Chronobiology International, Early Online: 1–8, (2014) ! Informa Healthcare USA, Inc. ISSN: 0742-0528 print / 1525-6073 online DOI: 10.3109/07420528.2014.906445

ORIGINAL ARTICLE

Association between circadian genes, bipolar disorders and chronotypes

Chronobiol Int Downloaded from informahealthcare.com by University Library Utrecht on 06/02/14 For personal use only.

B. Etain1,2,3, S. Jamain1,3, V. Milhiet1,2,4, M. Lajnef1, C. Boudebesse1,2,3, A. Dumaine1,3,4, F. Mathieu5, A. Gombert6, K. Ledudal6, S. Gard3,7, J. P. Kahn3,8, C. Henry1,2,3,4, A. Boland9, D. Zelenika9, D. Lechner9, M. Lathrop9, M. Leboyer1,2,3,4, and F. Bellivier1,3,10,11 1

Inserm U955, Equipe Psychiatrie Ge´ne´tique, Cre´teil, France, 2AP-HP, Hoˆpital H. Mondor – A. Chenevier, Poˆle de Psychiatrie, Cre´teil, France, 3Fondation Fondamental, Cre´teil, France, 4Faculte´ de Me´decine, Universite´ Paris Est, Cre´teil, France, 5 Faculte´ de Me´decine de Villemin, INSERM UMR-S958, Universite´ Paris 7, Paris, France, 6INSERM, Centre d’Investigation Clinique 006, Groupe hospitalier Henri Mondor, Poˆle Recherche Clinique Sante´ Publique, Cre´teil, France, 7Service de Psychiatrie Adulte, Hoˆpital Charles Perrens, Bordeaux, France, 8Service de Psychiatrie et Psychologie Clinique, CHU de Nancy, Hoˆpitaux de Brabois, Vandoeuvre Les Nancy, France, 9Centre National de Ge´notypage, CEA, Evry, France, 10 Service de Psychiatrie, AP-HP, Groupe Hospitalier Saint-Louis – Lariboisie`re – F. Widal, Paris, France, and 11UFR de Me´decine, Universite´ Denis Diderot (Paris 7), Paris, France

Abnormalities in circadian rhythms play an important role in the pathogenesis of bipolar disorders (BD). Previous genetic studies have reported discrepant results regarding associations between circadian genes and susceptibility to BD. Furthermore, plausible behavioral consequences of at-risk variants remain unclear since there is a paucity of correlates with phenotypic biomarkers such as chronotypes. Here, we combined association studies with a genotype/ phenotype correlation in order to determine which circadian genes variants may be associated with the circadian phenotypes observed in patients with BD. First, we compared the allele frequencies of 353 single nucleotide polymorphisms spanning 21 circadian genes in two independent samples of patients with BD and controls. The meta-analysis combining both samples showed a significant association between rs774045 in TIMELESS (OR ¼ 1.49 95%CI[1.18–1.88]; p ¼ 0.0008) and rs782931 in RORA (OR ¼ 1.31 95%CI[1.12–1.54]; p ¼ 0.0006) and BD. Then we used a ‘‘reverse phenotyping approach’’ to look for association between these two polymorphisms and circadian phenotypes in a subsample of patients and controls. We found that rs774045 was associated with eveningness (p ¼ 0.04) and languid circadian type (p ¼ 0.01), whereas rs782931 was associated with rigid circadian type (p ¼ 0.01). Altogether, these findings suggest that these variants in the TIMELESS and RORA genes may confer susceptibility to BD and impact on circadian phenotypes in carriers who thus had lower ability to properly adapt to external cues. Keywords: Bipolar disorders, chronotype, circadian rhythms, clock genes, RORA, TIMELESS

INTRODUCTION

neuroendocrine hormones (such as cortisol), core temperature circadian variations (Lee et al., 2010), melatonin levels (Lewy et al., 1985; Nurnberger et al., 1988, 2000; Pacchierotti et al., 2001) or fibroblast activity as assessed from the rhythmic expression patterns of clock genes (Yang et al., 2009). Most of these circadian dysfunctions are observed not only in remitted bipolar patients, but also among at-risk subjects (mainly unaffected first-degree relatives) (Milhiet et al., 2014). Therefore, these stable and trait-like circadian characteristics are potentially relevant biomarkers of BD since they are thought to represent the direct output of an underlying chronobiological and genetically determined

Bipolar disorder (BD) is a severe and chronic psychiatric illness that is characterized by alternating (hypo)manic and major depressive episodes, separated by euthymic periods (A.P.A., 1994). A novel avenue for pathophysiological research (Leboyer & Kupfer, 2010) has been recently opened since severe deregulations of circadian rhythms have been widely observed among patients and are thought to represent putative markers of the disorder (Etain et al., 2011; Milhiet et al., 2011, 2014). Indeed, numerous circadian physiological functions are abnormal in patients with BD both during acute and euthymic phases. This includes the secretion profile of

Submitted January 28, 2014, Returned for revision March 4, 2014, Accepted March 17, 2014

Correspondence: Bruno Etain, Poˆle de Psychiatrie, Hoˆpital Albert Chenevier, 40, rue de Mesly, 94000 Cre´teil Cedex, France. Tel: + 33 1 49 81 32 90. Fax: + 33 1 49 81 30 99. E-mail: [email protected]

1

Chronobiol Int Downloaded from informahealthcare.com by University Library Utrecht on 06/02/14 For personal use only.

