Genetic contributions to circadian activity rhythm and sleep pattern phenotypes in pedigrees segregating for severe bipolar disorder Lucia Pagania, Patricia A. St. Clairb, Terri M. Teshibab, Susan K. Serviceb, Scott C. Fearsb, Carmen Arayac, Xinia Arayac, Julio Bejaranoc, Margarita Ramirezc, Gabriel Castrillónd, Juliana Gomez-Makhinsone, Maria C. Lopeze, Gabriel Montoyae, Claudia P. Montoyae, Ileana Aldanab, Linda Navarrob, Daniel G. Freimerb, Brian Safaieb, Lap-Woon Keungb, Kiefer Greenspanb, Katty Choub, Javier I. Escobarf, Jorge Ospina-Duquee, Barbara Kremeyerg, Andres Ruiz-Linaresg, Rita M. Cantorb, Carlos Lopez-Jaramilloe,h, Gabriel Macayac, Julio Molinai, Victor I. Reusj, Chiara Sabattik, Carrie E. Beardenb, Joseph S. Takahashia,l,1, and Nelson B. Freimerb,1 aDepartment of Neuroscience, University of Texas Southwestern Medical Center, Dallas, TX 75390; bDepartment of Psychiatry and Biobehavioral Sciences, University of California, Los Angeles, CA 90095; cCell and Molecular Biology Research Center, Universidad de Costa Rica, San Pedro de Montes de Oca, San José, Costa Rica 11501; dInstituto de Alta Tecnología Médica de Antioquia, Medellín, Colombia 050026; eGrupo de Investigación en Psiquiatría (Research Group in Psychiatry; GIPSI), Departamento de Psiquiatría Facultad de Medicina, Universidad de Antioquia, Medellín, Colombia 050011; fDepartment of Psychiatry and Family Medicine, Rutgers-Robert Wood Johnson Medical School, New Brunswick, NJ 08901; gDepartment of Genetics, Evolution and Environment, University College London, London WC1E 6BT, United Kingdom; hMood Disorders Program, Hospital San Vicente Fundacion, Medellín, Colombia 050011; iBioCiencias Lab, 01010 Guatemala, Guatemala; jDepartment of Psychiatry, University of California, San Francisco, CA 94143; kDepartment of Health Research and Policy, Division of Biostatistics, Stanford University, Stanford, CA 94305; and lHoward Hughes Medical Institute, University of Texas Southwestern Medical Center, Dallas, TX 75390 Contributed by Joseph S. Takahashi, November 24, 2015 (sent for review September 21, 2015; reviewed by Maja Bucan, Kathleen Ries Merikangas, and Emmanuel J. M. Mignot) Abnormalities in sleep and circadian rhythms are central features We report here the delineation of sleep and activity BP endo- of bipolar disorder (BP), often persisting between episodes. We report phenotypes through investigations of 26 pedigrees (n = 558) here, to our knowledge, the first systematic analysis of circadian ascertained for severe BP (BP-I), from the genetically related rhythm activity in pedigrees segregating severe BP (BP-I). By analyzing populations of the Central Valley of Costa Rica (CR) and Anti- actigraphy data obtained from members of 26 Costa Rican and oquia, Colombia (CO) (7–9). Pedigrees ascertained for multiple Colombian pedigrees [136 euthymic (i.e., interepisode) BP-I individuals cases of severe BP (BP-I) should be enriched for extreme and 422 non–BP-I relatives], we delineated 73 phenotypes, of which 49 demonstrated significant heritability and 13 showed significant values of quantitative traits that are BP endophenotypes, en- trait-like association with BP-I. All BP-I associated traits related to ac- hancing their utility for genetic mapping studies of such pheno-– tivity level, with BP-I individuals consistently demonstrating lower types. Additionally, such pedigrees derived from recently expanded activity levels than their non–BP-I relatives. We analyzed all 49 heri- table phenotypes using genetic linkage analysis, with special empha- Significance sis on phenotypes judged to have the strongest impact on the biology underlying BP. We identified a locus for interdaily stability of activity, Characterizing the abnormalities in sleep and activity that are at a threshold exceeding genome-wide significance, on chromosome associated with bipolar disorder (BP) and identifying their cau- 12pter, a region that also showed pleiotropic linkage to two addi- sation are key milestones in unraveling the biological underpin- tional activity phenotypes. nings of this severe and highly prevalent disorder. We have conducted the first systematic evaluation of sleep and activity bipolar disorder | endophenotypes | circadian rhythms | actigraphy | phenotypes in pedigrees that include multiple BP-affected behavior members. By delineating specific sleep and activity measures that are significantly heritable in these families, and those whose vari- Quantitative sleep and activity measures are hypothesized to ation correlated with the BP status of their members, and by de-be endophenotypes for bipolar disorder (BP). Disturbance termining the chromosomal position of loci contributing to many of sleep and circadian activity typically precedes and may pre- of these traits, we have taken the first step toward discovery of cipitate the initial onset of BP (1, 2). Decreased sleep and in- causative genetic variants. These variants, in turn, could provide creased activity occur before and during manic and hypomanic clues to new approaches for both preventing and treating BP. episodes. Conversely, increased sleep and decreased activity characterize BP–depression. Extreme diurnal variation in mood Author contributions: L.P., P.A.S.C., J.I.E., J.O.-D., B.K., A.R.-L., C.L.-J., G. Macaya, V.I.R., C.E.B., features prominently in both mania and depression, whereas shifts J.S.T., and N.B.F. designed research; L.P. performed research; T.M.T., C.A., X.A., J.B., M.R.,G.C., J.G.-M., M.C.L., G. Montoya, C.P.M., I.A., D.G.F., B.S., L.-W.K., K.G., K.C., and J.M. in circadian phase (the time within the daily activity cycle at which contributed new reagents/analytic tools; C.A., X.A., J.B., M.R., G.C., M.C.L., and C.P.M. periodic phenomena such as bed time or awakening occur) can interviewed patients and managed clinical databases; J.G.