Molecular insights into human daily behavior Steven A. Brown*†‡, Dieter Kunz§, Amelie Dumas†, Pål O. Westermark¶, Katja Vanselow*, Amely Tilmann-Wahnschaffe§, Hanspeter Herzel¶, and Achim Kramer* *Laboratory of Chronobiology, Charite´-Universita¨tsmedizin Berlin, Hessische Strasse 3–4, D-10115 Berlin, Germany; †Institute for Pharmacology and Toxicology, University of Zurich, Winterthurerstrasse 190, CH-8057 Zurich, Switzerland; §Department of Psychiatry and Psychotherapy, Charite´-Universita¨tsmedizin Berlin (PUK), Grosse Hamburger Strasse 5–11, D-10115 Berlin, Germany; and ¶Institute for Theoretical Biology, Humboldt University, Invalidenstrasse 43, D-10115 Berlin, Germany Edited by Joseph S. Takahashi, Northwestern University, Evanston, IL, and approved December 10, 2007 (received for review August 17, 2007) Human beings exhibit wide variation in their timing of daily The circadian clock is organized in a hierarchical fashion. A behavior. We and others have suggested previously that such master clock tissue in the suprachiasmatic nucleus of the brain differences might arise because of alterations in the period length hypothalamus receives light input via the retinohypothalamic of the endogenous human circadian oscillator. Using dermal fibro- tract, and peripheral ‘‘slave’’ oscillators in the cells of most other blast cells from skin biopsies of 28 subjects of early and late tissues receive entrainment signals from the SCN. These signals chronotype (11 ‘‘larks’’ and 17 ‘‘owls’’), we have studied the are probably redundant, and include indirect cues such as body circadian period lengths of these two groups, as well as their ability temperature and feeding time, and direct ones such as hormone to phase-shift and entrain to environmental and chemical signals. secretion and sympathetic enervation (7). The clock mechanism We find not only period length differences between the two itself is cell-autonomous, and in mammals is probably driven by classes, but also significant changes in the amplitude and phase- an interlocked network of transcriptional feedback loops. These shifting properties of the circadian oscillator among individuals loops involve gene products that act both positively (CLOCK, with identical ‘‘normal’’ period lengths. Mathematical modeling BMAL1) and negatively (CRY1-2, PER1-3, REV-ERB␣) upon shows that these alterations could also account for the extreme cis-acting elements (E-boxes and ROR elements) (8). behavioral phenotypes of these subjects. We conclude that human Such elements have been used separately in the context of chronotype may be influenced not only by the period length of the luciferase fusion constructs to permit bioluminescent measure- circadian oscillator, but also by cellular components that affect its ment of circadian clock phase and period, both in transfected amplitude and phase. In many instances, these changes can be cells and in explanted tissues from transgenic reporter animals studied at the molecular level in primary dermal cells. (9, 10). Because the fundamental nature of the circadian ‘‘clock- work’’ in the SCN and peripheral tissues is highly similar and chronotype ͉ circadian ͉ fibroblast ͉ genetics based upon identical components, mutations in circadian clock genes affect clock function comparably both in the SCN and in rom photosynthetic bacteria to man, organisms throughout isolated peripheral tissues (11, 12). Phenotypes of circadian Fevolution have evolved biological clocks to adapt better to the mutations are often exaggerated in explanted peripheral tissues, 24-h period of the solar day. These so-called ‘‘circadian’’ clocks probably because circadian phase among these cells is less tightly (from the Latin circa diem, ‘‘about a day’’) permit both the coupled than in neurons (13, 14). anticipation of daily environmental changes and the segregation Previously, we have used a lentivirally delivered reporter to different time periods of antagonistic biological processes. In system to measure circadian period length in fibroblasts isolated mammals, many aspects of physiology show circadian regulation: from mice and from human subjects. Fibroblasts furnish an sleep–wake cycles and cognitive performance, cardiac function excellent model system because they are easy to obtain and (heartbeat and blood pressure), renal function, and most aspects cultivate, and in explants from transgenic Per2:luc mice, their of digestion (enzyme secretion, gastric throughput, and detoxi- period closely matched that of the SCN (9). In our studies of mice fication). In total, Ϸ10% of genes have circadian patterns of and humans of different genetic background, fibroblast period expression (1). length seemed to be a property of the individual from whom they In addition to its roles in