes/ alleles. 58.5 and 93.5 % of patients were treated with mood stabilizers, 44.2 and 50.7 % were treated with antipsychotics, and 69.4 and 46.1 % were treated with antidepressants in the COPE-BD and STEP-BD, respectively. Our findings suggested that ST8SIA2 may be associated with psychic ADRs, as shown in the COPE-BD (rs4777989 p = 0.0017) and STEP-BD (rs56027313, rs13379489 and rs10852173). A cluster of RORA SNPs around rs2083074 showed an effect on psychic ADRs in the STEP-BD. Trends supporting the association between HTR2A and autonomic ADRs were found in both samples.

A. Chiesa Department of Clinical and Experimental Medicine, University of Messina, Messina, Italy O. Sentissi De´partement de Psychiatrie Hoˆpitaux Universitaires de Gene`ve, Faculte´ de Me´decine de Gene`ve, Geneva, Switzerland J. Mendlewicz Universite´ Libre de Bruxelles, Brussels, Belgium G. De Girolamo A. Serretti IRCCS Fatebenefratelli, Brescia, Italy

C. Crisafulli Department of Biomedical Science and Morphological and Functional Images, University of Messina, Messina, Italy A. Chierchia D. Albani G. Forloni Department of Neuroscience, IRCCS-Istituto di Ricerche Farmacologiche ‘‘Mario Negri’’, Milan, Italy

123

C. Fabbri et al.

Confirmations are needed particularly for ST8SIA2 and RORA since the few available data regarding their role in relation to psychotropic ADRs. Keywords Side effect Gene Pharmacogenetics Antipsychotic Antidepressant Mood stabilizer

Introduction Bipolar disorder (BD) is a chronic disease characterized by the alternation of periods of (hypo)mania and depression, resulting in high personal and socio-economic burden in terms of poor quality of life, increased rates of suicide, direct and indirect costs (Whiteford et al. 2013). The treatment of BD usually requires the use of polypharmacotherapy, that raises the critical issue of the emergence of unwanted side effects, and the consequent possibility of low treatment adherence (Baldessarini et al. 2008). Treatment non-adherence occurs at a rate between 12 and 64 % among individuals with the disorder, with consequent increase in the likelihood of relapse and reduction of quality of life (Leclerc et al. 2013). Given that polypharmacotherapy cannot be avoided in a relevant proportion of patients because of insufficient control of symptomatology with monotherapy, biological predictors of adverse drug reactions (ADRs) are pivotal to identify subjects at higher risk for such detrimental events. Genetic polymorphisms are hypothesized to be modulators of the both pharmacokinetics and pharmacodynamics of psychotropic drugs, influencing both efficacy and ADRs. In detail, it is estimated that genetics accounts for 20 to 95 % of variability in drug disposition and pharmacodynamics and about 50 % of ADRs in CNS disorders might be attributed to pharmacogenomic factors (Cacabelos et al. 2012). Adverse drug reactions due to psychotropic drugs that were most studied in terms of their genetic risk profile were antipsychotic-induced weight gain, tardive dyskinesia (Muller et al. 2013) and other extrapyramidal side effects (EPS) (Drago et al. 2011), antidepressant-induced gastrointestinal side effects and sexual dysfunction (Fabbri et al. 2013). On the other hand, genes modulating the risk of sedation and memory deficits have been marginally addressed by previous studies (Saiz et al. 2008), while genes implicated in the neurocognitive response to antipsychotics have received more attention (McClay et al. 2011). We focused on genes involved in the regulation of neuroplasticity (BDNF, ST8SIA2), second messenger cascades (GSK3B, MAPK1, and CREB1), circadian rhythms (RORA), transcription (SP4, ZNF804A), and monoaminergic system (HTR2A and COMT) which can be hypothesized to influence some of the multiple mechanisms

123

involved in the generation of neurological, psychic, autonomic, and other ADRs that are the object of the present study. These side effects are expected to be influenced by a number of genes but this study had a candidate approach. Thus, the selection of candidate genes was based on the previous pharmacogenetic findings but also on the knowledge about the targets of psychotropic drugs, their mechanisms of action, biological abnormalities found in treated subjects and the physiological role of the gene. Brain-derived neurotrophic factor (BDNF) gene was associated with neurocognitive functioning (Simons et al. 2013), including memory and learning (Heldt et al. 2014). In addition, BDNF is known to modulate several of the receptor systems and molecular pathways implicated in the development of EPS (e.g., D1 and DARPP-32, D3 receptors, 17N-methyl D-aspartate (NMDA) receptor subunit expression and GABAergic receptors) (Foltynie et al. 2009). Although it has been less studied in sites different from the brain, BDNF is also widely expressed in nonneuronal tissues including the gastrointestinal tract. Recently, BDNF was demonstrated to regulate the contraction of intestine longitudinal muscle (Al-Qudah et al. 2014). BDNF levels were found to be altered in the CNS of BD subjects (Ray et al. 2014) and they are influenced by lithium and olanzapine in animals (Hammonds and Shim 2009), suggesting that BDNF may be a modulator of both treatment response and ADRs. The sialyltransferase 8B (ST8SIA2) encodes a sialyltransferase that catalyzes the transfer of polysialic acid to neural cell adhesion molecule 1 (NCAM1), influencing neural plasticity, remodeling of neural connections, and neural generation also during adulthood (Sato and Kitajima 2013). The expression of ST8SIA2 is modulated in rats by antipsychotics (Shaw et al. 2014). Recent evidence suggested that ST8SIA2 is also involved in the functional regulation of ion channels and neurologically active molecules, such as BDNF, FGF2, and dopamine (Sato and Kitajima 2013). Thus, the functional implications of ST8SIA2 are still partially understood, but the known influence on neurocognition, dopaminergic system, as well as its modulation by antipsychotics, lead to the hypothesis of a possible impact of this gene on the investigated ADRs. The phosphorylation of glycogen synthase kinase 3 B (GSK3B) occurs in the context of the signaling cascades in response to serotonin, dopamine, lithium, and antidepressants (Beaulieu et al. 2009). It is involved in the control of gene expression and plays a major role in regulation of neuroplasticity and cell survival (Grimes and Jope 2001). As a known target of lithium and antidepressants, the gene may modulate ADRs through pharmacodynamic mechanisms. Mitogen-activated protein kinase 1 (MAPK1) signaling represents the major convergence point in all signal

Genetics of side effects in bipolar disorder

pathways that regulate cellular growth and differentiation, including synaptic plasticity and long-term potentiation through the involvement of CREB-BDNF activation (Racaniello et al. 2010). Given this key role, MAPK1 signaling has been implicated in the pharmacodynamics of antidepressants, mood stabilizers, and antipsychotics (Calati et al. 2013). Further, MAPK1 pathway is critical for maintaining the functionality of Leydig cells and thus testosterone and LH levels in males (Yamashita et al. 2011), as well as the activation of the pathway in lumbar spinothalamic cells is required for ejaculation (Staudt et al. 2010). CAMP responsive element binding protein 1 (CREB1) encodes a transcription factor that is phosphorylated by several protein kinases among which MAPK1. CREB1, and MAPK1 show a highly interrelated function and represent critical points of convergence in signaling pathways regulating neuronal plasticity and involving also BDNF (Calati et al. 2013). CREB1 seems to be involved in antidepressant, mood stabilizer, and antipsychotic mechanisms of action according to in vitro and in vivo studies (Calati et al. 2013). Consistently, several components of the MAPK1 pathway (CREB1 included) showed altered expression in BD (Yuan et al. 2010). In periphery, CREB1 is involved in adult enteric nervous system maintenance (Liu et al. 2009) and sexual steroid synthesis (Stojkov et al. 2013). RAR-related orphan receptor A (RORA) acts as transcription factor that is implicated in the regulation of circadian rhythms and shows neuro-protective and antiinflammatory properties (Journiac et al. 2009). RORA genotype may impact on circadian changes during lithium treatment (McCarthy et al. 2013) and variation in gene expression after exposure to antidepressants was reported (Lisowski et al. 2013). The mammalian circadian clock plays a pivotal role not only in the regulation of the sleepwave cycle, but in the general regulation of rhythmic physiology and behavior, such as locomotor activity and endocrine function (e.g., diurnal cycles of adrenocorticotropic hormone and corticosterone secretion, inflammatory cytokine production (Frederic et al. 2006)). Further, dopamine D3 receptor expression in the striatum was demonstrated to have a circadian oscillation that is modulated by RORA (Ikeda et al. 2013). Thus variants in this gene may influence the risk of ADRs such as those affecting memory and attention (van der Heijden et al. 2013), arousal and activity level, and potentially EPS. SP4—that is expressed in brain, epithelial tissues, and testis—regulates the expression of genes implicated in a variety of biological processes including neuronal development and function (Black et al. 2001; Bouwman and Philipsen 2002). Mice with reduced expression of SP4 exhibit behavioral abnormalities in both sensorimotor gating and contextual memory, suggesting that the Sp4