-M. and G. Montoya inter- induce mania and ameliorate symptoms of BP–depression (3). viewed patients and collected clinical data; I.A. managed databases and reviewed and Twin studies have identified multiple heritable sleep and ac- performed data quality control; T.M.T., D.G.F., B.S., L.-W.K., K.G., and K.C. reviewed filesand performed data quality control; J.M. collected clinical data; L.P., S.K.S., S.C.F., L.N., R.M.C., tivity phenotypes, including sleep duration, sleep quality, phase C.S., J.S.T., and N.B.F. analyzed data; and L.P., S.K.S., C.S., and N.B.F. wrote the paper. of activity preference and sleep pattern, and sleep architecture Reviewers: M.B., University of Pennsylvania; K.R.M., National Institutes of Health; and variables [e.g., the amount of slow wave and rapid eye movement E.J.M.M., Stanford University School of Medicine. (REM) sleep (4) and polysomnography profiles during non- The authors declare no conflict of interest. REM sleep (5)]. Euthymic BP individuals, compared with healthy Freely available online through the PNAS open access option. controls, display trait-like alterations in several such phenotypes— See Commentary on page 1477. for example, sleep time and time in bed, sleep onset latency, and 1To whom correspondence may be addressed. Email: joseph.takahashi@utsouthwestern. periods of being awake after sleep onset (6). However, no prior edu or nfreimer@mednet.ucla.edu. investigations have assayed the heritability of such phenotypes in This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. BP individuals and their relatives. 1073/pnas.1513525113/-/DCSupplemental. E754–E761 | PNAS | Published online December 28, 2015 www.pnas.org/cgi/doi/10.1073/pnas.1513525113 founder populations are likely to show increased frequencies for to the exclusion of 80 recordings (Fig. S2), we analyzed activity many deleterious alleles—another feature that will enhance their data from 558 individuals, including 136 BP-I individuals and 422 utility for mapping these traits (10, 11). We previously described, in of their non–BP-I relatives (Table 1). A series of algorithms these pedigrees, multiple heritable and BP-associated pheno- obtained from published sources (14–17) were then applied to types from the domains of temperament, neurocognition, and the activity data to obtain 116 quantitative sleep and activity neuroanatomy (10). phenotypes. These phenotypes can be classified into six broad Actigraphy (activity measurement using wrist-worn acceler- domains that quantified patterns of activity and sleep during the ometers) can be conducted over prolonged periods without im- major rest period of the day (i) and during the awake period (ii), pinging on an individual’s usual activities, enabling assessment of the fragmentation or consolidation of activity (iii), overall ac- sleep and activity on a scale sufficient for genetic investigations. tivity levels (iv), and the fit of daily activity patterns to curves Actigraphy data on sleep quality and duration correlate strongly based on sine and cosine functions (using two different ap- with those obtained through polysomnography, the gold standard proaches) (v and vi). Details on the construction of phenotypes for sleep research (12). Using actigraphy, one can estimate the that fall into each domain are provided in SI Methods. main circadian activity parameters, which follow a sinusoidal To reduce the multiple testing burden and eliminate completely waveform with a ∼24-h period: phase, amplitude (the strength of redundant variables, we calculated pair-wise correlations among circadian rhythms, as measured by the difference in the amount all 116 phenotypes (Fig. S3) and performed a hierarchical clus- of activity between active and inactive moments), and coherence tering analysis using 1 – correlation as the distance metric. We of the rhythm (the degree of consolidation and stability of ac- then selected one representative phenotype from each cluster, tivity, rest, and sleep). Finally, actigraphy enables quantification yielding 73 phenotypes (Fig. S3), which we analyzed further as of BP-associated features, such as fragmentation of rest and described below. The 43 phenotypes excluded at this stage were activity within and between days, that do not fit a sinusoidal very highly correlated with other phenotypes, all with r > 0.90 and waveform and cannot be analyzed parametrically (13). most with r > 0.99. We recorded activity in euthymic BP-I individuals and their non–BP-I relatives from the CR and CO pedigrees for, on av- Heritability of Phenotypes and Their Association to BP-I. Estimating the erage, 14 consecutive days. For each actigraphy phenotype, we familial aggregation of the 73 phenotypes (an indicator of herita- evaluated association with BP-I, assessed the heritability of each bility) and their relationship to BP-I allowed us to determine which trait, and performed genome-wide genetic linkage analyses on all phenotypes have a significant genetic component, to proceed with significantly heritable traits. analyses to identify genes contributing to phenotypes that are po- tentially important in the etiology of BP-I. We subjected each Results phenotype to an inverse-normal transformation, and to control for Through actigraphy, we obtained activity recordings (illustrated covariates, we regressed [in the software SOLAR (Sequential in Fig. S1) from 638 members of 26 CR and CO pedigrees. After Oligogenic Linkage Analysis Routines) (18)] the transformed applying quality control (QC) procedures (SI Methods) that led phenotypes on age, gender, and country. The residuals from this Table 1. Sample