2

B. Etain et al.

susceptibility to BD (Harvey, 2008, 2011; McClung, 2013; Murray & Harvey, 2010). The observations described above have led to the hypothesis that genes encoding proteins implicated in the regulation of circadian rhythms may play a role in the susceptibility to BD, in particular those expressed within the suprachiasmatic nucleus (SCN). Many circadian genes have previously been associated with BD, including ARNTL1, CLOCK, NPAS2, NR1D1, PER3 and RORB (Kishi et al., 2008; Kripke et al., 2009; Lee et al., 2010; Mansour et al., 2006, 2009; McCarthy et al., 2012; McGrath et al., 2009; Nievergelt et al., 2006; Severino et al., 2009; Shi et al., 2008; Soria et al., 2010), multi locus interactions between BHLHB2-CSNK1"-CLOCK (Shi et al., 2008), and copy number variations within the GSK3B gene (Lachman et al., 2007). Furthermore, we have recently reported an association between common and rare variants in ASMT and BD (Etain et al., 2012). Convergent functional genomics integrating data from various animal and human studies also suggested that ARNTL1, GSK3B, RORA and RORB are among the most plausible candidate genes for BD (Le-Niculescu et al., 2009). To consolidate the level of evidence in favor of the implication of circadian genes in BD susceptibility, replications are required. Furthermore, correlations between genotypes and phenotypic markers (such as chronotypes) have not been sufficiently investigated to elucidate the links between genetic susceptibility and associated clinical/behavioral outputs. Some of these circadian phenotypes can be easily and non-invasively assessed using self-reports, such as scales for diurnal preference. This kind of measure is of simple use and has the major advantage to correlate with many endogenous circadian phase markers, such as melatonin peak time, temperature nadir, and measures of activity/rest (Duffy, Rimmer et al., 2001). Several previous studies demonstrated that patients with BD were more likely than healthy subjects to present with an evening chronotype (Ahn et al., 2008; Chung et al., 2012; Giglio et al., 2010; Mansour et al., 2005; Wood et al., 2009). This circadian preference is heritable and may thus be considered as genetically determined (Koskenvuo et al., 2007; Maxwell, 1992; Vink et al., 2001). Therefore, such chronotypes represent wellsuited candidate intermediate phenotypes for BD to be tested in association with circadian gene variants. Only two studies have attempted to associate genotypes and chronotypes among patients with BD and they reported a correlation between abnormal circadian preferences and susceptibility variants in CLOCK, PER3 and CSNK1" genes (Kripke et al., 2009; Lee et al., 2010). Therefore, combining classical genetic association studies of circadian gene polymorphisms and chronotypes as relevant intermediate phenotypes, may help the identification of susceptibility markers of BD and physiological pathways from gene to phenotypic outputs in a so-called ‘‘reverse phenotyping’’ approach (Leboyer, 2003; Schulze, 2010). We performed a genetic

association study between a large set of polymorphisms covering 21 circadian genes in two sample of patients with BD and control subjects and looked at possible correlations between susceptibility variants and circadian phenotypes (phase preference, stability and amplitude of circadian rhythms).

MATERIALS AND METHODS Case-control sample for the genetic association study Sample 1: Two hundred and thirty-nine remitted outpatients with BD (104 males and 135 females) were recruited in three university-affiliated psychiatric departments in France (Paris-Cre´teil, Bordeaux and Nancy). Patients met DSM-IV criteria (A.P.A., 1994) for BD (172 type I, 61 type II, 6 not otherwise specified). All subjects were assessed during a remission period (at least 3 months elapsed since the last major mood episode) by trained psychiatrists using the French version (Preisig et al., 1999) of the Diagnostic Interview for Genetic Studies (DIGS version 3.0) (Nurnberger et al., 1994). Eight hundred and seventy-three control individuals (360 males and 513 females) were sampled from the French general population. These controls were not screened for the presence of psychiatric disorders. Sample 2: Two hundred and forty additional patients (99 males and 141 females) with BD (195 type 1, 39 type 2, 6 not otherwise specified) and 886 control individuals (359 males and 527 females) were used for replication study. Patients and controls met identical clinical criteria as described for sample 1 and were recruited in the same psychiatric departments. All individuals were of French origin. This clinical research received the approval of the research ethics board of the Pitie´-Salpeˆtrie`re hospital (Paris, France). Written informed consent was obtained from all subjects prior to study participation. The experimental protocol is conform to international ethical standards (Portaluppi et al., 2010). Chronotype assessment A subsample of 100 patients with BD (70 type 1, 28 type 2 and 2 not otherwise specified) have been assessed for circadian phenotypes using the French validated version of the Composite Scale of Morningness (CSM) and the Circadian Type Inventory (CTI). The CSM is a validated adaptation of the Horne-Ostberg Scale of Morningness that assesses preferred times of day for achieving various activities (Caci et al., 1999; 2000; Smith et al., 1989). It is composed of 13 items. Score (ranging from 13 to 55) was used here as a continuous variable with lower scores indicating proneness towards evening typology. The CTI (Di Milia et al., 2005) measures the stability (Flexible/Rigid typology) and the amplitude (Languid/ Vigorous typology) of circadian rhythms. Two continuous scores were calculated: CTI-FR ranging from 5 to 25 (a higher score indicates greater flexibility) and CTI-LV Chronobiology International

Chronobiol Int Downloaded from informahealthcare.com by University Library Utrecht on 06/02/14 For personal use only.

