Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2014 Oct;11(4):345-62.
doi: 10.4306/pi.2014.11.4.345. Epub 2014 Oct 20.

Circadian polymorphisms in night owls, in bipolars, and in non-24-hour sleep cycles

Daniel F Kripke  1 , Walter T Klimecki  2 , Caroline M Nievergelt  3 , Katharine M Rex  3 , Sarah S Murray  4 , Tatyana Shekhtman  5 , Gregory J Tranah  6 , Richard T Loving  7 , Heon-Jeong Lee  8 , Min Kyu Rhee  9 , Farhad F Shadan  10 , J Steven Poceta  10 , Shazia M Jamil  10 , Lawrence E Kline  10 , John R Kelsoe  5
Affiliations

Circadian polymorphisms in night owls, in bipolars, and in non-24-hour sleep cycles

Daniel F Kripke et al. Psychiatry Investig. 2014 Oct.

Abstract

People called night owls habitually have late bedtimes and late times of arising, sometimes suffering a heritable circadian disturbance called delayed sleep phase syndrome (DSPS). Those with DSPS, those with more severe progressively-late non-24-hour sleep-wake cycles, and those with bipolar disorder may share genetic tendencies for slowed or delayed circadian cycles. We searched for polymorphisms associated with DSPS in a case-control study of DSPS research participants and a separate study of Sleep Center patients undergoing polysomnography. In 45 participants, we resequenced portions of 15 circadian genes to identify unknown polymorphisms that might be associated with DSPS, non-24-hour rhythms, or bipolar comorbidities. We then genotyped single nucleotide polymorphisms (SNPs) in both larger samples, using Illumina Golden Gate assays. Associations of SNPs with the DSPS phenotype and with the morningness-eveningness parametric phenotype were computed for both samples, then combined for meta-analyses. Delayed sleep and "eveningness" were inversely associated with loci in circadian genes NFIL3 (rs2482705) and RORC (rs3828057). A group of haplotypes overlapping BHLHE40 was associated with non-24-hour sleep-wake cycles, and less robustly, with delayed sleep and bipolar disorder (e.g., rs34883305, rs34870629, rs74439275, and rs3750275 were associated with n=37, p=4.58E-09, Bonferroni p=2.95E-06). Bright light and melatonin can palliate circadian disorders, and genetics may clarify the underlying circadian photoperiodic mechanisms. After further replication and identification of the causal polymorphisms, these findings may point to future treatments for DSPS, non-24-hour rhythms, and possibly bipolar disorder or depression.

Keywords: BHLHE40; Bipolar disorder; Delayed sleep phase syndrome; NFIL3; Non-24 hour sleep-wake disorder; RORC.