characteristics and recording days by country and family Family n (BP-I cases) Female Mean age (SD), range Mean recorded days (SD), range CO All 269 (66) 58% 46.3 (16.6), 18–83 15.1 (2.1), 7–24 CR All 290 (70) 53% 49.2 (16.0), 17–88 15.8 (2.9), 6–27 CO4 33 (7) 61% 42.0 (17.2), 18–76 15.1 (2.3), 13–24 CO7 96 (23) 53% 44.8 (17.0), 18–80 14.8 (1.6), 7–21 CO8 5 (2) 60% 40.8 (12.8), 24–54 14 (4.9), 7–21 CO10 22 (5) 73% 53.4 (15.4), 32–77 15.4 (2), 14–21 CO13 14 (4) 57% 42.1 (14.7), 18–66 17.1 (3.1), 14–21 CO14 13 (4) 46% 43.8 (15.4), 20–74 14.1 (2.4), 10–21 CO15 19 (4) 58% 41.7 (15.3), 19–73 14.5 (1.6), 12–20 CO18 20 (6) 55% 59.3 (14.2), 34–77 15.2 (1.9), 14–20 CO23 21 (6) 67% 45.0 (15.2), 18–83 15.3 (1.9), 14–20 CO25 8 (2) 63% 56.5 (13.5), 45–82 15.6 (2.6), 14–21 CO27 18 (3) 67% 46.7 (16.8), 18–74 15.4 (1.8), 14–21 CR001 5 (1) 60% 56.0 (11.3), 44–68 19.6 (2.6), 15–21 CR004 37 (7) 51% 55.6 (13.5), 30–84 15 (1.9), 12–21 CR006 7 (2) 29% 52.7 (11.8), 38–66 13.9 (4.1), 6–20 CR007 4 (2) 50% 47.0 (6.6), 39–55 13.8 (0.5), 13–14 CR008 9 (3) 44% 43.7 (16.1), 21–68 15.7 (2.5), 14–20 CR009 25 (6) 60% 40.6 (14.4), 20–74 15.6 (2.8), 12–22 CR010 12 (3) 58% 43.9 (15.9), 21–74 14.2 (2.2), 11–19 CR011 9 (2) 56% 48.8 (23.2), 21–86 18.2 (3), 14–21 CR012 19 (5) 68% 40.8 (15.3), 20–68 17 (3.2), 14–21 CR013 5 (2) 80% 52.2 (19.3), 35–74 15.4 (2.2), 14–19 CR014 2 (1) 50% 45.0 (2.8), 43–47 14.5 (0.7), 14–15 CR015 8 (1) 63% 51.6 (14.2), 38–71 18.4 (2.2), 16–21 CR016 13 (4) 38% 50.1 (14.5), 19–66 16.8 (2.6), 13–21 CR201 125 (28) 51% 50.6 (16.2), 17–88 15.8 (3.1), 7–27 CR277 9 (3) 56% 49.3 (11.6), 37–71 14.1 (0.3), 14–15 Grand total 558 (136) 56% 47.8 (16.3), 17–88 15.4 (2.6), 6–27 Pagani et al. PNAS | Published online December 28, 2015 | E755 GENETICS SEE COMMENTARY PNAS PLUS regression were assessed for heritability and for a mean difference BP-I subjects awoke later and slept longer than non–BP-I sub- between BP-I individuals and their non–BP-I relatives. jects (phenotypes, mean of sleep offset time and mean sleep Of the 73 phenotypes, 49 (67%) demonstrated significant duration). Outside of the rest period, BP-I individuals were, on heritability. Heritable phenotypes included measures related to average, awake fewer minutes than non–BP-I individuals (phe- sleep and activity duration, timing, fragmentation, and consoli- notypes, mean of awake duration and mean of total minutes dation; activity levels and variability; and the timing and peri- scores as awake) and had more variability in the time during the odicity of mean daily activity (Fig. 1). awake period scored as sleeping (phenotype, SD of the total Thirteen phenotypes (18%) were significantly associated with minutes scored as sleep). Similar to previous studies (16, 19), we BP-I, of which 12 (92%) were also heritable (Fig. 1 and Fig. S4). found that euthymic BP-I individuals display lower activity levels Sle Legend: ep p h2 estimate (inner histogram) hen Positive BP regression coefficient otyp (central histogram) es w Negative BP regression coefficient ithi (central histogram) n th Significantly heritable e s re Significantly associated to BP-I ing st ke n gth pe wa t le n ri f a o o bo u d . p gt h .500 o e en +/- 0 f n sle l 0.375 t t /- D o u + of bo p b ou 50 S an ep e t +/- 0 .2 6− e sle st s le u 1 M f ep bo 0.125/- 7 − D o ong e t sle+ 5 1 S l s t0.7 8− D o f hor te ke b ou 0 1 −S of s a 0.6 19 st w 45 Me an nge 0. 20− f lo bout an o e 0.30 −Me gest wak 21 0.15 D of lon ngth 2−S ake bo ut le 0 2 an of w e 23−M ut lengt h of wake bo 24−SD 01−Mean of awake duration 02−SD of awake duration 03−Mean of no. of 0 sl4 ee− p S boD uts o 0 f 5 n− o.M of e sa len ep o b0 f o6 u to ts−S 0 D ta o l m7− f in M to s sco 08 e tal m red as − an o inS f s s a c w o a09 D o mea r ke − ed 1 M f m n l as a 0 ea ea en ng w th a− n k1 1 SD of le o ef 1 − n aw 2 M o lon gth a − e fa lo ge on s f ke b SD n ge t aw oaw a u k t o o s a e f f t b t so h a o w ke o t b u a rt ak o t l e ut m est bin a os w ua t sc ko er e bd o ua ts sleep Fig. 1. Polar histogram of heritability traits and association to BP-I. The inner histogram represents the heritability estimate (h2) in yellow–green. The middle histogram represents the association to BP-I; positive associations with BP-I are presented in red (i.e., the trait has a higher value in those with BP-I), whereas negative associations are presented in blue (indicating those with BP-I have lower values on the trait). The outer histogram summarizes the heritable traits (black) and the phenotypes associated with BP-I (green). hrs, hours; mins, minutes; no., number. E756 | www.pnas.org/cgi/doi/10.1073/pnas.1513525113 Pagani et al. tting traits Four l fi ier f nd H il itt a i ninor g p os hC enotypes ffect BP-I E nset time leep o an of s time imat e 01−M e leep o nset 2 es t f s h 02−SD o of midsl eep time 03−SD 04−Mean of sleep offset time 05−SD of sleep offset time 06−Mean of sleep duration 07−SD of slee 0 p8 d− urM atioe na 0 n9 o− f sM lee epa o10 n n so et−S f s lal te eD e ncy 1 p 1 of s in− er 1 M le tia 2− e e a p S n in o er13− D ti fo s a M f le1 s ep 1 45 − ea ln eS e e o p ff − D f i e c M W f ie o fi n f A c c e iW S e y a nn A O cy of S n Oo. of awakenings activity measureme nts Rest- Overall activity level phenotypes igh t Mi dn to PM 6 om fr ivi ty ta ct er Be ma of a me t m D ra r Ga −S r p a tete 9 ino ra m s Co or pa − u m 1 osi n inim C m 2− osin or C met er m 3− para G 4−H ill mete r paraill 5−H eter Bm 6−Hill para eter minm 7−Hill pa ra 8−Hill acropha se 1−sin1 (period 24 hrs) 2−cos1 (period 24 3 hr− s)sin2 (p 4 e− rioc do 1s 22 h r( s) 5 p− ers ioi d 6 n3 1 − 2 c ( p he rso r ) io 7 s− 3s d ( i p 6 e hr rs8 n4 i− o ) c (p d 6 9 o− s eri ho r4 d sM ) i (n pim e 3 r i hrs u o )m d 3 