Decoding the rhythm of life

An international team of systems biologists led by researchers from the RIKEN Center for Developmental Biology in Kobe has used statistical methods to predict, seek out and finally build new DNA sequences that can regulate daily or circadian rhythms in cells. The group used data accumulated across a range of genomes to gain a greater understanding of the general biological principles of the control of circadian systems.

A network of genes ensures that the rhythms of organisms—sleep and wakefulness, changes in body temperature and blood pressure, the secretion of hormones and regulation of fertility—are attuned to daily and seasonal cycles. In humans common problems, such as jet lag and lack of alertness of shift workers, arise when the body’s circadian rhythms are not properly adjusted to the external environment (Fig. 1). Permanent disruption can lead to more serious disorders and has been implicated in bipolar disorder.

The RIKEN researchers conducted studies on the regulation of the estimated 5 to 10% of mammalian genes that show circadian rhythms. They worked in collaboration with scientists from the University of Pennsylvania, Kinki University in Osaka and INTEC Systems Institute in Tokyo and published their findings in the Proceedings of the National Academy of Sciences¹.

Members of the group began by constructing a database of promoters and enhancers—segments of DNA involved in regulating genes—from several mammalian genomes. They then used a statistical technique known as the Hidden Markov Model, with which they could search their database for core DNA sequences associated with three kinds of transcription factors involved in regulating activity with different daily peaks—E-box (morning), D-box (daytime) and RRE (night). This approach provided an estimate of the probability that any sequence it found was associated with such a regulator.

After picking the 10 most probable candidate genes in each category, the researchers were able to add them into seven different mouse tissues. Thirteen of the 30 regulatory candidates demonstrated strong circadian activity.

Taking the work a step further, they then used their model to design two artificial DNA sequences for each of the three kinds of regulatory genes—one predicted highly rhythmic activity, the other lower activity, but neither of which appears naturally in mouse or human genomes. When the researchers tested the activity of genes containing these synthetic sequences, their results showed that the sequences flanking the core of the regulators are significant in determining the amplitude of their impact on gene activity.