PubMed Disclaimer

Figures

Figure 1
Figure 1
Flow chart of participant selection for the meta-analyses. Selection of samples for the main meta-analysis associating DSPS phenotype with SNPs is diagrammed. At progressive stages, samples were excluded if the DSPS phenotype determination was missing or ambiguous (e.g., a control volunteer who was rated possible DSPS). Because logistic regressions of the mixed-ancestry groups demonstrated excessive genomic inflation, those of non-European ancestry or uncertain ancestry were excluded from the meta-analysis. Samples were also dropped if covariate data were missing or if the genotype call for a particular SNP failed quality control. However, in supplementary analyses, we considered those with non-European ancestry (a mixture of African, Asian, Pacific Island, and Native American ancestry and mixed ancestries) as a group, and we considered all participants regardless of ancestry. DNA selections were similar for the DSPS-case-control phenotype and for the BALM quantitative phenotype.
Figure 2
Figure 2
Patterns of associated BHLHE40 SNPs with phenotypes. Rows represent each of the 45 participants resequenced (8 bipolars and 37 DSPS), and columns describe the SNPs. Position is the chromosome-3 base position from assembly GRCH37/hg19. Alleles list the minor allele first, as indicated in the forward DNA strand, and MAF shows the minor allele frequency listed in the UCSC browser 7/6/2013. For the associated SNP pattern: Red shading highlights a minor allele homozygote, and pink shading highlights a minor allele heterozygote for the associated pattern of alleles. Green shading highlights a SNP that was not inferred to be one of the 18 interlinked polymorphisms. Purple highlights the associated allele group (top of columns) and participants with ≥2 such alleles (left columns). *for specification of SNPs without "rs" designations (not appearing in dbSNP) see http://www.ncbi.nlm.nih.gov/nuccore/EF015895, **regarding non-24-hour sleep-wake components or bipolar disorder, symbols in columns 2-3 for a participant indicate. P: possible, L: likely, D: definite, or ?: undetermined.
Figure 3
Figure 3
Linkage disequilibrium of BHLHE40 by ancestry. Linkage (r2) of BHLHE40 SNP rs11130215 is shown as an example for the haplotype group, compiled and edited from SNAP plots for CEU, CHBJPT, and YRI ancestry groups within the 1000 Genome Pilot 1 data base (hg18 locations). The range plotted is from 25K 5' to the SNP (just at the limit of CEU significant linkage) to 25K 3' to the SNP. The sample with CEU ancestry had far stronger 5' linkage than the Asian or YRI samples, but linkage dropped sharply within a few thousand nucleotides downstream of the 3' UTR in all groups. BHLHE40 is a small gene (green arrow). Its protein binds to E-box transcription factor binding sites of several circadian genes, deactivating promoter activity by competitive inhibition., BHLHE40 forms heterodimers with BHLHE41. BHLHE40 transcription is enhanced by light pulses, probably due to a CRE-responsive element in its promoter. According to the Scripps Genome Adviser, several of the BHLHE40-associated alleles may influence transcription binding sites of the adjacent gene.
Figure 4
Figure 4
Actigraphic record provides an example of a non-24-hour sleep-wake rhythm. This patient wore a wrist activity-light monitor for 4 weeks (starting and ending about 2 PM). On each day (consecutive horizontal lines), the red vertical bars show relative wrist activity, indicating wakefulness. Purple rectangles highlight intervals of predominant inactivity, inferred as sleep with assistance of the patient's sleep log. Vertical yellow bars display the patient's illumination surroundings, plotted on a semi-logarithmic scale of lux (from 1 lux, e.g., full moonlight to 10,000 lux, e.g., bright sunlight.) Note that on many days, the patient experienced little or no daylight. This patient habitually retired to bed an hour or two later each day and awakened correspondingly later (25.4 hr. cycle.) His sleep drifted around the clock, with only minimal pattern irregularities suggesting interactions with the environment. Brief intervals when the actigraph was removed for bathing were not edited.
Figure 5
Figure 5
Q-Q plot for the meta-analysis of the DSPS phenotype. This quantile-quantile plot shows SNP association p values from the meta-analysis of the DSPS phenotype in the combined DSPS-case-control and Sleep Center data sets. See the first worksheet of Supplement 1 for a listing of these p values. Quantile-quantile plots can be useful in visualizing the import of SNPs associations by comparing the distributions of nominal p values with the random expectation, and they are useful in appraising possible confounding from sample stratification. The ordinate plots the observed probability (P) values on a minus LOG10 scale, and the abscissa represents the expected probabilities of a random distribution on a comparable scale. The dashed red line highlights the random expectation of P statistic distributions. Pink circles highlight the two Bonferroni-significant SNPs. Note that the next 5-10 nominally-significant SNPs were sufficiently elevated above the expectation to suggest possible association.
Figure 6
Figure 6