pathway is essential for hippocampal integrity (Zhou et al. 2005). SP4 null mice exhibit retarded growth; males do not breed despite complete spermatogenesis, while females have a smaller thymus, spleen, and uterus (Gollner et al. 2001). Altered SP4 protein level in the cerebellum and prefrontal cortex in BD subject was demonstrated, and lithium could control SP4 expression (Pinacho et al. 2011). ZNF804A (zinc finger protein 804A) encodes for another transcription factor that has been mainly studied in regard to the risk of schizophrenia and BD, since its impact on brain architecture and cognition (Hess and Glatt 2014). In detail, polymorphisms in this gene were associated with working memory, (Esslinger et al. 2011), speed of processing (Van Den Bossche et al. 2012), and attention (Balog et al. 2011). COMT and DRD2—that were associated with haloperidol-induced EPS (Zivkovic et al. 2013)—are among the genes which expression is regulated by ZNF804A (Girgenti et al. 2012). HTR2A (5-hydroxytryptamine (serotonin) receptor 2A, G protein-coupled) is the main excitatory serotonergic receptor and was implicated in ADRs risk in patients treated with selective and nonselective serotonin reuptake inhibitors (Staeker et al. 2014) and in SSRI-induced sexual dysfunction (Garfield et al. 2013). HTR2A is also expressed in the basal ganglia, a brain region that plays a critical role in antipsychotic-induced movement disorders. Consistently, previous pharmacogenetic studies suggested that HTR2A variants may affect the risk of antipsychotic EPS (Basile et al. 2001; Segman et al. 2001; Gunes et al. 2007; Wilffert et al. 2009). COMT is an enzyme that inactivates dopamine and noradrenaline through their methylation. The Val158Met polymorphism (rs4680 or 472G/A) has been particularly studied since its functional relevance, indeed the Met allele entails the production of a three- to fourfold less active enzyme than the Val variant (Lotta et al. 1995). COMT participates in the metabolic inactivation of dopamine, thus genetic variability affecting this gene and dopamine availability has been studied for putative involvement in the risk of EPS. Several pharmacogenetic studies consistently suggested that this polymorphism may be a modulator of antipsychotic-induced EPS (Srivastava et al. 2006; Lafuente et al. 2008; Zai et al. 2010; Zivkovic et al. 2013). On the other hand, the most frequent non-dopaminergic side effect of COMT inhibition is diarrhea, followed by nausea, dizziness, headaches, and abdominal pain due to an abnormal prolonged action of catecholamines (Hauser et al. 1998). Further, rs4680 interacts with other dopaminergic genes and impacts on prefrontal cognition by affecting the stability of neural networks in the prefrontal cortex, with an effect on working memory and cognitive flexibility (Van Rheenen and Rossell 2013).

123

C. Fabbri et al.

Samples

with an age C15 years were entered into a study registry. All patients received a systematic assessment battery at study entry and were treated in a naturalist setting according to current guidelines. At each follow-up visit, the treating psychiatrist completed a standardized assessment and assigned an operationalized clinical status based on Diagnostic and Statistical Manual of Mental Disorders, 4th ed. (DSM-IV) criteria. 659 patients were included and the 98.33 % of them was of Caucasian ethnicity (Supplementary Table 1). For the purposes of the present study, we focused on the naturalistic part of the study and we included only BD patients for which reports of side effects were available.

COPE-BD sample

Side effects assessment

Patients were recruited in the ‘Psy Pluriel’ center, Centre Europe´en de Psychologie Me´dicale and the Department of Psychiatry of Erasme Hospital in Brussels. A detailed description of the sample has been reported elsewhere (Souery et al. 2011, 2012). In brief, the COPE-BD (Clinical Outcome Measures for Bipolar Disorder) project enrolled patients that met DSM-IV criteria for a diagnosis of BD type I or II or major depressive disorder. Only patients that met diagnostic criteria for BD were included in the present study. A structured examination tool was used to assess essential elements of disease, such as socio-demographic characteristics, psychiatric antecedents, diagnosis, current and previous treatments, quality of life and functioning. Lifetime and current diagnosis, course of illness and comorbidities were assessed through the Mini-International Neuropsychiatric Interview (MINI) (Sheehan et al. 1998). 147 patients were included and 144 (97.96 %) of them showed Caucasian ethnicity (Supplementary Table 1).

In the STEP-BD study, the presence and severity of side effects were rated on a Clinical Monitoring Form (CMF) for a series of side effects (Sachs et al. 2002), including constipation, diarrhea, dry mouth, extrapyramidal signs, tremors, headache, sedation, memory reduction, weight gain, and sexual dysfunction. Side effects were rated on a five-point scale ranging from 0 (no side effect) to 4 (extremely severe side effect). As the STEP-BD had a longitudinal course, to allow for a higher comparability with the COPE-BD study (see below) and to maximize the sensitivity to the detection of side effects in this sample, we considered the visit in which subjects showed the highest number of side effects. For subjects who did not show any side effect over the study period, we chose the visit that was closer to the average time of the overall follow-up period, as previously reported (Serretti et al. 2013). In the COPE-BD study, the presence and severity of side effects were rated using the Udvalg for Kliniske Undersøgelser (UKU) Side Effect Rating Scale (Lingjaerde et al. 1987). The UKU assesses a total of 48 symptoms arranged into four groups: psychic, neurologic, autonomic, and other side effects. Each symptom is scored from 0 (no side effect) to 3 (side effect that markedly interferes with patient’s performance). As this study was cross-sectional, side effects were recorded at a single time point. To match the presence of side effects in the STEP-BD and in the COPE-BD samples as much as possible, we dichotomized the presence of side effects in both studies. Furthermore, to match the structure of the UKU employed in the COPE-BD study, we categorized the side effects available for the STEP-BD sample into four different classes including psychic (sedation and memory reduction), neurologic (tremors and extrapyramidal effects), autonomic (dry mouth, constipation, and diarrhea) and other side effects (headache, weight gain, and sexual dysfunction). First of all, in accordance with a number of previous studies employing the UKU scale as a measure of side

Almost all the described genes have not been previously investigated as modulators of ADRs induced by psychotropic drugs, with the exception of HTR2A, COMT, and ZNF804A. Given the biological basis supporting a putative involvement of the selected genes on the risk of ADRs, the present study aimed to investigate them in terms of association with psychic, autonomic, neurological and other ADRs in two independent samples of BD subjects.

Materials and Methods

STEP-BD sample Clinical data were retrieved from the National Institute of Mental Health (NIMH) funded STEP-BD project (https:// www.nimhgenetics.org/available_data/bipolar_disorder/). The STEP-BD (Systematic Treatment Enhancement Program for Bipolar Disorder) was a long-term outpatient study designed to investigate which treatments, or combinations of treatments, are most effective for treating episodes of depression and mania and for preventing recurrent episodes in BD. The study had very broad inclusion criteria. Indeed, both bipolar patients and patients with the bipolar subtype of schizoaffective disorder in every phase of their disease were enrolled. A detailed description of inclusion/exclusion criteria can be found elsewhere (Sachs et al. 2003). Briefly, the STEP-BD used a hybrid design to collect longitudinal data as patients made transitions between naturalistic studies and RCTs. Eligible subjects

123


effects related to psychotropic drugs [e.g., (de Leon et al. 2005; Demyttenaere et al. 2005)], we dichotomized patients experiencing no or mild side effects that did not interfere with patient’s function as not having side effects, and patients experiencing moderate to severe side effects that had a significant impact on patient’s functioning as having side effects. Similarly, in the STEP-BD sample, we dichotomized patients experiencing none or mild side effects that did not interfere with patient’s functioning (score 0 or 1) as not having side effects and patients experiencing moderate to severe side effects (score from 2 to 4) as having side effects. Secondly, as an example, a patient was rated as experiencing psychic side effects if he/ she reported at least one psychic side effect. Moreover, a patient was rated as experiencing at least one side effect if he/she experienced at least one psychic, neurologic, autonomic, or other side effects. Phenotypes under investigation In both samples, the following dichotomous (mild-absent compared to moderate-severe) categories of side effects were investigated: psychic, neurological, autonomic, other and total side effects. Individual side effects included in each of these categories were determined according to the UKU scale for the COPE-BD and CMF form in the STEPBD (see paragraph 2.2 Side effects assessment). Genotyping Regarding the COPE-BD sample, genomic DNA was purified from peripheral blood samples with an automated workstation (Maxwell, Promega) and checked for quality and quantity by a small scale spectrophotometer (Nanodrop, Thermo Scientific). The genotyping was performed using a Sequenom MassArray platform (Sequenom, California, USA) in conjunction with the iPLEX assay (http:// www.sequenom.com). Genotyping was then performed according to the manufacturer’s standard protocols. MassArrayTyper V.4.0 3.4 was used to read the extended mass and genotype calls. For some SNPs, genotyping was performed by PCR followed by restriction enzyme analysis. Forward and reverse primers’ sequences are available upon request. SNPs were chosen among those (1) with a reported prevalence of at least 5 % for the variant allele among Caucasians (data from http://hapmap.ncbi.nlm.nih.gov/, R2 = 0.8 and MAF = 0.05), and (2) with availability of a validated assay in our laboratory. We also considered variants not investigated before. The list of genotyped SNPs is shown in Supplementary Table 2. The STEP-BD genome-wide dataset was available through the NIMH genetics initiative (https://www.nimh genetics.org/available_data/bipolar_disorder/).