Circadian genes, bipolar disorders and chronotypes ranging from 6 to 30 (a higher score indicates a greater languidness). We used the French validated version of the CTI (Boudebesse et al., 2013). The level of daytime sleepiness was assessed using the Epworth Sleepiness Scale (Johns, 1991). Seventy-two super-normal controls free of any personal history of DSM-IV axis I psychiatric disorders (as assessed with the DIGS) and of first-degree relatives with history of mood disorders, schizophrenia or suicide attempts (as assessed with the Family Interview for Genetic Studies) (Maxwell, 1992) have been specifically recruited for circadian phenotype assessment, using aforementioned questionnaires. Patients with BD and super-normal controls have been assessed for circadian phenotypes between January and August 2009. Euthymic state as well as residual mood symptoms were evaluated using the Quick Inventory of Depressive SymptomatologySelf-Rated 16 items (QIDS-SR16) (Corruble et al., 1999; Rush et al., 2003) and the Altman Self-Rating Mania Scale (Altman et al., 1997). Cut-off values for exclusion were set at 16 for the QIDS-SR16 and 6 for the Altman Self-Rating Mania Scale. Participants exposed to life events with a potential impact on wake/sleep rhythms (shift work, transmeridian flight, birth of a child, grief or psycho-traumatic events) within the two months before the assessment were not included. Patients with a current or recent (less than 3 months) DSM-IV criteria mood episode were not included. Details of current prescribed psychotropic drugs were obtained from the patients.

GENOTYPING AND QUALITY CONTROL Genomic DNA samples were extracted from peripheral blood leukocytes or cell lines by standard procedures. Genotyping was performed at the Centre National de Ge´notypage (CNG – France) using HumanHap550 or 610-Quad Beadchips (Illumina Inc., San Diego, CA) for patients with BD, and HumanHap300 Beadchips (Illumina Inc.) for controls. Three hundred and fiftythree single nucleotide polymorphisms (SNPs) spanning 21 circadian genes were analyzed (CLOCK, NPAS2, ARNTL1, ARNTL2, PER1, PER2, PER3, CRY1, CRY2, TIMELESS, NR1D1, RORA, RORB, RORC, CSNK1, CSNK1", GSK3, DBP, BHLHB2, BHLHB3, PPARGC1A) (Supplementary Table 1). Genes were selected for their involvement in the molecular mechanisms of the regulation of circadian rhythms. We collected data for all available SNPs on DNA chips within each gene as defined in the RefSeq Database (National Center for Biotechnology Information) as well as within the 10-kilobase pairs (kbp) upstream and downstream flanking regions The mean number of SNPs per gene was 16 (±31). All available SNPs with minor allele frequencies (MAF) higher than 1% and a genotyping call rate higher than 95% were used in this study. For all DNA chips, the call rate averaged 99.6%. In the control !

Informa Healthcare USA, Inc.

3

population, deviations from the Hardy–Weinberg equilibrium were tested for each SNP using a global significance threshold of p ¼ 0.01 and only SNPs that fulfilled this criterion were studied.

Statistical analysis Analyses were conducted SNP by SNP using classical single-marker association analyses using Fisher’s exact test and have been performed using PLINK software 1.07 (Purcell et al., 2007). Only SNPs for which allele frequencies showed a nominal p-value 50.05 for the difference between cases and controls in sample 1 were tested for replication in sample 2. A meta-analysis of SNPs associated with a nominal p50.05 in both samples was performed. The fixed effect model (Woolf’s method) was used to obtain combined ORs (Greenland, 1987). The fixed effect model assumes that the effect is the same in the two samples pooled and that the differences observed between samples are only due to random measurement errors. When this model is used, only results for which the homogeneity hypothesis cannot be rejected deserve to be reported. We used Review Manager 5.2 (Cochrane Collaboration, Copenhagen, Denmark). CSM and CTI scores were compared between patients with bipolar disorders and super-normal control subjects. Normal distribution of continuous variables was tested using the Shapiro–Wilk test. Group-wise comparisons included parametric (Student’s t test) or non-parametric (Mann–Whitney test) tests as appropriate. Association between circadian scores and genotypes was then tested using linear regression, using adjustment for covariates. Analyses were performed using the Statistical Analysis Software (SAS). RESULTS A graphical overview of the methodological approach is described in supplementary figure 1.

Case-control association study Allele frequencies of 353 SNPs spanning 21 circadian genes were compared between 239 patients with BD and 873 controls (sample 1). Fourteen SNPs showed a nominal p value 50.05 and were then used for a replication study using sample 2, including 240 patients with BD and 886 controls (Table 1). Three SNPs out of 14 showed a nominal p value 50.05 in the same direction than observed for sample 1. Two of them were located in TIMELESS (rs1082214 and rs7740455) and one in RORA (rs782931). Note that rs1082214 is located 120 bp upstream from TIMELESS and actually maps in the last intron of MIP, a gene encoding a major intrinsic protein of lens fibers. This SNP was in strong linkage disequilibrium with rs774045 (r2 ¼ 0.97) and was thus not further investigated. For rs774045 and rs782931, the meta-analysis showed no significant heterogeneity

4

B. Etain et al.

TABLE 1. Circadian gene polymorphisms and bipolar disorders. Sample 1

Chronobiol Int Downloaded from informahealthcare.com by University Library Utrecht on 06/02/14 For personal use only.