hrs) Activity outside of the rest period gth t le n bo u h ep t e len g sl ut n o f bop ea lee ou t b of m of s n lee p an a t se m e ge s bou t −M3 D of 1 of lon eep −S l 4 an s t s 1 Me lon ge 5−1 D o f 16− S bility sta terd aily iabilit y 1−In y var −Intra dail 2 amplit ude ive 3−Rela t −Amplitud e 4 5−Lowest 5 hrs of activ ity 6−Onset phase of lowest 5 hrs of activity 7−Onset phase of highest 10 hrs of activity 1−Medi 2 an− oM f ae ca tivi3− n ty S of acti 4 D of vit− y M ac ft ri o5 ean v m i − t midn SD o y f f rom ight t 6 o− M o ac mf tivi id 6 a ty n AM 7− ea ctiv f ig ro htn 8 SD o i m tt oy 6 f −M o f a r 6 A Ao M M e fa a c c ti m vi t 6 tA on t 1 o iv y f M 2 f P i r ta ty om o M c ft ro 1 12iv m 2 P ity 1 P M Mf 2r o PM tom 6 6 to P MP 6M P t Mo Midnight than non–BP-I individuals, by multiple measures: amplitude; L5, (LOD = 3.35); lower interdaily stability indicates a weaker 5-h period of minimal activity; median activity; mean activity in rhythm, while a lower value of mean number of sleep bouts is 6-h windows; and two parameters related to the amplitude and associated to a more consolidated rhythm, leading to a negative phase of curve fits to mean daily activity (parameter B from the correlation between these variables. Amplitude and interdaily fit of the Hill transformation to the daily activity profile and stability display phenotypic and genotypic correlation of a Cosinor parameter Beta). similar magnitude, albeit in a positive direction (rhop = 0.51 and rhog = 0.92); their joint linkage analysis (LOD = 3.41) Linkage Analysis of Selected Heritable Traits. To identify genetic loci suggests near-complete pleiotropy between these phenotypes, at with the largest impact on sleep and activity phenotypes in the the same location. BP-I pedigrees, we conducted genome-wide linkage analysis using A second (nonsignificant) linkage peak on chromosome 1q a dense set of SNPs. Our primary analyses focused on 13 pheno- displayed LOD > 3.0 for four correlated phenotypes related to types that we considered most relevant to BP. These phenotypes the SD of sleep onset and activity (SD of the time of sleep onset, included traits with prior evidence of BP association, traits that LOD 3.9 at 183 Mb; SD of the active phase, LOD 3.6 at 185 Mb; showed strong BP-I association in our pedigrees, and traits bi- SD of the time of midsleep, LOD 4.3 at 186 Mb; SD of the total ologically relevant to fragmentation or consolidation of circa- minutes awake, LOD 3.4 at 186 Mb). These phenotypes all have a dian activity or with biological relevance to phase (Table 2). Using SOLAR, we performed multipoint linkage analysis on pair-wise genetic correlation >0.85, as is reflected in joint linkage these 13 phenotypes. The strongest linkage was for the rest results suggesting complete pleiotropy among them (joint LOD– activity phenotype Interdaily Stability (IS), that represents the scores range from 3.1 to 4.2 at 185 Mb; Table S2). degree of variability of activity level on an hourly basis from day- Discussion to-day (Fig. 2), for which we observed a maximum LOD [loga- rithm (base 10) of odds] score (4.73, 4 Mb from chromosome We report here, to our knowledge, the first large-scale delineation 12pter), exceeding the traditional genome-wide significance of sleep and activity phenotypes in BP-affected individuals and threshold (P < 10−4). Interdaily stability linkage remains genome- their relatives. More generally, it is the first genetic investigation wide significant at the 0.05 alpha level, after correcting for the of such a comprehensive set of sleep and circadian measures in 13 phenotypes (empirical P values associated to the highest any human study. LOD score and to the smallest Simes P value are 0.03 and 0.02, Phenotypes significantly associated to BP-I paint a consistent’ respectively). The linkage evidence for interdaily stability di- picture; activity is lower in euthymic BP-I individuals than in their minishes only slightly when including BP-I status as a covariate non–BP-I relatives, reflecting a longer sleep duration, with a later (LOD = 4.65; Table S1). time of sleep offset and rest offset, resulting in a shorter duration Fig. 3 presents a summary of the significance of linkage results of the active phase. Furthermore, during the active phase, BP-I considering all 49 heritable phenotypes (20). Two linkage peaks individuals have fewer total minutes scored as awake and more stand out. The largest peak includes the interdaily stability variability in the total minutes scored as asleep. Such individuals linkage on chromosome 12pter, having a Simes P value < 0.0001. also display lower amplitude, mainly due to their lower activity In the same region of linkage to interdaily stability, we observed level during the least active hours of the day. suggestive linkage for two additional phenotypes (Fig. 4): the As the BP-I individuals who participated in the study were all mean number of sleep bouts in the awake period, and amplitude euthymic at the time of recording, the phenotypes that we ob- (the difference in activity between the 5 most active hours and served could be considered representative of the remission phase the 10 least active hours), with peak LOD scores, respectively, of of the disorder. Previous studies have suggested that distur- 2.53 (8 Mb from pter) and 2.13 (1 Mb from pter). Interdaily sta- bances in sleep and circadian activity are early signs of manic bility and mean number of sleep bouts show moderately negative episodes, particularly in individuals affected with rapid-cycling phenotypic correlation (rhop = –0.51) but strongly negative ge- forms of BP (1, 2). We did not observe a high rate of subsequent netic correlation (rhog = –0.93); joint multipoint