Statistical analysis In the STEP-BD genome-wide (GWAS) the effect of individual markers (alleles and genotypes) on phenotypes (see paragraph 2.3. Phenotypes under investigation) was tested through logistic regression models. Odds ratios (OR) with 95 % confidence intervals (CI) were estimated for the effects of high-risk genotypes/alleles. Covariates were selected according to their impact on outcomes, while gender, age, and ancestry were used as covariates in all the analyses. A complete agglomerative clustering was applied, based on a multidimensional scaling of a matrix of pairwise identity-by-state (IBS) values between samples, and ancestry clusters were defined according to the pairwise population concordance test (PCC \0.0001 (Purcell et al. 2007)). Identity-by-descent (IBD) analysis was used to identify related subjects (IBD [0.1875 (Anderson et al. 2010)). Available SNPs were extracted according to their physical positions (Genome Build 36.3), including SNPs up to 100.000 bp near gene boundaries. The plink software was used for the analysis (Purcell et al. 2007). Secondly, genes harboring SNPs with suggestive associations with phenotypes (p B 0.01) were imputed using IMPUTE2 (http://mathgen.stats.ox.ac.uk/impute/impute_v2.html) and 1,000 Genomes data (NCBI Build 36 (dbSNP b126)) as reference panel. An info value threshold C0.8 was applied to prune poorly imputed SNPs. Imputed genes were analyzed through the Snptest software (https://mathgen.stats. ox.ac.uk/genetics_software/snptest/snptest.html) according to the same principles reported above. Alpha value was set to 0.0006 since 88 independent SNPs were retrieved in the STEP-BD GWAS (SNPs were considered in linkage disequilibrium [LD] and thus not independent when R2 [ 0.8). The STEP-BD sample provided a power of 0.80 to detect risk alleles with OR C 2.9 setting alpha value to 0.0006 two-tailed. In the COPE-BD sample the effect of genotypes and alleles was investigated through logistic regression models. Age and gender were included as covariates in all analyses in addition to variables that showed evidence of impact on phenotypes. Alpha value was set to 0.002 since 29 independent polymorphisms were genotyped (linkage disequilibrium plots are shown in Supplementary Figure. 1). The COPE-BD sample provided a power of 0.80 to detect risk alleles with OR C 4.2 setting alpha value to 0.002 twotailed. We, therefore, conservatively corrected for multiple testing in each sample separately.

Results The clinical-demographic features of the analyzed samples are shown in Supplementary Table 1, while the clinical

123

C. Fabbri et al.

characteristics of samples split by presence/absence of side effects of each category are reported in Supplementary Table 3. In the COPE-BD sample, age, gender, and polytherapy versus monotherapy (impact on total side effect risk: Chi2 = 9.97, df = 1, p = 0.0016) were used as covariates. The phase of disease did not show evidence of impact on side effect level (p = 0.11). In the STEP-BD sample 659 subjects fulfilled inclusion criteria, 116 SNPs were available in the genes of interest (Supplementary Table 2). Unfortunately, no SNP was available within the SP4 gene, while for CREB1 only 7 SNPs near the gene (from 109,018 to 44,577 bp) were available but no one was an intragenic SNP. Polytherapy versus monotherapy had a clear impact on the risk of side effects (for total side effects: T = -5.71, p = 2.36e-08), as well as current disease phase (euthymia vs. acute phase: F = 7.27, p = 0.0008), thus both were used as covariates in all regression models in addition to age, gender, and ancestry. No related subjects were retrieved in the STEPBD, according to IBD. In the COPE-BD sample, BDNF SNPs showed a selective trend of association with autonomic side effects (rs11030101 TT vs. AA p = 0.025 and T vs. A p = 0.01; rs11030104 G vs. A p = 0.035; rs12273363 T vs. C p = 0.033), while rs11030101 showed also a trend of association with all (genotypic p = 0.009) and neurological ADRs (genotypic p = 0.02). HTR2A rs6313 T vs. C allele had a trend of association with other ADRs (allelic p = 0.040), as well as RORA rs809736 (allelic p = 0.019) and CREB1 rs889895 (allelic p = 0.038). rs809736 may be also related to the risk of neurological side effects (GG vs. AA p = 0.028 and G vs. A p = 0.022). SP4 rs10272006 and rs2282888 showed an interesting trend of association with psychic side effects (allelic p = 0.004 and 0.008, respectively). ST8SIA2 rs4777989 represented the only SNPs that reached the significance threshold for association with psychic ADRs (AG vs. AA p = 0.00167), while other two SNPs showed modest trends of association with the same phenotype (ZNF804A rs7603001 G vs. A p = 0.029 and MAPK1 rs6928 GC vs. CC p = 0.019). Finally, MAPK1 rs8136867 had a trend of impact on the risk of neurological side effects. Results with p \ 0.05 are summarized in Table 1, while an overview of all results in the COPE-BD is provided in Supplementary Table 4. In the STEP-BD sample, no SNP reached the significance threshold (Supplementary Table 5). rs8025225 in the ST8SIA2 gene (p = 0.0019) and rs1923882 in HTR2A gene (p = 0.002) showed a trend of association with autonomic side effects. SNPs available after imputation of the genes of interest are shown in Supplementary Table 6. For ZNF804A and CREB1 imputation was not possible because of the

123

insufficient number of the available SNPs in the original dataset. A group of RORA polymorphisms in the first intron was associated with psychic side effects (especially rs8037669 and rs2083074, p = 0.00018, see Table 2; Fig. 1), while other SNPs showed a trend of association with autonomic side effects. A cluster of SNPs within the HTR2A gene showed suggestive p values for association with autonomic side effects (p \ 0.005, see Table 2), as suggested by the analysis of the genotyped SNPs. A group of SNPs in ST8SIA2 showed a trend of association with psychic side effects (Table 2), possibly confirming the top result obtained for rs4777989 in the COPE-BD sample.

Discussion The present study investigated the putative impact of genes involved in the regulation of neuroplasticity (BDNF and ST8SIA2), second messenger cascades (GSK3B, MAPK1, and CREB1), circadian rhythms (RORA), transcription (SP4 and ZNF804A), and monoaminergic system (HTR2A and COMT) on the risk of neurological, psychic, autonomic, and other ADRs during treatment with psychotropic drugs. Our results suggested that the risk of psychic side effects (sedation and memory deficits) may be influenced by ST8SIA2, as outlined by interesting trends in both the COPE-BD and STEP-BD (with COPE-BD rs4777989 located *80 Kbp from findings in the STEP-BD), and by RORA, as emerged from the analysis of imputed SNPs in the STEP-BD. rs4777989 is a ST8SIA2 intronic SNP that is located in the promoter of C15orf32 (chromosome 15 open reading frame 32), while the top ST8SIA2 SNPs in the STEP-BD are unfortunately still uncharacterized. RORA top SNPs in the STEP-BD were all located in an intronic region that spans about 10 Kbp. No previous study investigated the role of these genes on the risk of psychic ADRs, while evidence suggested that both of them are regulators of neural plasticity and remodeling of neural connections, with a consistent impact on memory and attention (Sato and Kitajima 2013; van der Heijden et al. 2013). Further, RORA has a key role in the regulation of circadian rhythms and genetic variants were associated with sleep disturbances (Utge et al. 2010), with a predictable impact on diurnal levels of vigilance and activity. The modulation of corticosterone circadian production exerted by RORA can also be implicated in the risk of memory deficits, since prolonged disturbances in hypothalamo– pituitary–adrenal function can negatively affect hippocampal long-term potentiation (Joels et al. 2013). A marginal effect or no effect on the risk of psychic ADRs can be hypothesized for the SP4 gene, according to the present findings. Indeed, the intronic SNP rs10272006 within this gene showed an interesting trend of association in the