SNP rs356642 rs1082214 rs774045 rs774049 rs2553234* rs12438355 rs782909 rs782931 rs7870498 rs1570500 rs9826 rs3828057 rs1521177 rs4845604

2

Gene

Variant (1/2)

f(BD)-f(C)



NPAS2 TIMELESS TIMELESS TIMELESS RORA RORA RORA RORA RORB RORB RORC RORC RORC RORC

A/G T/C A/G T/C T/C A/G T/C A/G G/A C/A C/T T/C G/T A/G

0.15–0.19 0.12–0.08 0.12–0.08 0.09–0.05 0.12–0.16 0.19–0.24 0.10–0.14 0.29–0.35 0.41–0.46 0.22-0.16 0.41-0.35 0.38-0.46 0.55-0.46 0.17-0.13

3.9 5.5 5.9 6.3 4.1 5.7 4.7 5.9 4.4 8.2 5.2 8 11.7 4.2

Sample 2 2

p

f(BD)-f(C)



0.05 0.02 0.02 0.01 0.02 0.02 0.03 0.02 0.03 0.004 0.02 0.005 0.0006 0.04

0.14–0.16 0.11–0.08 0.11–0.08 0.07–0.06 0.19–0.14 0.25–0.23 0.12–0.14 0.31–0.37 0.48–0.48 0.16–0.17 0.37–0.36 0.43–0.45 0.47–0.46 0.15–0.15

0.8 4.9 4.1 1.5 5.6 0.9 1.3 5.3 0.0 0.1 0.1 0.9 0.3 0.1

Meta-analyses of samples 1 and 2 p 0.38 0.03 0.04 0.22 0.02 0.35 0.26 0.02 0.89 0.73 0.75 0.33 0.60 0.78

2hetero

phetero

2asso

passo

0.13 0.13

0.72 0.72

11.21 11.21

0.0008 0.0008

0.001

0.97

11.71

0.0006

SNP: Single Nucleotide Polymorphism; Variant 1 and 2: minor and major allele nucleotides respectively; f: minor allele frequency; BD: patients with bipolar disorders – C: controls; 2hetero and phetero: Chi-square and p value for the heterogeneity between samples; 2asso and passo: Chi-square and p value for association. SNPs associated in the two samples are indicated in bold. *rs2553234 showed associations in opposite direction in samples 1 and 2 and was not included in the meta-analysis.

between samples 1 and 2 and a significant association between BD and the rs774045 allele A (OR ¼ 1.49 95%CI[1.18–1.88], p ¼ 0.0008) and the rs782931 allele G (OR ¼ 1.31 95%CI[1.12–1.54]; p ¼ 0.0006) (Table 1). These results resisted to Bonferroni’s correction for multiple testing (rs774045, pBonferroni ¼ 0.02; rs782931, pBonferroni ¼ 0.01).

Association between genotypes and chronotypes One hundred patients with BD and 72 super-normal control subjects completed the circadian phenotypes assessment. The CSM, CTI-FR and CTI-LV scores were normally distributed (data not shown). Supplementary table 1 showed the differences observed between cases and controls. Patients with BD exhibited an evening preference (Student’s t test, t ¼ 2.6, p ¼ 0.01) and a rigid and languid circadian type (CTI-FR, Student’s t test, t ¼ 2.1, p ¼ 0.04; CTI-LV Student’s t test, t ¼ 4.5, p50.0001). Among patients, there was no significant association between circadian scores and the three major categories of currently prescribed psychotropic medication (mood stabilizers, antidepressants or/and sedative/hypnotics) (data not shown). Three distinct stepwise regressive multivariate models were designed with circadian scores (CSM, CTI-LV, CTI-FR) as dependent variables, and genotypes (rs774045 and rs782931), clinical status (patients versus controls) and potential confounding factors (depressive symptoms and gender) as independent variables. When taking into account clinical status and potential confounding variables, eveningness and languid circadian type were associated with the rs774045 the ‘‘AA’’ and ‘‘AG’’ genotypes in TIMELESS (p ¼ 0.04 and p ¼ 0.01, respectively) and rigid circadian type was associated with the rs782931 the ‘‘GG’’ and ‘‘AG’’ genotypes in RORA (p ¼ 0.03) (Table 2 and Supplementary Figures 2

and 3). As this ‘‘reverse phenotyping’’ step was exploratory and pilot, we applied no correction for multiple testing.

DISCUSSION In this study, we explored allelic frequencies of 353 SNPs in 21 circadian genes using two independent samples and reported significant associations between SNPs located in TIMELESS and RORA genes and BD. Previous studies have only provided suggestive signals for these two genes due to inconsistent findings (lack of replication in an independent sample or results not remaining significant after correction for the number of tested SNPs). Regarding TIMELESS, Shi et al. found no association with BD in a family sample; nevertheless, they found an association between two SNPs in TIMELESS and a BD sub-phenotype (insomnia during mania) (Shi et al., 2008). Mansour et al. used the transmission disequilibrium test (TDT) in two family samples and found modest associations (nominal p values 50.05 before correction for multiple findings) (Mansour et al., 2009). A similar situation has been observed for RORA with suggestive associations obtained in family samples (McGrath et al., 2009) or case-control studies (Soria et al., 2010), however neither replicated in an independent sample, nor significant after correction for multiple testing. The phenotypic analysis as regards genotyping data of the associated SNPs suggested that they might influence the circadian phenotypes among patients and controls. Indeed, the rs774045 ‘‘AA’’ and ‘‘AG’’ at-risk genotypes in TIMELESS were associated with phase delay (eveningness) and sensitivity to sleep inertia (languid type), whereas rs782931 ‘‘GG’’ and ‘‘AG’’ at-risk genotype in RORA were associated with less adaptability of circadian rhythms (rigid type). Chronobiology International