linkage analysis mania among those BP-I individuals in whom we detected the suggests complete pleiotropy, indicating a common genetic most extreme sleep and activity phenotypes; however, it is pos- component to both phenotypes, centered 4 Mb from 12pter sible that we may have missed such a relationship given that we Table 2. Thirteen phenotypes chosen for the primary linkage analysis Phenotype Trait Importance Amplitude Relative amplitude BP-I associated (19) Amplitude Association with BP-I in the present study Median of activity level Association with BP-I in the present study Phase Time of sleep onset BP-I associated (17) Time of sleep offset BP-I associated (17) Hill acrophase Biologically relevant for phase Fragmentation/ consolidation IV, Intradaily Variability in activity BP-I associated (17) IS, Interdaily Stability in activity BP-I associated (17) Mean of the number of sleep bouts Biologically relevant to during the awake period fragmentation/consolidation Mean of the length of sleep bouts Biologically relevant to during the sleep period fragmentation/consolidation Efficiency of sleep WASO, total minutes in awake bouts BP-I associated (24) after sleep onset Mean of awake duration Association with BP-I in the present study Mean total minutes scored as awake Association with BP-I in the present study during the awake period Pagani et al. PNAS | Published online December 28, 2015 | E757 GENETICS SEE COMMENTARY PNAS PLUS Fig. 2. Representation of the multipoint LOD scores on different chromosomes (which are depicted in alternating violet and black) for 13 traits that we considered of highest biological significance. Blue line indicates an LOD score of 3.3, corresponding to a P value of 5 × 10−5. Mins, minutes; no., number. did not systematically monitor clinical state after the 2-wk re- agents. When subjects receiving neuroleptics were removed from cording period. the analysis, however, activity levels remained significantly lower Our observation of longer sleep duration in BP-I individuals in BP-I subjects for the majority of variables (Table S1). Anti- accords with the results of a meta-analysis conducted in euthymic depressant medication and lithium treatment appeared to have BP cases compared with controls (6). However, although the meta- no significant effect on the group differences except for one analysis observed a difference between cases and controls in sleep variable, mean activity from midnight to 6:00 AM, which was onset latency, we did not observe such a difference between BP-I lower in subjects taking lithium. The finding of decreased activity individuals and their non–BP-I relatives. Our study differs from the in BP-I subjects also remained significant, when lithium-treated previous investigation in two ways: First, whereas it compared BP patients were removed from the analysis. cases (defined broadly) to normal controls, we compared BP cases The size and composition of the pedigree set enabled us to (defined narrowly) with participants defined only by the absence of identify significant heritability for most measures that we evalu- BP-I. Second, our comparisons involved close relatives rather than ated. This finding encouraged us to conduct, to our knowledge, independent participants. the first genome-wide mapping study of quantitative traits repre- We evaluated the possible effects of medication use on the senting the most important features of human circadian behavior: association of activity phenotypes to BP-I. Twelve variables, in- phase, amplitude, and rhythm coherence or robustness. cluding several measuring mean activity levels, were lower for Previous studies of rare, autosomal dominant circadian rhythm BP-I subjects on neuroleptics than for those not prescribed these disorders have implicated genes [including PER2 (PERIOD2), E758 | www.pnas.org/cgi/doi/10.1073/pnas.1513525113 Pagani et al. influenced our results. An alternative study design might have been to analyze sleep and circadian parameters in traditional laboratory conditions, such as constant routine or forced desyn- chrony (33). We did not, however, consider such a design feasible. Not only were we concerned that exposure of BP-I–affected in- dividuals to such conditions might trigger an acute episode, but such laboratory studies are extremely expensive and labor- intensive, making the recording and analysis of circadian activity and sleep patterns in such a big cohort virtually impossible. To counter the limitations noted above, future investigations of circadian rhythms in these pedigrees will use molecular assays that record circadian rhythms in skin fibroblasts, from affected and unaffected individuals, in real time for several days. From this analysis, we will obtain information on the period length of Fig. 3. Multipoint linkage analysis across all 49 heritable phenotypes. Chromosomes are represented in alternating colors (light gray dark gray). the cells as well as parameters such as amplitude, phase, and– entrainment. Compared with behavioral phenotypes, circadian phenotypes resulting from this analysis will more directly reflect CK1δ (Casein kinase 1 delta), and DEC2 (BHLHE41, basic the underlying genetic properties of the clock (34, 35). helix-loop-helix family member e41)] already known to function in In summary, this is the first large-scale analysis of activity the regulation of the circadian clock (21–23). In contrast, the phenotypes in pedigrees ascertained for BP. We demonstrate linkage regions identified here for quantitative activity traits do lower activity in euthymic BP-I individuals compared with their not include any known clock genes. Similarly, prior work has non–BP-I relatives and heritability for phenotypes assaying mul- found little