Reppert, at Massachusetts General Hospital at the time, and his team showed that the murine circadian clock is orchestrated by positive and negative transcription and post-translation feedback loops. Using mutant mice, they found that the Period2 protein positively regulates the Bmal1 gene loop, and that cryptochrome proteins negatively regulate the Period and Cryptochrome loops. The Takahashi lab paper showed that a mutation of a specific kinase that modifies circadian clock proteins has a clear effect on the mammalian circadian clock.
Through the Loop-D-Loop
Reppert surmised that "there is this discrepancy between the molecular alterations in Cryptochrome-deficient mice and Period 2-deficient animals that we thought could be explained by a second loop." The other factor the team considered is the clock's two essential transcription factors, Clock and Bmal1. Bmal1 is known to be rhythmic and its rhythm is out of phase with the Period and Cryptochrome genes. "Bells went off," Reppert remembers. The investigators thought this gene expression discrepancy could be explained by Period 2 coming into the nucleus to regulate Bmal1 transcription.
Reppert's team proposed a model of interacting molecular loops, one governing Period and Cryptochrome gene transcriptions, and a second controlling Bmal1 transcription. At the beginning of the day, a Clock-Bmal1 protein complex, bound to the promoter regions, transcribes the Period and Cryptochrome genes. The translated Period and Cryptochrome proteins form complexes necessary for their transport to the nucleus, where the Cryptochrome proteins turn off Period and Cryptochrome transcription. At the same time, nuclear-bound Period2 protein enhances Bmal1 gene transcription. Over the next several hours, Bmal1 protein again becomes available to partner with the Clock protein to restart the clock-cycle transcription.
A Mutant Effect
"We had a lot of hurdles in this project," Takahashi remembers. "It took seven years to complete."4 Phillip Lowrey, a postdoctoral fellow in Takahashi's lab, notes that "people used to shake their heads when we explained we were trying to clone tau. That's because, unlike the case for mouse and rat, there are virtually no resources for cloning genes in the hamster. We had to create our own." The first hurdle was to map the hamster genome around the mutation region, but first the researchers had to find a Syrian hamster strain with DNA polymorphisms so they could cross the mutant animals. Most of the existing Syrian hamster strains, however, are 70-plus years old, derived in that country from a single capture of just three littermates. The investigators eventually found one: In 1971, Michael Murphy, an MIT psychology graduate student, captured a strain that had more genetic variation. (The National Institutes of Health now maintains this strain.)
After crossing the two hamster strains, one of which carried the tau mutation, researchers measured the offsprings' circadian behavior, noting whether the hamsters were homozygous mutant, with a 20-hour clock; heterozygous, with a 22-hour clock; or normal, with a 24-hour clock. Next, Takahashi's group needed to find genetic markers to map the mutant gene to one of the 22 hamster chromosomes. Using a method called genetically directed representational difference analysis, the team searched for genetic markers in regions around the mutation. This strategy detects differences in restriction sites linked to the mutation, enabling the researchers to clone small pieces of linked DNA. This produced the first markers, which they called RDA650 and RDA750.
With these markers in hand, they screened hamster genomic libraries to find additional sequences near the markers. Eventually, a hamster sequence was found that corresponded to a gene on murine chromosome 15. "We knew we were on to something because we had two markers that went to exactly the same place," says Takahashi.
That finding allowed the group to go back to the hamster and clone the hamster version of additional murine genes in the region around the tau mutation. Owing to the dense genetic maps available for mouse and human, once they knew the gene's position on mouse chromosome 15, the re-searchers could find the human conserved segment. There turned out to be three--on human chromosomes 8, 12, and 22.
"When we went to humans a candidate gene popped out," says Takahashi. That prospect turned out to be casein kinase I epsilon, located on human chromosome 22. "It became interesting because of the earlier double-time casein kinase I mutation cloned in flies."
|Courtesy of Raw Sienna Inc.|
Luck Lends a Hand
"Using information that came from fly genetics, the double-time mutation, and what we found in hamsters, it looked like the role of the hamster enzyme was the same," says Takahashi. "What's incredible is that the enzyme appears to be doing exactly the same in fly as in hamster. So this is a highly conserved system." Lowrey also points out that "recent work on a human sleep disorder called familial advanced sleep phase syndrome by University of Utah researchers5 shows that in affected individuals there's a mutation in the casein kinase I epsilon binding site on one of the enzyme's substrates." This underscores the importance of clock genes like casein kinase I epsilon in regulating human sleep-wake behavior and suggests that future therapies for genetic sleep disorders may, in part, target the clock proteins.
Reppert says he thinks that his group's work is highly cited because it provides a useful model of how the mammalian circadian clock functions. "The model had evolved to a certain point about a year before the paper came out," he recalls. "But this paper put everything that we knew in a context in which these loops were working and assigned functions to the various clock genes and their protein products, which hadn't been done before."
Takahashi says he believes that his lab's paper made the Hot Paper list because it is such a complete story, using many different techniques from many disciplines: behavior, genetics, molecular biology, and comparative genomics. "Of course it's of interest to chronobiologists, but what I found from colleagues is that the ones most interested [work in] genetics because we showed that you could clone a gene in an organism with few genetic resources by using comparative information, in this case from mouse and human genomes," concludes Takahashi.
In the last two years, the Reppert lab has been involved with further defining the functions of the Period genes and elucidating post-translational mechanisms in the mammalian circadian clock. A recent paper6 looks more rigorously at protein modification within the mammalian circadian system. "Now that we have the loops, we're asking how the time delays are built into the system through the protein modification to get the 24-hour kinetic to the clock," he says.
The field, say the researchers, is really coming together. Reppert sums it up: "There were no mammalian clock genes characterized six years ago. There are now seven, probably eight, and we can talk about an actual mechanism."
1. X. Jin et al., "A molecular mechanism regulating rhythmic output from the suprachiasmatic circadian clock," Cell, 96:57-68, 1999.
2. K. Kume et al., "mCRY1 and mCRY2 are essential components of the negative limb of the circadian clock feedback loop," Cell, 98:193-205, 1999.
3. K.Y. Kreeger, "The quest for perfect timing," The Scientist, 15:16, July 9, 2001.
4. E. Russo, "Circadian rhythms," The Scientist, 13:16, Oct. 25, 1999.
5. K.L. Toh et al., "An hPer2 phosphorylation site mutation in familial advance sleep phase syndrome," Science, 291:1040-3, 2001.
6.C. Lee et al., "Posttranslational mechanisms regulate the mammalian circadian clock," Cell, 107:855-67, December 2001.
J.C. Dunlap, "Molecular bases for circadian clocks," Cell, 96, 271-90, Jan. 22, 1999.