Circadian genes, bipolar disorders and chronotypes

5

TABLE 2. Correlations between rs774045 in TIMELESS and rs782931 in RORA genotypes with circadian phenotypes taking into account clinical status and potential confounding factors (gender and depressive symptoms). CSM Variables Intercept TIMELESS rs774045 [AG+AA versus GG*] RORA rs782931 [GG versus AA*] RORA rs782931 [GG versus AG*] Disease Status [Bipolar versus Controls*] QIDS score Gender [Female versus Male*]

Beta (SE) 36.12 2.91 1.83 1.97 1.97 0.24 0.05

a

(2.32) (1.45) (2.13) (2.07) (1.28) (0.16) (1.17)

CTI FR p

Beta (SE)

0.00 0.04 0.39 0.34 0.12 0.14 0.96

17.53 0.98 2.74 3.05 1.46 0.09 1.40

b

(1.38) (0.86) (1.26) (1.23) (0.76) (0.09) (0.69)

CTI LV p

Beta (SE)c

p

0.00 0.25 0.03 0.01 0.06 0.30 0.04

18.54 (1.70) 2.67 (1.06) 2.46 (1.56) 0.97 (1.52) 3.21 (0.94) 0.09 (0.11) 0.74 (0.86)

0.00 0.01 0.11 0.52 0.0009 0.43 0.38

Chronobiol Int Downloaded from informahealthcare.com by University Library Utrecht on 06/02/14 For personal use only.

*Reference categories for the regression: a: Beta50 indicates eveningness; b: Beta50 indicates rigidity; c: Beta40 indicates languidness.

TIMELESS is suspected to create a 24-hours rhythm in drosophila (Tomioka & Matsumoto, 2010) and conditional knockdown of TIMELESS protein expression in the rat suprachiasmatic nucleus disrupts neuronal activity rhythms and altered levels of known core clock elements (Barnes et al., 2003). This crucial role of TIMELESS in the regulation of circadian rhythms explain why some variants of this gene might be associated with a proneness to deregulation of the 24-hours rhythm, leading to evening preference and misalignment with external clues. One alternative mechanism by which TIMELESS acts on circadian rhythms may involve its close partnership with glycogen synthase kinase 3 b (GSK3B), another key candidate actor for BD susceptibility and, more generally, in mood regulation (Lachman et al., 2007; Li & Jope, 2010). Indeed, in Drosophila melanogaster, Shaggy, the GSK3B ortholog, phosphorylates Timeless and promotes nuclear translocation of the Period-Timeless complex (Harms et al., 2003). This is consistent with phosphorylation mechanisms of clock proteins regulating the period, phase and amplitude of the circadian clock (Bae & Edery, 2006). The polymorphism in TIMELESS described in this study may, either directly influence or be in linkage disequilibrium with another variation that may impact on this interaction. Nominally suggestive signals for RORA have been reported in several genome-wide association studies for BD and this gene was among the top candidates in convergent functional genomics (Baum et al., 2008; LeNiculescu et al., 2009; Sklar et al., 2008; Wellcome Trust Case Control Consortium, 2007). The RAR-related orphan receptor a (RORA) is a nuclear hormone receptor required for the consolidation of daily locomotor activity and is regulated by the core clock in the suprachiasmatic nucleus (Akashi & Takumi, 2005; Sato et al., 2004). When entrained in light–dark cycles, Staggerer mutant mice that are deficient for RORA activity show shorter free-running activity period lengths under constant darkness conditions (Gold et al., 2007). We recently showed that patients with BD when assessed for phase preference and for circadian typologies exhibited significant differences as compared to controls (Boudebesse et al., 2013). Three main !

Informa Healthcare USA, Inc.

characteristics can be measured using questionnaires: phase (CSM), amplitude (CTI Languidness/Vigor factor) and stability (CTI Flexibility/Rigidity factor). Rigid typology was associated with a greater need for sleep, greater difficulties with going to bed early and to sleep at unusual hours (Di Milia et al., 2005). Languid typology was associated with more lethargic feelings following reduced sleep, more difficulties with overcoming drowsiness and a greater need for sleep (Di Milia et al., 2005). These phenotypes are potentially related to patients’ difficulties to resynchronize following disruption in their social routines resulting in the onset of new mood episodes. The ‘‘social zeitgeber’’ theory suggests that patients with BD may present a biological alteration of internal circadian oscillators, resulting in less adaptability to endogenous and exogenous challenges, which thus contributes to affective deregulation (Ehlers et al., 1988; Grandin et al., 2006). Such circadian typologies are thought to be heritable traits (Klei et al., 2005; Koskenvuo et al., 2007; Vink et al., 2001), and thus influenced by circadian gene variants (Kripke et al., 2009; Lee et al., 2010). Our study highlights that critical components (phase, amplitude and stability) of circadian rhythms are likely to be influenced by circadian genes such as TIMELESS and RORA. Our results extend previous works showing associations between circadian preferences and some variants of CLOCK, PER3 and CSNK1" (Kripke et al., 2009; Lee et al., 2010) among patients with BD and between PER3 and morningness among healthy subjects (Lazar et al., 2012). Our results are consistent with recent etiological hypotheses in which circadian rhythm abnormalities are core vulnerability dimensions of BD (Etain et al., 2011, 2012; Milhiet et al., 2011; Murray & Harvey, 2010; Waddington et al., 2010). They are also consistent with the ‘‘social zeitgeber’’ theory suggesting that patients with BD show poor adaptability to endogenous and exogenous challenges. Harvey proposed a model for BD, which integrates genetic susceptibility, sleep and circadian functioning, neurotransmitter output and mood deregulation (Harvey, 2011). Genetic variants of circadian genes may predispose individuals to be relatively less able to adapt their circadian rhythms appropriately to their environment and to be prone to sleep

Chronobiol Int Downloaded from informahealthcare.com by University Library Utrecht on 06/02/14 For personal use only.