evidence that such genes play a role in quantitative tiple facets of sleep and activity. The genome-wide significant circadian activity phenotypes in mice (24). linkage to interdaily stability, a phenotype associated with BP in Our most striking genetic finding was the genome-wide sig- case-control studies (17), provides an opportunity to identify se- nificant linkage for interdaily stability, a measure of day-to-day quence variants contributing to the biological underpinnings of variability of the waveform of activity, near chromosome 12pter. this disorder. The suggestive linkage peaks in this region for two additional phenotypes, amplitude and mean number of sleep bouts, which Methods show pleiotropy with interdaily stability in bivariate linkage analy- Activity Recording Procedures. We used the Actiwatch Spectrum (Philips sis, underline its importance for the regulation of circadian activity. Respironics) to record activity count (in 1-min epochs) and ambient light level.At the time of purchase and after annual servicing and battery replacement, Several genes in this region could plausibly influence activity- we performed two independent procedures, to calibrate devices and mini- related behaviors, including the histone lysine demethylase mize interdevice variability (SI Methods). Project staff in CR and CO provided JARID1a (KMD5A) and calcium channel subunit 1C (CACNA1C). calibrated Actiwatches to all participants, whom we ascertained as reported JARID1a forms a complex with the core clock proteins CLOCK previously (10) and who provided informed consent, as approved by US and and BMAL1, thereby recruiting them to the Per2 promoter. On this local Institutional Review Boards (University of California–Los Angeles promoter, JARID1a enhances Per2 transcription through a de- Medical Institutional Review Board, the Ethics Committees of the University methylase-independent mechanism. Depletion of JARID1a, in of Costa Rica, the Ethics Committees of the University of Antioquia, and mammalian cells in culture, shortens the circadian period (25). University of Texas Southwestern Medical Center Institutional ReviewBoard). As close as possible to the time of other phenotypic assessments, we CACNA1C demonstrates a circadian expression pattern. Its placed the Actiwatch on the nondominant wrist and instructed participants loss affects the ability to phase advance wheel running behavior to not remove it for 14 d, press the marker button when they lay down to in mouse after a light pulse and impairs the induction of Per2 and Per1 expression (26). In multiple GWASs (genome-wide associ- ation studies), CACNA1C has shown genome-wide significant associations to BP (27–29) as well as to other psychiatric dis- orders (27). Studies in two cohorts have suggested a role of variants in CACNA1C in sleep habits and insomnia (30, 31), and CACNA1C knockout mice display lower EEG spectral power and impaired REM sleep recovery (32). These diverse GWAS findings, together with the pleiotropic effects that we observed, and the moderate association in our dataset between interdaily stability and BP-I, suggest that variants in this region could have a complex phenotypic impact beyond BP; however, SNPs in CACNA1C known to be associated to BP (30, 31) are not associated to interdaily stability and when included as covariates in our linkage analysis do not decrease our evidence for linkage to chromosome 12. Whole-genome sequencing un- derway in these pedigrees will enable us to evaluate, in relation to 12pter-linked sleep and activity phenotypes, variants in the genes noted above as well as other genes in this region. Although the study demonstrates the feasibility of large-scale genetic investigation of human circadian activity, its limitations reflect the imprecision of actigraphy as a representation of cir- cadian rhythm. By conducting the recordings while individuals were carrying out their usual activities, we were unable to exclude Fig. 4. Multipoint linkage analysis of the most biological significant phe- the possibility that various masking and confounding factors, such notypes; depiction of the region of pleiotropic linkage on chromosome 12p. as social and natural/artificial light entrainment, could have no., number. Pagani et al. PNAS | Published online December 28, 2015 | E759 GENETICS SEE COMMENTARY PNAS PLUS sleep and when they got out of bed, and keep a sleep log, registering bed Genome-Wide Genotyping and Quantitative Trait Linkage Analysis. Genotyping times and nap times. of 856 individuals, performed in three batches, used the Illumina Omni 2.5 chip (for QC details, see SI Methods). We implemented genome-wide mul- Activity Data Analysis. Two research assistants visually inspected activity re- tipoint linkage analysis in SOLAR, which uses a variance component cordings (Fig. S1) for gross abnormalities or deficiencies in data collection approach to partition the genetic covariance between relatives for each trait that would exclude them from analyses (SI Methods). For acceptable re- into locus-specific heritability (h2q) and residual genetic heritability (h2r). cordings, we first delineated the rest period for each 24 h (the interval from The software package Loki (39, 40), which implements Markov Chain Monte the time an individual gets into bed until he/she gets out of bed), combining Carlo, provided estimated multipoint identical by decent (MIBD) allele- actigraphy data with information from sleep logs using an algorithm written sharing among family members from genotype data, using a linkage dis- in R (36) (SI Methods and Figs. S5 and S6). We analyzed sleep parameters equilibrium (LD)-pruned subset of markers that passed QC procedures (SI using a script written in R, based