Like watchmakers, biologists have hunkered down over their respective model organisms, meticulously seeking out biological timekeepers, the genes important for regulating life's internal clock. Up until now, classical approaches had not uncovered the finest details of the machinery that synchronizes life processes with light and darkness, let alone how these rhythms affect behavior and metabolism. "[They] haven't identified genes other than the main components such as the central transcription factors," says Paul Etter at the University of Arizona, Tucson. He and adviser Mani Ramaswami recently wrote a review on the topic.1
Now, full genome sequences for such creatures as the fruitfly and mouse have allowed these so-called chronobiologists to pull off the clock's proverbial face and reveal the intricate inner workings. The verdict: It's still pretty complicated in there. Sophisticated tools such as the microarray, however, offer new power in screening for genes that cycle over time. Indeed, within the last two years more than a dozen papers have been published on this subject.2-8
Reports on gene expression in fruit flies and mammals using arrays point to key candidate genes for further downstream study, which scientists hope will shed light on such diverse topics as hibernation and drug metabolism. But while there is seemingly not much agreement--concordance, in chronobiological terms--on the types of genes that these array surveys pick up, researchers have extracted a score or more of new clock and output gene candidates. These genes in both flies and rodents represent a diverse set of functions, including basic metabolic processes and neurological pathways.
TOP TWENTY Studying gene-expression changes that occur over a 24-hour period is time-consuming and expensive. So, first-generation tissue studies have focused on a relatively small-parts list, including fly heads, rodent brains, and livers. "Everyone's only looked at the adult fly head so far, and of course that's a heterogeneous tissue," says Michael Young, professor of genetics and director of the Center for Biological Timing at Rockefeller University, New York City.
These heterogeneous heads offer specific challenges. "To some extent, we expect to get an abbreviated list of what's going on, just because things get diluted if only a subset of the brain is expressing," says Young. Still, the first paper covering this topic in Young's lab showed a continuum of gene expression with peaks at different times of the day and night.2 "It's like a drum in a music box where you have this complicated melody, and in this case a complicated pattern of gene expression, [which] you accomplish with every turn of the drum, [and] that touches all aspects of biology," says Young. The aspects his group found: metabolism, synapse function involved in learning and memory, vision, and olfaction.
"The field had good reasons to believe that many genes would be regulated by the clock; that there would be many 'cycling' genes evident by microarray analysis," says Paul Taghert, professor of neurobiology, Washington University Medical School in St. Louis. Taghert's group measured gene expression in fruitfly heads, comparing controls with flies with a mutant period gene.3 "We found that only a small number of genes could reliably be called circadian cyclers." Seventy-two genes in control flies showed diurnal rhythms in light-dark cycles.
Although a few groups have studied fly chronobiology using arrays, there is no precise mathematical definition of what constitutes a circadian pattern of gene expression, so each group had to devise its own methods of analysis and its own definitions, notes Taghert. "The lists of genes called circadian were largely discordant, and we think most genes called circadian are actually false-positives," he explains. "There is a small group of about 20 genes that all groups agree cycle robustly; we consider these the most reliable." This group includes all the known circadian genes, and their presence serves as an internal control.
But the researchers also have to look for possible false- negatives, that is, genes cycling in the fly head but that for various reasons were not cleanly identified as such. What's needed now, says Taghert, is first to confirm the reliable cyclers and test their precise roles in vivo, and then find out how they contribute and what counterparts exist in mammalian systems.
MAMMALIAN METABOLISM In mammals, investigators have compared gene expression in the liver, which governs much of mammalian metabolism, with expression in the suprachiasmatic nucleus (SCN), a knot of cells in the hypothalamus that controls peripheral organ cycles. Charalambos P. Kyriacou, a University of Leicester, UK, geneticist, and Michael H. Hastings, a neurobiologist at MRC Laboratory of Molecular Biology, Cambridge, UK, say several results surprised them.4 They found a high proportion of genes that cycled at the mRNA level in the murine liver, representing about 9% of the several thousand sampled and covering many types of cellular activity. About 5% of the sampled SCN genes cycled, but all told, only about 10% overlapped with cyclers in the liver. Still, the group found that cyclic gene expression in peripheral organs, including that for both canonical circadian and novel genes identified in arrays, was dampened or destroyed when the mouse SCN was surgically ablated.