RORArs809736

HTR2Ars6313

BDNFrs12273363

BDNFrs11030104

TA vs. AA: E = -1.18, SE = 0.45, z = -2.60, p = 0.009* (OR = 0.31; 95 % CI = 0.12-0.73)

BDNFrs11030101

TT vs. AA: p = 0.28

TT vs. AA: E = -1.25, SE = 0.56, z = -2.24, p = 0.025* (OR = 0.29, 95 % CI = 0.09–0.82)

GA vs. AA: p = 0.62 GG vs. AA: E = 1.91, SE = 0.76, z = 2.52, p = 0.012* (OR = 6.79, 95 % CI = 1.64–35.43) G vs. A: E = 0.71, SE = 0.30, z = 2.34, p = 0.019* (OR = 2.04, 95 % CI = 1.12–3.70)

GA vs. AA: p = 0.41 GG vs. AA: E = 2.09, SE = 0.95, z = 2.21, p = 0.028* (OR = 1.78, 95 % CI = 4.16–7.05) G vs. A: E = 1.06, SE = 0.46, z = 2.29, p = 0.022* (OR = 2.90. 95 % CI = 1.14–7.19)

GA vs. AA: p = 0.74 GG vs. AA: p = 0.90 G vs. A: p = 0.91

GA vs. AA: p = 0.49 GG vs. AA: p = 0.24 G vs. A: p = 0.19

GA vs. AA: p = 0.34

GG vs. AA: p = 0.36

G vs. A: p = 0.94

TT vs. CC: p = 0.05 T vs. C: E = -0.55, SE = 0.27, z = -2.05, p = 0.040* (OR = 0.58, 95 % CI = 0.34–0.97)

T vs. C: p = 0.64

T vs. C: p = 0.47

T vs. C: p = 0.39

TT vs. CC: p = 0.59

TC vs. CC: p = 0.36

T vs. C: p = 0.73

TT vs. CC: p = 0.83

TC vs. CC: p = 0.93

T vs. C: p = 0.22

TT vs. CC: p = 0.48

TC vs. CC: p = 0.14

T vs. C: E = -0.75, SE = 0.35, z = -2.14, p = 0.033* (OR = 0.47, 95 % CI = 0.24–0.95)

T vs. C: p = 0.45

T vs. C: p = 0.99

TC vs. CC: p = 0.89

TT vs. CC: p = 0.12

TT vs. CC: p = 0.33

TT vs. CC: p = 0.94

TT vs. CC: p = 0.42

T vs. C: p = 0.46

TC vs. CC: p = 0.40

TC vs. CC: p = 0.38

TC vs. CC: p = 0.92

G vs. A: p = 0.24

GG vs. AA: p = 0.11

TC vs. CC: p = 0.45

TC vs. CC: p = 0.052 TT vs. CC: p = 0.092

G vs. A: E = 0.64, SE = 0.30, z = 2.11, p = 0.035* (OR = 1.90, 95 % CI = 1.04–3.43)

G vs. A: p = 0.81

G vs. A: p = 0.10

AG vs. AA: p = 0.80

TT vs. CC: p = 0.24

G vs. A: p = 0.34

GG vs. AA: p = 0.066

GG vs. AA: p = 0.52

T vs. A: p = 0.27

TT vs. AA: p = 0.37

TA vs. AA: p = 0.075

Other side effects

TC vs. CC: p = 0.64

AG vs. AA: p = 0.70 GG vs. AA: p = 0.99

AG vs. AA: p = 0.15

AG vs. AA: p = 0.37

T vs. A: E = -0.71, SE = 0.28, z = -2.56, p = 0.01* (OR = 0.49, 95 % CI = 0.29–0.84)

TA vs. AA: E = -1.59, SE = 0.69, z = -2.29, p = 0.02* (OR = 0.20, 95 % CI = 0.05–0.75)

TA vs. AA: E = -1.08, SE = 0.43, z = -2.53, p = 0.012* (OR = 0.34, 95 % CI = 0.14–0.78) T vs. A: p = 0.13

Neurological side effects

Autonomic side effects

GG vs. AA: p = 0.13

T vs. A: p = 0.39

TT vs. AA: p = 0.43

TA vs. AA: p = 0.37

Psychic side effects

AG vs. AA: p = 0.33

T vs. A: p = 0.068

TT vs. AA: p = 0.10

All side effects

Gene-SNP

Table 1 SNPs with p \ 0.05 in the COPE-BD sample were reported


123

123

ST8SIA2rs11632521

SP4rs12673091

SP4rs10272006

GG vs. AA: p = 0.49 GA vs. AA: p = 0.17 G vs. A: p = 0.44

GA vs. AA: p = 0.39

G vs. A: p = 0.98

G vs. A: p = 0.16

GA vs. AA: p = 0.23

GG vs. AA: p = 0.18

G vs. A: p = 0.16

G vs. A: p = 0.15

G vs. A: p = 0.60

GG vs. AA: p = 0.92

GG vs. AA: p = 0.05

GG vs. AA: p = 0.14

GG vs. AA: p = 0.63

G vs. A: E = -0.84, SE = 0.29, z = -2.88, p = 0.004* (OR = 0.43, 95 % CI = 0.24–0.75)

GG vs. AA: E = -1.42, SE = 0.61, z = -2.32, p = 0.020* (OR = 0.24, 95 % CI = 0.06–0.74)

AG vs. AA: p = 0.58

G vs. A: p = 0.40

G vs. A: p = 0.91

AG vs. AA: p = 0.66

GG vs. AA: p = 0.36

GG vs. AA: p = 0.74

AG vs. AA: E = -1.03, SE = 0.43, z = -2.40, p = 0.016* (OR = 0.36, 95 % CI = 0.15–0.82)

G vs. A: E = -0.75, SE = 0.28, z = -2.64, p = 0.008* (OR = 0.47, 95 % CI = 0.27–0.82)

AG vs. AA: p = 0.81

AG vs. AA: p = 0.94

AG vs. AA: p = 0.41

G vs. A: p = 0.16

G vs. A: p = 0.68 GG vs. AA: E = -1.25, SE = 0.58, z = -2.18, p = 0.029* (OR = 0.29, 95 % CI = 0.08–0.83)

GG vs. AA: p = 0.20

GG vs. AA: p = 0.71

GA vs. AA: E = 2.64, SE = 1.17, z = 2.25, p = 0.025* (OR = 14.01, 95 % CI = 2.04–299.11) G vs. A: p = 0.20

GG vs. AA: p = 0.25

G vs. A: E = 1.05, SE = 0.42, z = 2.49, p = 0.013* (OR = 2.85, 95 % CI = 1.25–6.53)

GG vs. AA: p = 0.05

AG vs. AA: E = 1.64, SE = 0.73, z = 2.25, p = 0.02* (OR = 5.16, 95 % CI = 1.35–25.6)

G vs. A: p = 0.90

GG vs. AA: p = 0.93

AG vs. AA: p = 0.50

G vs. A: p = 0.52

GG vs. AA: p = 0.56

AG vs. AA: p = 0.74

AG vs. AA: E = -0.93, SE = 0.43, z = -2.17, p = 0.03*

AG vs. AA: p = 0.39

AG vs. AA: p = 0.78

SP4rs2282888




All side effects

Gene-SNP

Table 1 continued

G vs. A: p = 0.60

GA vs. AA: p = 0.66

GG vs. AA: p = 0.58

G vs. A: p = 0.91

GG vs. AA: p = 0.79

AG vs. AA: p = 0.84

G vs. A: p = 0.47

GG vs. AA: p = 0.39

AG vs. AA: p = 0.84

G vs. A: p = 0.54

GG vs. AA: p = 0.46

AG vs. AA: p = 0.79

Other side effects

C. Fabbri et al.

MAPK1rs6928

CREB1rs2254137

CREB1rs889895

ZNF804Ars7603001

AG vs. AA: p = 0.21

ST8SIA2rs4777989

G vs. A: p = 0.78

G vs. A: p = 0.71

G vs. A: p = 0.26

G vs. C: p = 0.61

G vs. C: p = 0.31

GG vs. CC: p = 0.36

GC vs. CC: E = 1.11, SE = 0.47, z = 2.36, p = 0.019* (OR = 3.03, 95 %CI = 1.24–7.97)