6

B. Etain et al.

disturbances. Since circadian and neurotransmission systems are tightly connected, circadian and/or sleeprelated abnormalities may affect the functioning of the dopamine or serotonin circuitry, which in turn may affect mood regulation. Some of the limitations of our study need to be considered. The first is the lack of psychiatric screening information concerning the controls used for the genetic association study. Second, our two steps strategy has the advantage of providing replication whilst it necessarily has reduced the power of detecting signals (of small magnitude) in other genes when performing the association study in one unique large sample. Third, the set of circadian genes that we selected based on previous publications may be considered as too limited since a recent study demonstrated the relevance of enrichment of this set by adding upstream circadian clock modulators and downstream clock controlled genes to the essential ‘‘core’’ clock genes (McCarthy et al., 2012). We only used questionnaires and the absence of more direct, biological measures of circadian alignment/phase was a possible limitation. Finally, the ‘‘reverse phenotyping’’ step of this study was considered as exploratory and pilot. As such, we applied no correction for multiple testing and this can represent a limitation. In conclusion, this genetic association and genotype/ phenotype correlation study suggests that variations in TIMELESS and RORA might act on some critical components of circadian rhythms (phase, amplitude and stability) that were observed in patients with BD. These findings may increase our understanding of the pathophysiological processes involved in BD and require further replication in independent samples of patients with BD assessed for chronotypes or other relevant circadian biomarkers. Indeed, novel etiological and clinical approaches based on chronobiology are strongly needed, both to identify clinically useful relevant biomarkers, and to correlate them with the genetic susceptibility to this disorder (Etain et al., 2011; Leboyer & Kupfer, 2010; Milhiet et al., 2011).

ACKNOWLEDGEMENTS We thank patients with BD and controls who agreed to participate in this study. We thank the staff at the inclusion sites in Paris-Cre´teil (A. Raust and B. Cochet for their active participation in the clinical assessment; E. Abadie for the organization of recruitment), Bordeaux (L. Zanouy) and Nancy (O. Wajsbrot-Elgrabli and RF. Cohen). We are also grateful to the Clinical Investigation Centre (O. Montagne and P. Le Corvoisier) and the Plateforme de Ressources Biologiques (B. Ghaleh) of Mondor Hospital, l’Etablissement Franc¸ais du Sang of Cre´teil (J.L. Beaumont and B Mignen), the Cochin Hospital cell library (J. Chelly), the Centre National de Ge´notypage (M. Lathrop) and J.R. Richard for technical assistance.

We thank S. Folkard who gave the authorization and instructions for using the CTI questionnaire.

DECLARATION OF INTEREST This work was supported by INSERM (Institut National de la Sante´ et de la Recherche Me´dicale), AP-HP (Assistance Publique des Hoˆpitaux de Paris), the ITMO Neurosciences, Cognitive Sciences, Neurology and Psychiatry, the Fondation Fondamental (fondation de coope´ration scientifique), the Fondation pour la Recherche Me´dicale (Grant to VM) and the Agence Nationale pour la Recherche (ANR NEURO2006, MANAGE_BPAD). This research was also supported the Investissements d’Avenir program managed by the ANR under reference ANR-11-IDEX-0004. The authors have no conflict of interest to report.

REFERENCES A.P.A. (1994). Diagnostic and statistical manual of mental disorders, 4th edition. Paris: Masson. Ahn YM, Chang J, Joo YH, et al. (2008). Chronotype distribution in bipolar I disorder and schizophrenia in a Korean sample. Bipolar Disord. 10:271–5. Akashi M, Takumi T. (2005). The orphan nuclear receptor RORalpha regulates circadian transcription of the mammalian core-clock Bmal1. Nat Struct Mol Biol. 12:441–8. Altman EG, Hedeker D, Peterson JL, Davis JM. (1997). The Altman self-rating mania scale. Biol Psychiatry. 42:948–55. Bae K, Edery I. (2006). Regulating a circadian clock’s period, phase and amplitude by phosphorylation: Insights from Drosophila. J Biochem. 140:609–17. Barnes JW, Tischkau SA, Barnes JA, et al. (2003). Requirement of mammalian timeless for circadian rhythmicity. Science. 302: 439–42. Baum AE, Akula N, Cabanero M, et al. (2008). A genome-wide association study implicates diacylglycerol kinase eta (DGKH) and several other genes in the etiology of bipolar disorder. Mol Psychiatry. 13:197–207. Boudebesse C, Lajnef M, Geoffroy PA, et al. (2013). Chronotypes of bipolar patients in remission: Validation of the French version of the circadian type inventory in the FACE-BD sample. Chronobiol Int. 30:1042–9. Caci H, Nadalet L, Staccini P, et al. (1999). Psychometric properties of the French version of the composite scale of morningness in adults. Eur Psychiatry. 14:284–90. Caci H, Nadalet L, Staccini P, et al. (2000). The composite scale of morningness: Further psychometric properties and temporal stability. Eur Psychiatry. 15:278–81. Chung JK, Lee KY, Kim SH, et al. (2012). Circadian rhythm characteristics in mood disorders: Comparison among bipolar i disorder, bipolar ii disorder and recurrent major depressive disorder. Clin Psychopharmacol Neurosci. 10: 110–16. Corruble E, Legrand JM, Duret C, et al. (1999). IDS-C and IDS-sr: Psychometric properties in depressed in-patients. J Affect Disord. 56:95–101. Di Milia L, Smith PA, Folkard S. (2005). A validation of the revised circadian type inventory in a working sample. Person Individual Differences. 39:1293–305. Duffy JF, Rimmer DW, Czeisler CA. (2001). Association of intrinsic circadian period with morningness-eveningness, Chronobiology International

Chronobiol Int Downloaded from informahealthcare.com by University Library Utrecht on 06/02/14 For personal use only.