Linkage disequilibrium of NFIL3 by ancestry. Linkage (r2) of NFIL3 rs2482705 is shown, compiled and edited from SNAP plots for CEU, CHBJPT, and YRI ancestry groups within the 1000 Genome Pilot 1 data base. Linkage with r2 >0.2 ranges in the 3 ancestry groups from approximately 100K 5' to rs2482705 to 100K 3' downstream to rs2482705. There is considerable variation in the linkage patterns in the different ancestry groups, which may be important, considering that the association of rs2482705 with the DSPS phenotype and the BALM in our samples of European ancestry were not supported among non-Europeans. Haploview lists 6 SNPs (one 5', four intronic, and one 3') with which rs2482705 is strongly linked in a sample of European ancestry (not counting rs33973463). Rs33973463, to which rs2482705 is well-linked (Spearman rho=0.669), is a synonymous coding SNP which may create an ESE splice promotion site. NFIL3 is a competitive inhibitor at D-box transcription factor binding sites. We have not discerned a specific role for rs2482705, an intronic SNP. It is plausible that one or more of the linked SNPs are responsible for the association of rs2482705 with DSPS.
Figure 7
Figure 7
Linkage disequilibrium of RORC by ancestry. Linkage (r2) of RORC rs3828057 is shown, compiled and edited from SNAP plots for CEU, CHBJPT, and YRI ancestry groups within the 1000 Genome Pilot 1 data base. Linkage with r2 >0.4 ranges in the 3 ancestry groups from the 5' part of RORC to SNPs and genes about 40K downstream to rs3828057. Green arrows show genes in the region with their direction of transcription. RORC binds at RRE transcription factor binding sites on CLOCK, ARNTL, NFIL3, and RORC itself, among others, promoting transcription with complex influences on circadian feedback loops. The roles of RORC in immunology have attracted the most interest. No specific role of rs3828057 has been discerned, and considering the substantial linkage to several adjacent genes, further investigation will be needed to confirm that RORC actively influences DSPS.
Figure 8
Figure 8
BALM morningness-eveningness scores are shown for DSPS cases and their controls. Alleles of rs2482705 (A) and Alleles of rs3828057 (B). BALM scores <28 indicate marked eveningness, e.g., DSPS. Scores ≥28 indicate the normal range or greater morningness. The means (blue circles) and 95% confidence intervals (red bars) are shown for homozygotes with the common alleles (left), heterozygotes (center), and for those uncommon homozygotes with the minor alleles (right). For both SNPs, those with 1 or 2 minor alleles (AG or AA) had BALM means outside the confidence limits for the homozygotes (GG) of the common allele, suggesting that minor alleles were associated with dominant protection from DSPS.
Figure 9
Figure 9
Q-Q plot for the meta-analysis of the BALM phenotype. Quantile-quantile plot of SNP association probabilities for meta-analysis of the BALM phenotype in the combined DSPS-case-control and Sleep Center data sets. See the second worksheet of Supplement 1 for a listing of these p values. The dashed red line shows random expectation for the probability distribution. The pink circle highlights rs2482705. Only 3 other SNPs (rs2292910, rs11605924, and rs3820587 from top to bottom) were notably elevated above the trend line, but apparently almost all the nominally significant SNPs (with observed -LOG10[P]>1.3) were slightly above the trend line, suggesting the possibility of dozens of SNP associations each having small effects.
Figure 10
Figure 10
Mechanisms by which DSPS-associated SNPs might lead to photoperiodic disturbances. A: Part of the circadian gene network is shown which promotes transcription through pathways leading to E-box activation (green) or which deactivates transcription and E-box promoter action (red). B: The yellow line illustrates normal melatonin secretion commencing shortly before the preferred nocturnal sleep time and terminating about the time of awakening near dawn, so that preferred sleep times and sleepiness normally correspond. The yellow dotted line illustrates how in DSPS, melatonin secretion may become delayed, consequently delaying sleepiness. C: The gene EYA3 reaches a sharp peak in pars tuberalis transcription about 12 hours after darkness onset (solid orange line), but if melatonin is still elevated (in winter or DSPS), the EYA3 peak is largely suppressed (dashed orange line). D: After short nights in summer, EYA3 and SIX1 coactivate TEF at a D-box on the TSHβ promoter. TSHβ hybridizes with TSHα, releasing active TSH into 3rd ventricle CSF. E: TSH circulates retrograde to promote deiodinase 2 which converts T4 to T3. F: T3 promotes synthesis and release of gonadotropin hormones, implementing summer reproduction and good mood.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Gale C, Martyn C. Larks and owls and health, wealth, and wisdom. BMJ. 1998;317:1675–1677. - PMC - PubMed
    1. Roberts RD, Kyllonen PC. Morningness-eveningness and intelligence: early to bed, early to rise will likely make you anything but wise. Pers Individ Dif. 1999;27:1123–1133. - PubMed
    1. Merikanto I, Lahti T, Kronholm E, Peltonen M, Laatikainen T, Vartiainen E, et al. Evening types are prone to depression. Chronobiol Int. 2013;30:719–725. - PubMed
    1. Lee HJ, Rex KM, Nievergelt CM, Kelsoe JR, Kripke DF. Delayed sleep phase syndrome is related to seasonal affective disorder. J Affect Disord. 2011;133:573–579. - PMC - PubMed
    1. Giglio LM, Magalhaes PV, Andersen ML, Walz JC, Jakobson L, Kapczinski F. Circadian preference in bipolar disorder. Sleep Breath. 2010;14:153–155. - PubMed