C vs. A: p = 0.62

C vs. A: p = 0.61

GC vs. CC: p = 0.08 GG vs. CC: p = 0.71

CA vs. AA: p = 0.33 CC vs. AA: p = 0.69

CA vs. AA: p = 0.76

CA vs. AA: p = 0.18

GC vs. CC: p = 0.63 GG vs. CC: p = 0.19 G vs. C: p = 0.19 G vs. C: p = 0.22 G vs. C: p = 0.35

C vs. A: E = 0.53, SE = 0.26, z = 2.00, p = 0.046* (OR = 1.69, 95 % CI = 1.01–2.85)

CC vs. AA: E = 1.11, SE = 0.55, z = 2.01, p = 0.045* (OR = 3.03, 95 % CI = 1.04–9.28)

GC vs. CC: p = 0.068 GG vs. CC: p = 0.17

C vs. A: p = 0.63

CC vs. AA: p = 0.56

GC vs. CC: p = 0.86 GG vs. CC: p = 0.34

C vs. A: p = 0.96

CC vs. AA: p = 0.98

CA vs. AA: p = 0.80

G vs. A: E = -0.68, SE = 0.33, z = -2.08, p = 0.038* (OR = 0.51, 95 % CI = 0.26–0.95) G vs. A: p = 0.76

G vs. A: p = 0.40

G vs. A: p = 0.20

G vs. A: p = 0.92

CC vs. AA: p = 0.61

GG vs. AA: p = 0.099

GG vs. AA: p = 0.75

GG vs. AA: p = 0.58

GG vs. AA: p = 0.08

GG vs. AA: p = 0.44

CA vs. AA: p = 0.70

AG vs. AA: p = 0.30

AG vs. AA: p = 0.95

AG vs. AA: p = 0.49

AG vs. AA: p = 0.44

AG vs. AA: p = 0.14

G vs. A: E = 0.63, SE = 0.29, z = 2.19, p = 0.029* (OR = 1.87, 95 % CI = 1.07–3.30)

GG vs. AA: p = 0.76

GG vs. AA: p = 0.71

GG vs. AA: p = 0.32

G vs. A: p = 0.34

AG vs. AA: p = 0.59

AG vs. AA: p = 0.49

AG vs. AA: p = 0.34

GG vs. AA: E = 1.35, SE = 0.61, z = 2.20, p = 0.028* (OR = 3.87, 95 % CI = 1.20–13.72)

G vs. A: p = 0.16

G vs. A: p = 0.76

G vs. A: p = 0.41

AG vs. AA: p = 0.09

GG vs. AA: p = 0.14

GG vs. AA: p = 0.97

GG vs. AA: p = 0.31

GG vs. AA: p = 0.33

G vs. A: p = 0.49

GG vs. AA: p = 0.96

AG vs. AA: p = 0.79

AG vs. AA: p = 0.40

AG vs. AA: p = 0.78

AG vs. AA: E = -1.46, SE = 0.46, z = -3.14, p = 0.00167** (OR = 0.23, 95 % CI = 0.09–0.57)

Other side effects




AG vs. AA: p = 0.33

G vs. A: p = 0.53

GG vs. AA: p = 0.32

All side effects

Gene-SNP

Table 1 continued


123

Results with p \ 0.05 and p [ 0.002. ** results with p \ 0.002. For a summary of all results in the COPE-BD sample see Supplementary Table 4

HWE Hardy–Weinberg equilibrium

*

GG vs. AA: E = -1.96, SE = 0.96, z = -2.05, p = 0.04* (OR = 0.14, 95 % CI = 0.02–0.94) G vs. A: p = 0.13

For each SNP genotypic and allelic analyses were performed. For genotypic analysis, the homozygote of reference was compared to the heterozygote and other homozygote

G vs. A: p = 0.84 G vs. A: p = 0.46 G vs. A: p = 0.57

123

G vs. A: p = 0.22

GG vs. AA: p = 0.39 GG vs. AA: p = 0.81 GG vs. AA: p = 0.84

GG vs. AA: p = 0.90

GA vs. AA: p = 0.14

GA vs. AA: E = -2.45, SE = 1.14, z = -2.14, p = 0.03* (OR = 0.09, 95 % CI = 0.007–0.76) GA vs. AA: p = 0.18 GA vs. AA: p = 0.75 GA vs. AA: p = 0.85

MAPK1rs8136867

Other side effects Neurological side effects Autonomic side effects Psychic side effects All side effects Gene-SNP

Table 1 continued

C. Fabbri et al.

COPE-BD, but no signal was retrieved in the STEP-BD. Anyway, the possibility of a genuine effect of rs10272006 on the risk of psychic ADRs is encouraged by the observed abnormalities in contextual memory in mice with reduced expression of SP4 (Zhou et al. 2005). Some interesting trends of association were found among two intronic SNPs in the HTR2A and ST8SIA2 genes and the risk of autonomic side effects in the STEPBD sample. After imputation, several HTR2A SNPs showed trends of association (Table 2) and some of them (rs732821, rs731245) are located in the promoter of the gene. Serotonin is an important signaling molecule in the gut targeting enterocytes, smooth muscles, and enteric neurons. Serotonin initiates responses such as nausea, vomiting, intestinal secretion, and peristalsis; contractions of gut smooth muscle cells can be evocated by stimulating 5-HT2 receptors on muscle (Sikander et al. 2009). Thus, the association between HTR2A variants and gastrointestinal ADRs seems well supported by the role of serotonin on smooth muscle activity and intestinal secretion. ST8SIA2 involvement in the risk of autonomic ADRs is more difficult to explain, indeed the gene is widely expressed during embryonic development, while expression in the adult appears confined to limited brain regions of persistent neural plasticity (Angata et al. 1997). However, ST8SIA2 is a pivotal regulator of cell–cell interactions by the modulation of intercellular spaces and the docking between ligands and receptors on cell surfaces, besides being involved in Ca2? transporters regulation (Sato and Kitajima 2013). Thus, we speculate that ST8SIA2 polymorphisms may entail different patterns of polysialic acid distribution resulting in different susceptibility of the autonomic system to pharmacological stimulation/inhibition of target receptors. Findings regarding neurological ADRs—that are doubtless among the most clinically relevant—did not suggest any major involvement of the investigated genes. Some trends of association were found for BDNF rs11030101, RORA rs809736, and MAPK1 rs8136867 in the COPE-BD sample that anyway did not show any replication in the STEP-BD sample. Despite all these genes are hypothesized to modulate pathways involved in the risk of EPS (Foltynie et al. 2009; Crisafulli et al. 2013; Ikeda et al. 2013), the present results do not support their involvement. Other ADRs—including headache, weight gain, and sexual dysfunction—did not demonstrate evidence of correlation with the genes investigated in the present study. Trends of association were retrieved for BDNF rs11030101, HTR2A rs6313, RORA rs809736, and CREB1 rs889895 in the COPE-BD, but no confirmation was found in the STEP-BD. HTR2A rs6313 CC genotype may be correlated to the risk of ADRs according to

Recessive

Additive

0.8030 0.8800 0.8957 0.9177

rs11635008

rs11631786

rs1002147

rs12898479

0.8514 0.8995

rs341407

0.8743 0.8815

rs35464030

rs12900971

rs8041438

Additive

0.8903

rs2899667

Additive

Dominant

Additive

Additive

Additive

Additive

Additive

Additive

Additive

1.0000 0.8874

rs2278507

Recessive

Recessive Recessive

rs4775376

0.8520 0.9651

rs873961

0.9867

rs57567672

rs59938185

Recessive

0.9757 1.000

rs7173460 rs11632352

Recessive Recessive

0.9738 0.9752

rs60601176

Recessive

Recessive

Recessive

Recessive

Recessive

Recessive

Additive

Additive

Additive

Additive

Additive

Dominant

Recessive

Model

rs7173279

0.9573 0.9778

rs2280594

rs6494246

0.9760 0.9766

rs8029330

rs8029226

0.9047

0.9757

rs7173460 0.9316

0.9738

rs60601176

rs8037669

0.9867

rs57567672

rs2083074

0.9779 0.9752

0.8830

rs752915

rs6494246

0.9655 0.9649

rs708682 rs782920

rs7173279

0.9534 0.9942

rs2438127

RORA

Info

rs782929

SNP

Gene



All side effects

Phenotype

Cases MAF = 0.1609; controls MAF = 0.1781; OR = 1.13, CI = 0.84–1.52, p = 0.0043












Cases MAF = 0.3406; controls MAF = 0.4052; OR = 0.76, CI = 0.60–0.95, p = 0.00036*