Circadian genes, bipolar disorders and chronotypes usual wake time, and circadian phase. Behav Neurosci. 115: 895–9. Ehlers CL, Frank E, Kupfer DJ. (1988). Social zeitgebers and biological rhythms. A unified approach to understanding the etiology of depression. Arch Gen Psychiatry. 45:948–52. Etain B, Dumaine A, Bellivier F, et al. (2012). Genetic and functional abnormalities of the melatonin biosynthesis pathway in patients with bipolar disorder. Hum Mol Genet. 21: 4030–7. Etain B, Milhiet V, Bellivier F, Leboyer M. (2011). Genetics of circadian rhythms and mood spectrum disorders. Eur Neuropsychopharmacol. 21:S676–82. Giglio LM, Magalhaes PV, Andersen ML, et al. (2010). Circadian preference in bipolar disorder. Sleep Breath. 14:153–5. Gold DA, Gent PM, Hamilton BA. (2007). ROR alpha in genetic control of cerebellum development: 50 staggering years. Brain Res. 1140:19–25. Grandin LD, Alloy LB, Abramson LY. (2006). The social zeitgeber theory, circadian rhythms, and mood disorders: Review and evaluation. Clin Psychol Rev. 26:679–94. Greenland S. (1987). Quantitative methods in the review of epidemiologic literature. Epidemiol Rev. 9:1–30. Harms E, Young MW, Saez L. (2003). CK1 and GSK3 in the Drosophila and mammalian circadian clock. Novartis Found Symp. 253:267–77. Harvey AG. (2008). Sleep and circadian rhythms in bipolar disorder: Seeking synchrony, harmony, and regulation. Am J Psychiatry. 165:820–9. Harvey AG. (2011). Sleep and circadian functioning: Critical mechanisms in the mood disorders? Annu Rev Clin Psychol. 7:297–319. Johns MW. (1991). A new method for measuring daytime sleepiness: The Epworth sleepiness scale. Sleep. 14:540–5. Kishi T, Kitajima T, Ikeda M, et al. (2008). Association analysis of nuclear receptor Rev-erb alpha gene (NR1D1) with mood disorders in the Japanese population. Neurosci Res. 62: 211–15. Klei L, Reitz P, Miller M, et al. (2005). Heritability of morningnesseveningness and self-report sleep measures in a family-based sample of 521 hutterites. Chronobiol Int. 22:1041–54. Koskenvuo M, Hublin C, Partinen M, et al. (2007). Heritability of diurnal type: A nationwide study of 8753 adult twin pairs. J Sleep Res. 16:156–62. Kripke DF, Nievergelt CM, Joo E, et al. (2009). Circadian polymorphisms associated with affective disorders. J Circadian Rhythms. 7:1–10. Lachman HM, Pedrosa E, Petruolo OA, et al. (2007). Increase in GSK3beta gene copy number variation in bipolar disorder. Am J Med Genet B Neuropsychiatr Genet. 144B:259–65. Lazar AS, Slak A, Lo JC, et al. (2012). Sleep, diurnal preference, health, and psychological well-being: A prospective singleallelic-variation study. Chronobiol Int. 29:131–46. Le-Niculescu H, Patel SD, Bhat M, et al. (2009). Convergent functional genomics of genome-wide association data for bipolar disorder: Comprehensive identification of candidate genes, pathways and mechanisms. Am J Med Genet B Neuropsychiatr Genet. 150B:155–81. Leboyer M. (2003). Searching for alternative phenotypes in psychiatric genetics. Methods Mol Med. 77:145–61. Leboyer M, Kupfer DJ. (2010). Bipolar disorder: New perspectives in health care and prevention. J Clin Psychiatry. 71:1689–95. Lee KY, Song JY, Kim SH, et al. (2010). Association between CLOCK 3111T/C and preferred circadian phase in Korean patients with bipolar disorder. Prog Neuropsychopharmacol Biol Psychiatry. 34:1196–201. Lewy AJ, Nurnberger Jr JI, Wehr TA, et al. (1985). Supersensitivity to light: possible trait marker for manic-depressive illness. Am J Psychiatry. 142:725–7. !

Informa Healthcare USA, Inc.