Joseph S. Takahashi, a Howard Hughes Medical Institute investigator at Northwestern University, Chicago, Ill., and colleagues were also taken aback by the 10% overlap in the cycling genes between the SCNs and livers they studied.5 "We thought there'd be a larger conserved set of cycling genes," says Takahashi. "It's surprising, not because the gene is expressed in one tissue and not the other ... it was just that one oscillated in one tissue and not the other." In the SCN, they found that the largest group of cycling transcripts was involved in peptide processing, protein transport and turnover, and neurosecretion, but these weren't under circadian control in the liver.
Jay C. Dunlap, genetics department chairman at Dartmouth Medical School, and colleagues took a different approach; they surveyed genes in an immortalized cell line of rat fibroblasts, a cell type with more consistent physiological history than live animal samples.6 They found that about 2% of the genes they studied, including the canonical clock genes, showed consistent circadian expression. Cyclic genes were diverse, including transcription factors, ubiquitin-associated factors, proteasome components, and RAS/MAPK signaling pathway components. While not matching the complexity of a living organism, fibroblasts are a biochemical workhorse and have all the regular clock genes, explains Dunlap. Comparing the cultured fibroblasts to in vivo cell types from other studies showed a similar range of overlap; 16% of the rhythmic genes from fibroblasts were also rhythmic in the hypothalamus or the liver, and 5% were common to all three tissue types.
MISSING PROTEINS "Microarrays are the latest wave at being able to do the same thing better and better," says Dunlap, comparing array technology's contribution to the field to older techniques such as subtractive hybridization. "I don't believe that anyone thinks this set of papers is the final answer, because not all the genes are on the arrays and others are underexpressed. Also, it doesn't get at proteomics."
Protein production and turnover rates and epigenetic controls throw another wrench in the complicated circadian systems. Several groups aim to take the next proteomic step. Hiroki R. Ueda, head of the laboratory for systems biology at RIKEN, Kobe, Japan, and colleagues have published papers on both mammals and flies.7,8 In the latter flies they found more than 100 genes cycling in both light-dark and constant-dark conditions. In mice they found more than 100 clock-controlled genes in the SCN and nearly 400 clock-controlled genes in the liver, with candidate genes again covering a range of functions, including transcription factors, RNA-binding factors, receptors, ligands, and enzymes.
Ueda's group is starting to look at protein expression based on this work for practical purposes, in a sense looking to move from watch-hobbyists pulling together parts lists, to timepiece master craftsmen. Says Ueda: "We are much interested in chronotherapy and microarray expression data of liver, which may provide precious information to understand how our metabolism, including drug metabolism, is controlled by the circadian clock."
Karen Kreeger (firstname.lastname@example.org) is a freelance writer in Media, Pa.
1. P.D. Etter, M. Ramaswami, "The ups and downs of daily life: profiling circadian gene expression in Drosophila," BioEssays, 24:494-8, 2002.
2. A. Claridge-Chang et al., "Circadian regulation of gene expression systems in the Drosophila head," Neuron, 32:657-71, 2001.
3. Y. Lin et al., "Influence of the period-dependent circadian clock on diurnal, circadian, and aperiodic gene expression in Drosophila melanogaster," Proc Natl Acad Sci, 99:9562-7, 2002.
4. R.A. Akhtar et al., "Circadian cycling of the mouse liver transcriptome, as revealed by cDNA microarray, is driven by the suprachiasmatic nucleus," Curr Biol, 12:540-50, 2002.
5. S. Panda et al., "Coordinated transcription of key pathways in the mouse by the circadian clock," Cell, 109:307-20, 2002.
6. G.E. Duffield et al., "Circadian programs of transcriptional activation, signaling, and protein turnover revealed by microarray analysis of mammalian cells," Curr Biol, 12:551-7, 2002.
7. H.R. Ueda et al., "A transcription factor response element for gene expression during circadian night," Nature, 418:534-9, 2002.
8. H.R. Ueda et al., "Genome-wide transcriptional orchestration of circadian rhythms in Drosophila," J Biol Chem, 277:14048-52, 2002.