Cases MAF = 0.2038; controls MAF = 0.2796; OR = 0.66, CI = 0.51–0.85, p = 0.00021* Cases MAF = 0.3388; controls MAF = 0.4029; OR = 0.76, CI = 0.61–0.95, p = 0.00029*















Cases MAF = 0.0520; controls MAF = 0.0921; OR = 0.54,CI = 0.31–0.93, p = 0.0016 Cases MAF = 0.0581; controls MAF = 0.0521; OR = 0.54, CI = 0.32–0.96, p = 0.0019



Statistics

Table 2 results after imputation of the genes of interest. Only SNPs with info[0.8 were included in the analysis. SNPs with p B 0.005 are shown. No SNP reached this threshold for the BDNF (113 SNPs available in this gene)


123

123 0.9731 0.9705 0.8004 0.8091

rs9567745

rs2760348

rs732821

rs731245

0.9698 0.9731

0.9673

rs4942583

rs1928039

0.8967

rs9534478

rs3945573

0.9614 0.9792 1.0000

rs13379489

0.8735

rs56027313

rs10852173 rs1928038

0.8804 1.0000

rs1455775

rs1352323

0.9447 0.8865

rs7166370

rs3848120

Dominant

0.9362 0.8830

rs339996

rs2414679

Recessive

Recessive

Additive

Additive

Additive

Additive

Additive

Additive

Dominant Additive

Dominant

Dominant

Recessive

Dominant

Recessive

Recessive

Recessive

Dominant Recessive

0.8257 0.8995

rs12440105

rs8041438

Model

Info

SNP



Other side effects


Cases

Cases

Phenotype

*

Results with p \ 0.0006

Cases presence of side effect(s), controls absence of side effect(s), CI 95 % confidence interval

HTR2A

ST8SIA2

Gene

Table 2 continued







Cases MAF = 0.0354; controls MAF = 0.0681; OR = 1.99, CI = 1.15–3.44,p = 0.0027

Cases MAF = 0.0333; controls MAF = 0.0642; OR = 0.50, CI = 0.29–0.88,p = 0.0025

Cases MAF = 0.1908; controls MAF = 0.1444, OR = 0.72, CI = 0.53–0.95, p = 0.0030 Cases MAF = 0.0353; controls MAF = 0.0712; OR = 2.10,CI = 1.22–3.61, p = 0.0020

Cases MAF = 0.1923; controls MAF = 0.1450, OR = 0.71,CI = 0.53–0.96, p = 0.0032

Cases MAF = 0.2043; controls MAF = 0.1566, OR = 0.72, CI = 0.54–0.96, p = 0.0029



Cases MAF = 0.3252; controls MAF = 0.3855; OR = 1.30; CI = 1.003–1.69, p = 0.0039




MAF = 0.4258; controls MAF = 0.3483; OR = 0.72; CI = 0.57–0.91, p = 0.0047

MAF = 0.3358; controls MAF = 0.3056; OR = 0.87; CI = 0.69–1.10, p = 0.0045

Statistics

C. Fabbri et al.


of the same group of ADRs induced by different psychotropic medications. For example, both some antidepressants and antipsychotics can induce gastrointestinal ADRs through HTR2A receptor antagonism. Finally, no polymorphism in the SP4 gene was available in the STEP-BD and imputation was not possible for ZNF804A and CREB1 because of the low number of genotyped SNPs. Taking into consideration the discussed limitations of the study, the main findings that may help in the identification of genes involved in ADRs induced by psychotropic medications are ST8SIA2 and RORA for the risk of psychic ADRs, and HTR2A for the risk of autonomic and other ADRs. Confirmations are needed particularly for ST8SIA2 and RORA given the poor available knowledge regarding the role of these genes in the pharmacodynamics of psychotropic medications. Fig. 1 Results obtained for RORA SNPs after imputation in the STEP-BD. Top findings are in the first intron of the gene

previous studies including a meta-analysis (Kato and Serretti 2010), as well as in the present study the risk of ADRs is higher in CC, intermediate in TC, and lower in TT subjects (40, 31.3, and 19.4 % reported ADRs, respectively). The consistency of our finding with previous studies reduces the risk of a false positive regarding HTR2A, while no previous study investigated RORA and CREB1 genes in relation to ADRs. The main limitations of the present study are represented by the small size of the COPE-BD sample and the heterogeneity between the two investigated samples (Supplementary Table 1). The tentative replication of findings across two independent samples was used to balance the small size of the COPE-BD and the multiple testing correction considering the number of SNPs and nor the number of side effect categories neither the different genetic models analyzed. Different genetic models were analyzed since the lack of knowledge about the inheritance model of the investigated SNPs. The analysis of imputed SNPs in the STEP-BD was not subjected to a further multiple testing correction, assuming that these SNPs showed a certain degree of linkage disequilibrium with genotyped SNPs and results of imputation show a per se preliminary nature. The use of a candidate approach entailed the restriction of the focus on some genes and excluded other potential candidates. Further, few previous studies were available to guide the selection of genes (see also Introduction). Another limitation is represented by the use of a polypharmacological treatment in both samples and the use of the covariate monotherapy versus polytherapy did not allow to discriminate the effects due to the individual drugs or classes. Anyway, we hypothesized an overlap among the mechanisms involved in the physiology

Acknowledgments The STEP-BD project was funded in whole or in part with Federal Funds from the National Institute of Mental Health (NIMH), under contract N01MH8001. Collection of DNA from consenting participants in STEP-BD was supported by N01MH80001 (Gary S. Sachs, M.D., principal investigator). Sample collection funding was supported by NIH grants (MH067288, Pamela Sklar; MH063445, Jordan W. Smoller; MH63420, Vish Nimgaonkar). Genotyping was funded by grants from NIH (MH067288 Pamela Sklar, PI), Johnson & Johnson Pharmaceutical Research and Development, Sylvan C. Herman Foundation, Stanley Medical Research Foundation, and Merck Genome Research Institute (Edward Scolnick, PI). We thank Manuel Mayhaus (University of Saarlandes, Germany) for technical assistance with Sequenom MassArray platform.

References Al-Qudah M, Anderson CD, Mahavadi S, Bradley ZL, Akbarali HI, Murthy KS, Grider JR (2014) Brain-derived neurotrophic factor enhances cholinergic contraction of longitudinal muscle of rabbit intestine via activation of phospholipase C. Am J Physiol Gastrointest Liver Physiol 306:G328–G337 Anderson CA, Pettersson FH, Clarke GM, Cardon LR, Morris AP, Zondervan KT (2010) Data quality control in genetic casecontrol association studies. Nat Protoc 5:1564–1573 Angata K, Nakayama J, Fredette B, Chong K, Ranscht B, Fukuda M (1997) Human STX polysialyltransferase forms the embryonic form of the neural cell adhesion molecule. Tissue-specific expression, neurite outgrowth, and chromosomal localization in comparison with another polysialyltransferase, PST. J Biol Chem 272:7182–7190 Baldessarini R, Henk H, Sklar A, Chang J, Leahy L (2008) Psychotropic medications for patients with bipolar disorder in the United States: polytherapy and adherence. Psychiatr Serv 59:1175–1183 Balog Z, Kiss I, Keri S (2011) ZNF804A may be associated with executive control of attention. Genes Brain Behav 10:223–227 Basile VS, Ozdemir V, Masellis M, Meltzer HY, Lieberman JA, Potkin SG, Macciardi FM, Petronis A, Kennedy JL (2001) Lack of association between serotonin-2A receptor gene (HTR2A) polymorphisms and tardive dyskinesia in schizophrenia. Mol Psychiatry 6:230–234