on published Respironics definitions and Methods). We performed linkage analysis at 1-cM intervals focusing pri- algorithms (14–17). R scripts are available upon request. marily on phenotypes that we considered most relevant to BP; for these phenotypes, we also performed a secondary genome-wide linkage analysis Analysis of Heritability and Association to BP-I. As in the previous endophe- including BP status as a covariate. There were 558 individuals with genotype notyope analyses of these pedigrees (10), we analyzed heritability and asso- and phenotype data for all analyses (136 BP-I and 422 non–BP-I). Power ciation to BP-I in SOLAR (18), which implements a variance component method analysis in SOLAR, using simulated data, indicated that this sample size pro- to estimate the proportion of phenotypic variance due to additive genetic vided >80% power to detect a LOD score of 3, provided the estimate of locus- factors (narrow sense heritability). Under the null hypothesis that the value of specific heritability was ≥35%. the additive genetic variance is zero, testing significance of the estimate of To evaluate the significance of the strongest linkage finding while ap- genetic variance compared with the null value is a one-sided test. propriately accounting for multiple comparisons among phenotypes most Variance components analysis is sensitive to outliers and nonnormal trait relevant to BP, we used simulations. Specifically, we used gene dropping to distributions. To guard against statistical artifacts induced by skewed distri- generate genotypes consistent with the relationships between the pedigree butions, before analyses we used, in SOLAR, a standard rank-based procedure members, independent from the recorded phenotypic values. This approach (37) to inverse-normal transform all phenotypes and thereby avoid correla- allowed us to simulate null datasets that maintain the specific dependency tions between relatives or inflated heritability estimates (38). We regressed all phenotypes on three covariates [sex, age, and country structure existing across these 13 phenotypes. We used gene-drop simulations (CR vs. CO), also used in all analyses described below]; of six potential to construct a null distribution of LOD scores, rather than permutingphenotypic covariates that we evaluated, only these three significantly affected phe- values, because individuals in pedigrees are not exchangeable; phenotypic notype values (SI Methods). We initially considered household effects as a permutation would likely render most of our phenotypes nonheritable, potential source of phenotypic variation; however, we found that “house- therefore biasing the null distribution toward zero LOD scores. For each of 100 hold” was not a significant component of variation for any phenotype and datasets so generated, we carried out linkage analysis for each phenotype, therefore did not include it as a variable in further models. We implemented recording both the highest LOD score obtained and themost significant P value regressions in SOLAR, using pedigree structures, using residuals from these for the hypothesis of no linkage to any phenotype (Simes P value). models in all further analyses. We tested for BP-I association (difference in trait means between individuals with and without this diagnosis), using ACKNOWLEDGMENTS. We thank the members of CR and CO families for SOLAR to account for dependencies among relatives in a two-sided test of participating in this study and John Blangero and Thomas Dyer (Texas the null hypothesis of no association. As the non–BP-I category includes in- Biomedical Research Institute and University of Texas Health Science dividuals diagnosed with other psychiatric disorders, this is not a case-control Center) for calculating MIBDs. This research was supported by NationalInstitute of Health Grants R01MH075007, R01MH095454, and P30NS062691 (to comparison (and likely underestimates the degree of BP-I association of each N.B.F.), T32MH073526 (to P.A.S.C.), K23MH074644-01 (to C.E.B.), and K08MH086786 measure). We controlled family-wise error rate at the 0.05 level, using a (to S.C.F.), the Colciencias and Codi-University of Antioquia (to C.L.-J.), and the Bonferroni-corrected threshold for each test (heritability and BP-I associa- Joanne and George Miller Family Endowed Term Chair (to C.E.B.). J.S.T. is an tion; P < 6.7 × 10−4). investigator in the Howard Hughes Medical Institute. 1. Jackson A, Cavanagh J, Scott J (2003) A systematic review of manic and depressive 18. Almasy L, Blangero J (1998) Multipoint quantitative-trait linkage analysis in general prodromes. J Affect Disord 74(3):209–217. pedigrees. Am J Hum Genet 62(5):1198–1211. 2. Leibenluft E, Albert PS, Rosenthal NE, Wehr TA (1996) Relationship between sleep and 19. Harvey AG, Schmidt DA, Scarnà A, Semler CN, Goodwin GM (2005) Sleep-related mood in patients with rapid-cycling bipolar disorder. Psychiatry Res 63(2-3):161–168. functioning in euthymic patients with bipolar disorder, patients with insomnia, and 3. Benedetti F, Barbini B, Colombo C, Smeraldi E (2007) Chronotherapeutics in a psy- subjects without sleep problems. Am J Psychiatry 162(1):50–57. chiatric ward. Sleep Med Rev 11(6):509–522. 20. Simes RJ (1986) An improved Bonferroni procedure for multiple tests of significance. 4. Linkowski P (1999) EEG sleep patterns in twins. J Sleep Res 8(Suppl 1):11–13. Biometrika 73(3):751–754. 5. De Gennaro L, et al. (2008) The electroencephalographic fingerprint of sleep is ge- 21. Xu Y, et al. (2005) Functional consequences of a CKIdelta mutation causing familial netically determined: A twin study. Ann Neurol 64(4):455–460. advanced sleep phase syndrome. Nature 434(7033):640–644. 