regulating physiology, the circadian were taken. In mice, it was similar to the period of wheel-running oscillator is believed to control behavior: for example, daytime in the same animals: mutations in clock genes that shorten, preference (‘‘chronotype’’) in human beings. It has been sug- lengthen, or abolish the circadian period of wheel-running had gested that people with longer free-running circadian periods, like effects upon circadian clocks in fibroblasts taken from these i.e., the time taken for one circadian cycle under constant animals (15). environmental conditions, have later phases of behavior under In this article, we have used a fibroblast-based assay to address normal day–night conditions, and people with shorter periods the contribution of circadian clock properties to human daily have earlier phases. The validity of this model within the human behavior. Biopsies from subjects of early and late behavioral population has been challenging to assess because subjects must phase (as measured by a chronotype questionnaire) were culti- be kept for prolonged intervals in a controlled laboratory vated, and the clock properties of fibroblast pools were mea- environment. Under such conditions, a loose correlation has sured. Overall, a correlation was seen between fibroblast period been observed between human circadian period and chrono- length and subject behavioral phase. This correlation was espe- type (2). Twin studies show that daily behavior has a significant genetic component (3). Syndromes associated with extreme chronotype Author contributions: S.A.B. and A.K. designed research; S.A.B., D.K., A.D., K.V., and A.T.-W. (Familial Advanced Sleep Phase or Delayed Sleep Phase Syn- performed research; P.O.W. and H.H. contributed new reagents/analytic tools; S.A.B., D.K., drome) have been mapped to mutations in circadian clock genes P.O.W., H.H., and A.K. analyzed data; and S.A.B., P.O.W., H.H., and A.K. wrote the paper. (4, 5), and in one case have been specifically tied to shortened The authors declare no conflict of interest. period length in affected individuals (6). Nevertheless, the This article is a PNAS Direct Submission. separation of environmental factors from genetic ones in the ‡To whom correspondence should be addressed. E-mail: [email protected]. study of behavior remains a formidable obstacle in the identi- This article contains supporting information online at www.pnas.org/cgi/content/full/ fication of biological factors that affect human chronotype. In 0707772105/DC1. this study we have taken a cellular approach to this question. © 2008 by The National Academy of Sciences of the USA 1602–1607 ͉ PNAS ͉ February 5, 2008 ͉ vol. 105 ͉ no. 5 www.pnas.org͞cgi͞doi͞10.1073͞pnas.0707772105 Downloaded by guest on September 29, 2021 cially striking in individuals whose fibroblasts had extreme A B period lengths. Among individuals whose fibroblasts had ‘‘nor- Lentiviral circadian reporter dnoces dnoces 100 mal’’ period length, further analyses of some samples demon- pCMV-puroR 80 strated either an abnormal magnitude of circadian gene tran- 4xINSBglobin 60 23.1 h 40 scription (resulting in a lower oscillator amplitude), or abnormal r 20 BMAL1 BMAL1 lentistnuo ep p -luciferase-UTR LTR 70 phase-shifting properties in response to a chemical phase- 60 shifting agent. We show by simple mathematical models that 50 24.2 h 40 30 each of these differences would be able to change the phase of ecnec c( the circadian clock without affecting period length. Thus, our 105 Dermal fibroblasts 95 results are consistent with the hypothesis that human daytime 85 24.7 h 75 preferences within the general population can be influenced by s 65 enimul 70 multiple clock properties that can be studied in primary dermal 60 cells. 50 25.6 h 40 d5-d9 Measure bioluminescence o i 30 d1 d4 B) Results Plate Synchronize with 1 2 3 4 5 days There Is a Significant Correlation Between Human Chronotype and cells dexamethasone Dermal Fibroblast Period Length. A questionnaire about daytime C 25.5 preference [the Horne–Ostberg Chronotype Questionnaire, HOQ (16)] was administered to subjects responding to newspa- per and television advertisements or personalized recruitment )sruoh( htgnel doireP htgnel )sruoh( 25 efforts seeking individuals of extreme chronotype. Eleven indi- viduals of early type (‘‘larks’’) and 17 of late type (‘‘owls’’) were chosen for participation. The larks possessed a HOQ score of 24.5 Ͼ60, and the owls had a HOQ score of Ͻ40. These scores reflect responses to questions in which subjects are asked to characterize their preferred waking time and sleeping times, alertness at 24 different times of day, and vacation habits. GENETICS From these individuals, two 2-mm dermal punch biopsies were taken and separately cultivated.