7

Li X, Jope RS. (2010). Is glycogen synthase kinase-3 a central modulator in mood regulation? Neuropsychopharmacology. 35: 2143–54. Mansour HA, Talkowski ME, Wood J, et al. (2009). Association study of 21 circadian genes with bipolar I disorder, schizoaffective disorder, and schizophrenia. Bipolar Disord. 11:701–10. Mansour HA, Wood J, Chowdari KV, et al. (2005). Circadian phase variation in bipolar I disorder. Chronobiol Int. 22:571–84. Mansour HA, Wood J, Logue T, et al. (2006). Association study of eight circadian genes with bipolar I disorder, schizoaffective disorder and schizophrenia. Genes Brain Behav. 5:150–7. Maxwell ME. (1992). Family interview for genetic studies. Clinical Neurogenetics Branch, Intramural Research Program, NIMH. McCarthy MJ, Nievergelt CM, Kelsoe JR, Welsh DK. (2012). A survey of genomic studies supports association of circadian clock genes with bipolar disorder spectrum illnesses and lithium response. PLoS ONE. 7:e32091. McClung CA. (2013). How might circadian rhythms control mood? Let me count the ways. Biol Psychiatry. 74:242–9. McGrath CL, Glatt SJ, Sklar P, et al. (2009). Evidence for genetic association of RORB with bipolar disorder. BMC Psychiatry. 9: 70. Milhiet V, Boudebesse C, Bellivier F, et al. (2014). Circadian abnormalities as markers of susceptibility in bipolar disorders. Frontiers Biosci. 6:120–37. Milhiet V, Etain B, Boudebesse C, Bellivier F. (2011). Circadian biomarkers, circadian genes and bipolar disorders. J Physiol Paris. 105:183–9. Murray G, Harvey A. (2010). Circadian rhythms and sleep in bipolar disorder. Bipolar Disord. 12:459–72. Nievergelt CM, Kripke DF, Barrett TB, et al. (2006). Suggestive evidence for association of the circadian genes PERIOD3 and ARNTL with bipolar disorder. Am J Med Genet B Neuropsychiatr Genet. 141B:234–41. Nurnberger Jr JI, Adkins S, Lahiri DK, et al. (2000). Melatonin suppression by light in euthymic bipolar and unipolar patients. Arch Gen Psychiatry. 57:572–9. Nurnberger Jr JI, Berrettini W, Tamarkin L, et al. (1988). Supersensitivity to melatonin suppression by light in young people at high risk for affective disorder. A preliminary report. Neuropsychopharmacology. 1:217–23. Nurnberger Jr JI, Blehar MC, Kaufmann CA, et al. (1994). Diagnostic interview for genetic studies. Rationale, unique features, and training. NIMH Genetics Initiative. Arch Gen Psychiatry. 51:849–59; discussion 863–44. Pacchierotti C, Iapichino S, Bossini L, et al. (2001). Melatonin in psychiatric disorders: A review on the melatonin involvement in psychiatry. Front Neuroendocrinol. 22:18–32. Portaluppi F, Smolensky MH, Touitou Y. (2010). Ethics and methods for biological rhythm research on animals and human beings. Chronobiol Int. 27:1911–29. Preisig M, Fenton BT, Matthey ML, et al. (1999). Diagnostic interview for genetic studies (DIGS): Inter-rater and test-retest reliability of the French version. Eur Arch Psychiatry Clin Neurosci. 249:174–9. Purcell S, Neale B, Todd-Brown K, et al. (2007). PLINK: A tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 81:559–75. Rush AJ, Trivedi MH, Ibrahim HM, et al. (2003). The 16-Item Quick Inventory of Depressive Symptomatology (QIDS), clinician rating (QIDS-C), and self-report (QIDS-SR): A psychometric evaluation in patients with chronic major depression. Biol Psychiatry. 54:573–83. Sato TK, Panda S, Miraglia LJ, et al. (2004). A functional genomics strategy reveals Rora as a component of the mammalian circadian clock. Neuron. 43:527–37. Schulze TG. (2010). Genetic research into bipolar disorder: The need for a research framework that integrates sophisticated

Chronobiol Int Downloaded from informahealthcare.com by University Library Utrecht on 06/02/14 For personal use only.

8

B. Etain et al.

molecular biology and clinically informed phenotype characterization. Psychiatr Clin North Am. 33:67–82. Severino G, Manchia M, Contu P, et al. (2009). Association study in a Sardinian sample between bipolar disorder and the nuclear receptor REV-ERBalpha gene: A critical component of the circadian clock system. Bipolar Disord. 11: 215–20. Shi J, Wittke-Thompson JK, Badner JA, et al. (2008). Clock genes may influence bipolar disorder susceptibility and dysfunctional circadian rhythm. Am J Med Genet B Neuropsychiatr Genet. 147B:1047–55. Sklar P, Smoller JW, Fan J, et al. (2008). Whole-genome association study of bipolar disorder. Mol Psychiatry. 13:558–69. Smith CS, Reilly C, Midkiff K. (1989). Evaluation of three circadian rhythm questionnaires with suggestions for an improved measure of morningness. J Appl Psychol. 74:728–38. Soria V, Martinez-Amoros E, Escaramis G, et al. (2010). Differential association of circadian genes with mood disorders: CRY1 and NPAS2 are associated with unipolar major

depression and CLOCK and VIP with bipolar disorder. Neuropsychopharmacology. 35:1279–89. Tomioka K, Matsumoto A. (2010). A comparative view of insect circadian clock systems. Cell Mol Life Sci. 67:1397–406. Vink JM, Groot AS, Kerkhof GA, Boomsma DI. (2001). Genetic analysis of morningness and eveningness. Chronobiol Int. 18: 809–22. Waddington Lamont E, Coutu DL, Cermakian N, Boivin DB. (2010). Circadian rhythms and clock genes in psychotic disorders. Isr J Psychiatry Relat Sci. 47:27–35. Wellcome Trust Case Control Consortium. (2007). Genome-wide association study of 14,000 cases of seven common diseases and 3,000 shared controls. Nature. 447:661–78. Wood J, Birmaher B, Axelson D, et al. (2009). Replicable differences in preferred circadian phase between bipolar disorder patients and control individuals. Psychiatry Res. 166:201–9. Yang S, Van Dongen HP, Wang K, et al. (2009). Assessment of circadian function in fibroblasts of patients with bipolar disorder. Mol Psychiatry. 14:143–55.

Chronobiology International