6. Ng TH, et al. (2014) Sleep-wake disturbance in interepisode bipolar disorder and high- 22. Xu Y, et al. (2007) Modeling of a human circadian mutation yields insights into clock risk individuals: A systematic review and meta-analysis. Sleep Med Rev 20:46–58. regulation by PER2. Cell 128(1):59–70. 7. Carvajal-Carmona LG, et al. (2003) Genetic demography of Antioquia (Colombia) and 23. He Y, et al. (2009) The transcriptional repressor DEC2 regulates sleep length in the Central Valley of Costa Rica. Hum Genet 112(5-6):534–541. mammals. Science 325(5942):866–870. 8. Service S, et al. (2006) Results of a SNP genome screen in a large Costa Rican pedigree 24. Shimomura K, et al. (2001) Genome-wide epistatic interaction analysis reveals com- segregating for severe bipolar disorder. Am J Med Genet B Neuropsychiatr Genet plex genetic determinants of circadian behavior in mice. Genome Res 11(6):959–980. 141B(4):367–373. 25. DiTacchio L, et al. (2011) Histone lysine demethylase JARID1a activates CLOCK-BMAL1 9. Reich D, et al. (2012) Reconstructing Native American population history. Nature and influences the circadian clock. Science 333(6051):1881–1885. 488(7411):370–374. 26. Schmutz I, et al. (2014) A specific role for the REV-ERBα-controlled L-Type Voltage- 10. Fears SC, et al. (2014) Multisystem component phenotypes of bipolar disorder for Gated Calcium Channel CaV1.2 in resetting the circadian clock in the late night. J Biol genetic investigations of extended pedigrees. JAMA Psychiatry 71(4):375–387. Rhythms 29(4):288–298. 11. Stoll G, et al. (2013) Deletion of TOP3β, a component of FMRP-containing mRNPs, 27. Cross-Disorder Group of the Psychiatric Genomics Consortium (2013) Identification of contributes to neurodevelopmental disorders. Nat Neurosci 16(9):1228–1237. risk loci with shared effects on five major psychiatric disorders: A genome-wide 12. Acebo C, LeBourgeois MK (2006) Actigraphy. Respir Care Clin N Am 12(1):23–30, viii. analysis. Lancet 381(9875):1371–1379. 13. Witting W, Kwa IH, Eikelenboom P, Mirmiran M, Swaab DF (1990) Alterations in the cir- 28. Ferreira MA, et al.; Wellcome Trust Case Control Consortium (2008) Collaborative cadian rest-activity rhythm in aging and Alzheimer’s disease. Biol Psychiatry 27(6):563–572. genome-wide association analysis supports a role for ANK3 and CACNA1C in bipolar 14. Marler MR, Gehrman P, Martin JL, Ancoli-Israel S (2006) The sigmoidally transformed disorder. Nat Genet 40(9):1056–1058. cosine curve: A mathematical model for circadian rhythms with symmetric non- 29. Psychiatric GCBDWG; Psychiatric GWAS Consortium Bipolar Disorder Working Group sinusoidal shapes. Stat Med 25(22):3893–3904. (2011) Large-scale genome-wide association analysis of bipolar disorder identifies a 15. Wang J, et al. (2011) Measuring the impact of apnea and obesity on circadian activity new susceptibility locus near ODZ4. Nat Genet 43(10):977–983. patterns using functional linear modeling of actigraphy data. J Circadian Rhythms 9(1):11. 30. Byrne EM, et al. (2013) A genome-wide association study of sleep habits and in- 16. Salvatore P, et al. (2008) Circadian activity rhythm abnormalities in ill and recovered somnia. Am J Med Genet B Neuropsychiatr Genet 162B(5):439–451. bipolar I disorder patients. Bipolar Disord 10(2):256–265. 31. Parsons MJ, et al. (2013) Replication of genome-wide association studies (GWAS) loci 17. Jones SH, Hare DJ, Evershed K (2005) Actigraphic assessment of circadian activity and for sleep in the British G1219 cohort. Am J Med Genet B Neuropsychiatr Genet sleep patterns in bipolar disorder. Bipolar Disord 7(2):176–186. 162B(5):431–438. E760 | www.pnas.org/cgi/doi/10.1073/pnas.1513525113 Pagani et al. 32. Kumar D, et al. (2015) Ca1.2 modulates electroencephalographic rhythm and rapid 39. Heath SC (1997) Markov chain Monte Carlo segregation and linkage analysis for eye movement sleep recovery. Sleep 38(9):1371–1380. oligogenic models. Am J Hum Genet 61(3):748–760. 33. Nováková M, Sumová A (2014) New methods to assess circadian clocks in humans. 40. Heath SC, Snow GL, Thompson EA, Tseng C, Wijsman EM (1997) MCMC segregation Indian J Exp Biol 52(5):404–412. and linkage analysis. Genet Epidemiol 14(6):1011–1016. 34. Pagani L, et al. (2010) The physiological period length of the human circadian clock in 41. Jones SH, Tai S, Evershed K, Knowles R, Bentall R (2006) Early detection of bipolar vivo is directly proportional to period in human fibroblasts. PLoS One 5(10):e13376. disorder: A pilot familial high-risk study of parents with bipolar disorder and their 35. Yang S, Van Dongen HP, Wang K, Berrettini W, Bucan M (2009) Assessment of circadian adolescent children. Bipolar Disord 8(4):362–372. function in fibroblasts of patients with bipolar disorder. Mol Psychiatry 14(2):143–155. 42. Ankers D, Jones SH (2009) Objective assessment of circadian activity and sleep pat- 36. R Development Core Team (2014) R: A Language and Environment for Statistical terns in individuals at behavioural risk of hypomania. J Clin Psychol 65(10):1071–1086. Computing (R Foundation for Statistical Computing, Vienna, Austria). 43. Van Someren EJ, et al. (1999) Bright light therapy: Improved sensitivity to its effects 37. van der Waerden BL (1952) Order tests for the two-sample problem and their power. on rest-activity rhythms in Alzheimer patients by application of nonparametric Indag Math 14:453–458. methods. Chronobiol Int 16(4):505–518. 38. Pilia G, et al. (2006) Heritability of cardiovascular and personality traits in 6,148 Sar- 44. Nelson W, Tong YL, Lee JK, Halberg F (1979) Methods for cosinor-rhythmometry. dinians. PLoS Genet 2(8):e132. Chronobiologia 6(4):305–323. Pagani et al. PNAS | Published online December 28, 2015 | E761 GENETICS SEE COMMENTARY PNAS PLUS