123

C. Fabbri et al. Beaulieu JM, Gainetdinov RR, Caron MG (2009) Akt/GSK3 signaling in the action of psychotropic drugs. Annu Rev Pharmacol Toxicol 49:327–347 Black AR, Black JD, Azizkhan-Clifford J (2001) Sp1 and kruppellike factor family of transcription factors in cell growth regulation and cancer. J Cell Physiol 188:143–160 Bouwman P, Philipsen S (2002) Regulation of the activity of Sp1related transcription factors. Mol Cell Endocrinol 195:27–38 Cacabelos R, Martinez-Bouza R, Carril JC, Fernandez-Novoa L, Lombardi V, Carrera I, Corzo L, McKay A (2012) Genomics and pharmacogenomics of brain disorders. Curr Pharm Biotechnol 13:674–725 Calati R, Crisafulli C, Balestri M, Serretti A, Spina E, Calabro M, Sidoti A, Albani D, Massat I, Hofer P, Amital D, Juven-Wetzler A, Kasper S, Zohar J, Souery D, Montgomery S, Mendlewicz J (2013) Evaluation of the role of MAPK1 and CREB1 polymorphisms on treatment resistance, response and remission in mood disorder patients. Prog Neuropsychopharmacol Biol Psychiatry 44:271–278 Crisafulli C, Drago A, Sidoti A, Serretti A (2013) A genetic dissection of antipsychotic induced movement disorders. Curr Med Chem 20:312–330 de Leon J, Susce MT, Pan RM, Fairchild M, Koch WH, Wedlund PJ (2005) The CYP2D6 poor metabolizer phenotype may be associated with risperidone adverse drug reactions and discontinuation. J Clin Psychiatry 66:15–27 Demyttenaere K, Albert A, Mesters P, Dewe W, De Bruyckere K, Sangeleer M (2005) What happens with adverse events during 6 months of treatment with selective serotonin reuptake inhibitors? J Clin Psychiatry 66:859–863 Drago A, Crisafulli C, Serretti A (2011) The genetics of antipsychotic induced tremors: a genome-wide pathway analysis on the STEPBD SCP sample. Am J Med Genet B Neuropsychiatr Genet 156B:975–986 Esslinger C, Kirsch P, Haddad L, Mier D, Sauer C, Erk S, Schnell K, Arnold C, Witt SH, Rietschel M, Cichon S, Walter H, MeyerLindenberg A (2011) Cognitive state and connectivity effects of the genome-wide significant psychosis variant in ZNF804A. Neuroimage 54:2514–2523 Fabbri C, Di Girolamo G, Serretti A (2013) Pharmacogenetics of antidepressant drugs: an update after almost 20 years of research. Am J Med Genet B Neuropsychiatr Genet 162B:487–520 Foltynie T, Cheeran B, Williams-Gray CH, Edwards MJ, Schneider SA, Weinberger D, Rothwell JC, Barker RA, Bhatia KP (2009) BDNF val66met influences time to onset of levodopa induced dyskinesia in Parkinson’s disease. J Neurol Neurosurg Psychiatry 80:141–144 Frederic F, Chianale C, Oliver C, Mariani J (2006) Enhanced endocrine response to novel environment stress and lack of corticosterone circadian rhythm in staggerer (Rora sg/sg) mutant mice. J Neurosci Res 83:1525–1532 Garfield LD, Dixon D, Nowotny P, Lotrich FE, Pollock BG, Kristjansson SD, Dore PM,Lenze EJ (2013) Common selective serotonin reuptake inhibitor side effects in older adults associated with genetic polymorphisms in the serotonin transporter and receptors: data from a randomized controlled trial. Am J Geriatr Psychiatry [Epub ahead of print] Girgenti MJ, LoTurco JJ, Maher BJ (2012) ZNF804a regulates expression of the schizophrenia-associated genes PRSS16, COMT, PDE4B, and DRD2. PLoS One 7:e32404 Gollner H, Bouwman P, Mangold M, Karis A, Braun H, Rohner I, Del Rey A, Besedovsky HO, Meinhardt A, van den Broek M, Cutforth T, Grosveld F, Philipsen S, Suske G (2001) Complex phenotype of mice homozygous for a null mutation in the Sp4 transcription factor gene. Genes Cells 6:689–697

123

Grimes CA, Jope RS (2001) The multifaceted roles of glycogen synthase kinase 3beta in cellular signaling. Prog Neurobiol 65:391–426 Gunes A, Scordo MG, Jaanson P, Dahl ML (2007) Serotonin and dopamine receptor gene polymorphisms and the risk of extrapyramidal side effects in perphenazine-treated schizophrenic patients. Psychopharmacology 190:479–484 Hammonds MD, Shim SS (2009) Effects of 4-week treatment with lithium and olanzapine on levels of brain-derived neurotrophic factor, B-cell CLL/lymphoma 2 and phosphorylated cyclic adenosine monophosphate response element-binding protein in the sub-regions of the hippocampus. Basic Clin Pharmacol Toxicol 105:113–119 Hauser RA, Molho E, Shale H, Pedder S, Dorflinger EE (1998) A pilot evaluation of the tolerability, safety, and efficacy of tolcapone alone and in combination with oral selegiline in untreated Parkinson’s disease patients. Tolcapone De Novo Study Group. Mov Disord 13:643–647 Heldt SA, Zimmermann K, Parker K, Gaval M, Weinshenker D, Ressler KJ (2014) BDNF deletion or TrkB impairment in amygdala inhibits both appetitive and aversive learning. J Neurosci 34:2444–2450 Hess JL, Glatt SJ (2014) How might ZNF804A variants influence risk for schizophrenia and bipolar disorder? A literature review, synthesis, and bioinformatic analysis. Am J Med Genet B Neuropsychiatr Genet 165:28–40 Ikeda E, Matsunaga N, Kakimoto K, Hamamura K, Hayashi A, Koyanagi S, Ohdo S (2013) Molecular mechanism regulating 24-hour rhythm of dopamine D3 receptor expression in mouse ventral striatum. Mol Pharmacol 83:959–967 Joels M, Pasricha N, Karst H (2013) The interplay between rapid and slow corticosteroid actions in brain. Eur J Pharmacol 719:44–52 Journiac N, Jolly S, Jarvis C, Gautheron V, Rogard M, Trembleau A, Blondeau JP, Mariani J, Vernet-der Garabedian B (2009) The nuclear receptor ROR(alpha) exerts a bi-directional regulation of IL-6 in resting and reactive astrocytes. Proc Natl Acad Sci USA 106:21365–21370 Kato M, Serretti A (2010) Review and meta-analysis of antidepressant pharmacogenetic findings in major depressive disorder. Mol Psychiatry 15:473–500 Lafuente A, Bernardo M, Mas S, Crescenti A, Aparici M, Gasso P, Deulofeu R, Mane A, Catalan R, Carne X (2008) Polymorphism of dopamine D2 receptor (TaqIA, TaqIB, and-141C Ins/Del) and dopamine degradation enzyme (COMT G158A, A-278G) genes and extrapyramidal symptoms in patients with schizophrenia and bipolar disorders. Psychiatry Res 161:131–141 Leclerc E, Mansur RB, Brietzke E (2013) Determinants of adherence to treatment in bipolar disorder: a comprehensive review. J Affect Disord 149:247–252 Lingjaerde O, Ahlfors UG, Bech P, Dencker SJ, Elgen K (1987) The UKU side effect rating scale. A new comprehensive rating scale for psychotropic drugs and a cross-sectional study of side effects in neuroleptic-treated patients. Acta Psychiatr Scand Suppl 334:1–100 Lisowski P, Juszczak GR, Goscik J, Stankiewicz AM, Wieczorek M, Zwierzchowski L, Swiergiel AH (2013) Stress susceptibilityspecific phenotype associated with different hippocampal transcriptomic responses to chronic tricyclic antidepressant treatment in mice. BMC Neurosci 14:144 Liu MT, Kuan YH, Wang J, Hen R, Gershon MD (2009) 5-HT4 receptor-mediated neuroprotection and neurogenesis in the enteric nervous system of adult mice. J Neurosci 29:9683–9699 Lotta T, Vidgren J, Tilgmann C, Ulmanen I, Melen K, Julkunen I, Taskinen J (1995) Kinetics of human soluble and membranebound catechol O-methyltransferase: a revised mechanism and

Genetics of side effects in bipolar disorder description of the thermolabile variant of the enzyme. Biochemistry 34:4202–4210 McCarthy MJ, Wei H, Marnoy Z, Darvish RM, McPhie DL, Cohen BM, Welsh DK (2013) Genetic and clinical factors predict lithium’s effects on PER2 gene expression rhythms in cells from bipolar disorder patients. Transl Psychiatry 3:e318 McClay JL, Adkins DE, Aberg K, Bukszar J, Khachane AN, Keefe RS, Perkins DO, McEvoy JP, Stroup TS, Vann RE, Beardsley PM, Lieberman JA, Sullivan PF, van den Oord EJ (2011) Genome-wide pharmacogenomic study of neurocognition as an indicator of antipsychotic treatment response in schizophrenia. Neuropsychopharmacology 36:616–626 Muller DJ, Chowdhury NI, Zai CC (2013) The pharmacogenetics of antipsychotic-induced adverse events. Curr Opin Psychiatry 26:144–150 Pinacho R, Villalmanzo N, Lalonde J, Haro JM, Meana JJ, Gill G, Ramos B (2011) The transcription factor SP4 is reduced in postmortem cerebellum of bipolar disorder subjects: control by depolarization and lithium. Bipolar Disord 13:474–485 Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, Bender D, Maller J, Sklar P, de Bakker PI, Daly MJ, Sham PC (2007) PLINK: a tool set for whole-genome association and populationbased linkage analyses. Am J Hum Genet 81:559–575 Racaniello M, Cardinale A, Mollinari C, D’Antuono M, De Chiara G, Tancredi V, Merlo D (2010) Phosphorylation changes of CaMKII, ERK1/2, PKB/Akt kinases and CREB activation during early long-term potentiation at Schaffer collateral-CA1 mouse hippocampal synapses. Neurochem Res 35:239–246 Ray MT, Shannon Weickert C, Webster MJ (2014) Decreased BDNF and TrkB mRNA expression in multiple cortical areas of patients with schizophrenia and mood disorders. Transl Psychiatry 4:e389 Sachs GS, Guille C, McMurrich SL (2002) A clinical monitoring form for mood disorders. Bipolar Disord 4:323–327 Sachs GS, Thase ME, Otto MW, Bauer M, Miklowitz D, Wisniewski SR, Lavori P, Lebowitz B, Rudorfer M, Frank E, Nierenberg AA, Fava M, Bowden C, Ketter T, Marangell L, Calabrese J, Kupfer D, Rosenbaum JF (2003) Rationale, design, and methods of the systematic treatment enhancement program for bipolar disorder (STEP-BD). Biol Psychiatry 53:1028–1042 Saiz PA, Susce MT, Clark DA, Kerwin RW, Molero P, Arranz MJ, de Leon J (2008) An investigation of the alpha1A-adrenergic receptor gene and antipsychotic-induced side-effects. Hum Psychopharmacol 23:107–114 Sato C, Kitajima K (2013) Impact of structural aberrancy of polysialic acid and its synthetic enzyme ST8SIA2 in schizophrenia. Front Cell Neurosci 7:61 Segman RH, Heresco-Levy U, Finkel B, Goltser T, Shalem R, Schlafman M, Dorevitch A, Yakir A, Greenberg D, Lerner A, Lerer B (2001) Association between the serotonin 2A receptor gene and tardive dyskinesia in chronic schizophrenia. Mol Psychiatry 6:225–229 Serretti A, Chiesa A, Calati R, Fabbri C, Sentissi O, De Ronchi D, Mendlewicz J, Souery D (2013) Side effects associated with psychotropic medications in patients with bipolar disorder: evidence from two independent samples. J Psychopharmacol 27:616–628 Shaw AD, Tiwari Y, Kaplan W, Heath A, Mitchell PB, Schofield PR, Fullerton JM (2014) Characterisation of genetic variation in ST8SIA2 and its interaction region in NCAM1 in patients with bipolar disorder. PLoS One 9:e92556 Sheehan DV, Lecrubier Y, Sheehan KH, Amorim P, Janavs J, Weiller E, Hergueta T, Baker R, Dunbar GC (1998) The miniinternational neuropsychiatric interview (M.I.N.I.): the development and validation of a structured diagnostic psychiatric

interview for DSM-IV and ICD-10. J Clin Psychiatry 59((Suppl 20)):22–33 (quiz 34–57) Sikander A, Rana SV, Prasad KK (2009) Role of serotonin in gastrointestinal motility and irritable bowel syndrome. Clin Chim Acta 403:47–55 Simons CJ, van Winkel R, Group (2013) Intermediate phenotype analysis of patients, unaffected siblings, and healthy controls identifies VMAT2 as a candidate gene for psychotic disorder and neurocognition. Schizophr Bull 39:848–856 Souery D, Zaninotto L, Calati R, Linotte S, Sentissi O, Amital D, Moser U, Kasper S, Zohar J, Mendlewicz J, Serretti A (2011) Phenomenology of psychotic mood disorders: lifetime and major depressive episode features. J Affect Disord 135:241–250 Souery D, Zaninotto L, Calati R, Linotte S, Mendlewicz J, Sentissi O, Serretti A (2012) Depression across mood disorders: review and analysis in a clinical sample. Compr Psychiatry 53:24–38 Srivastava V, Varma PG, Prasad S, Semwal P, Nimgaonkar VL, Lerer B, Deshpande SN, Bk T (2006) Genetic susceptibility to tardive dyskinesia among schizophrenia subjects: IV. Role of dopaminergic pathway gene polymorphisms. Pharmacogenet Genomics 16:111–117 Staeker J, Leucht S, Laika B, Steimer W (2014) Polymorphisms in serotonergic pathways influence the outcome of antidepressant therapy in psychiatric inpatients. Genet Test Mol Biomark 18:20–31 Staudt MD, de Oliveira CV, Lehman MN, McKenna KE, Coolen LM (2010) Activation of MAP kinase in lumbar spinothalamic cells is required for ejaculation. J Sex Med 7:2445–2457 Stojkov NJ, Janjic MM, Baburski AZ, Mihajlovic AI, Drljaca DM, Sokanovic SJ, Bjelic MM, Kostic TS, Andric SA (2013) Sustained in vivo blockade of alpha(1)-adrenergic receptors prevented some of stress-triggered effects on steroidogenic machinery in Leydig cells. Am J Physiol Endocrinol Metab 305:E194–E204 Utge SJ, Soronen P, Loukola A, Kronholm E, Ollila HM, Pirkola S, Porkka-Heiskanen T, Partonen T, Paunio T (2010) Systematic analysis of circadian genes in a population-based sample reveals association of TIMELESS with depression and sleep disturbance. PLoS One 5:e9259 Van Den Bossche MJ, Docx L, Morrens M, Cammaerts S, Strazisar M, Bervoets C, Smolders S, Depreeuw V, Lenaerts AS, De Rijk P, Del-Favero J, Sabbe BG (2012) Less cognitive and neurological deficits in schizophrenia patients carrying risk variant in ZNF804A. Neuropsychobiology 66:158–166 van der Heijden KB, de Sonneville LM, Swaab H (2013) Association of eveningness with problem behavior in children: a mediating role of impaired sleep. Chronobiol Int 30:919–929 Van Rheenen TE, Rossell SL (2013) Genetic and neurocognitive foundations of emotion abnormalities in bipolar disorder. Cogn Neuropsychiatry 18:168–207 Whiteford HA, Degenhardt L, Rehm J, Baxter AJ, Ferrari AJ, Erskine HE, Charlson FJ, Norman RE, Flaxman AD, Johns N, Burstein R, Murray CJ, Vos T (2013) Global burden of disease attributable to mental and substance use disorders: findings from the Global Burden of Disease Study 2010. Lancet 382:1575–1586 Wilffert B, Al Hadithy AF, Sing VJ, Matroos G, Hoek HW, van Os J, Bruggeman R, Brouwers JR, van Harten PN (2009) The role of dopamine D3, 5-HT2A and 5-HT2C receptor variants as pharmacogenetic determinants in tardive dyskinesia in AfricanCaribbean patients under chronic antipsychotic treatment: curacao extrapyramidal syndromes study IX. J Psychopharmacol 23:652–659 Yamashita S, Tai P, Charron J, Ko C, Ascoli M (2011) The Leydig cell MEK/ERK pathway is critical for maintaining a functional

123

C. Fabbri et al. population of adult Leydig cells and for fertility. Mol Endocrinol 25:1211–1222 Yuan P, Zhou R, Wang Y, Li X, Li J, Chen G, Guitart X, Manji HK (2010) Altered levels of extracellular signal-regulated kinase signaling proteins in postmortem frontal cortex of individuals with mood disorders and schizophrenia. J Affect Disord 124:164–169 Zai CC, Tiwari AK, Muller DJ, De Luca V, Shinkai T, Shaikh S, Ni X, Sibony D, Voineskos AN, Meltzer HY, Lieberman JA, Potkin SG, Remington G, Kennedy JL (2010) The catechol-O-methyltransferase gene in tardive dyskinesia. World J Biol Psychiatry 11:803–812

123

Zhou X, Long JM, Geyer MA, Masliah E, Kelsoe JR, Wynshaw-Boris A, Chien KR (2005) Reduced expression of the Sp4 gene in mice causes deficits in sensorimotor gating and memory associated with hippocampal vacuolization. Mol Psychiatry 10:393–406 Zivkovic M, Mihaljevic-Peles A, Bozina N, Sagud M, NikolacPerkovic M, Vuksan-Cusa B, Muck-Seler D (2013) The association study of polymorphisms in DAT, DRD2, and COMT genes and acute extrapyramidal adverse effects in male schizophrenic patients treated with haloperidol. J Clin Psychopharmacol 33:593–599