PerspectiveCircadian Rhythms

PAS, Present, and Future: Clues to the Origins of Circadian Clocks

See allHide authors and affiliations

Science  02 May 1997:
Vol. 276, Issue 5313, pp. 753-754
DOI: 10.1126/science.276.5313.753

Last year, sales of melatonin in the United States surpassed those of vitamin C. Some of these pills were consumed in an attempt to overcome the recalcitrance of our internal circadian clock and force it to adapt more quickly to new day-night cycles—in a foreign time zone or in night shift work. The clock's self-sustained expression in such circumstances is a hallmark of its ubiquitous, and critical, role in all organisms from vertebrates to plants to single-cell life forms. The molecular circuitry of this internal timing system has become an intense field of study. The latest payoff for this intensity is the identification on page 763 of this issue of a known transcription factor, WC-2, as a clock protein in the fungus Neurospora (1). This news yields a two-for-one reward: It sheds light on how clocks whirl away inside us but also generates intriguing ideas about their origins.

Why do almost all organisms have circadian clocks? There is a potential adaptive advantage of timing behavioral, metabolic, and developmental processes to the appropriate phase of night or day. A simple biochemical sand timer might suffice for this task, but it could not cope with seasonal changes in day length. In contrast, a circadian clock could accommodate these conditions, by being sensitive to external signals. Thus the hands of the clock can be bumped forward or backward each day by light, maintaining the clock's relevance to the environment. Clock regulation and phototransduction are inextricably tied together.

Genetic analyses, notably in Neurospora and the fruit fly Drosophila (2), have led to the discovery of some of the clock's molecular cogs and to a general picture that clocks are constructed of transcription factors that feed back and inhibit their own transcription. Neurospora provides easy access to the inside of the clock. When inoculated into one end of a glass tube, the mycelium of this fungus will propagate along the tube and produce pigmented spores on top of aerial hyphae about every 24 hours. By using this phenotype, researchers have isolated arrhythmic mutants and mutants with altered period lengths, most notably in the frequency gene (frq), a bona fide clock component. To belong to this small and prestigious club, clock molecules must pass the admission criteria of molecular horology: A clock component must itself cycle; when it is held at a constant level, the clock should stop; and it must respond rapidly to signals that bump the phase of the clock, such as light (3). The sophisticated molecular and genetic tools available for Neurospora have shown that the frq gene passes on each of these criteria (2, 4). In addition, the FRQ protein feeds back and inhibits its own transcription, as has also been found for per and tim—also bona fide clock genes—in Drosophila (2, 5).

While the clock itself was being “brought to its knees” (5), other groups have studied how light affects its function. Macino and his colleagues, and others, have isolated Neurospora mutants that are blind to light, and they have now cloned two of the responsible genes (6, 7). These are white collar (wc)-1and -2, named not because of their social position, but because they lack carotenoid pigments at the fringes of colonies. Both genes are transcription factors containing zinc finger and transcriptional activation domain motifs. WC-1 and WC-2 thus appeared to be transcriptional components of the blue-light photosensory pathways in Neurospora. Furthermore, these proteins both have PAS domains. PAS was initially identified as two direct repeats (PAS-A and PAS-B) in the Drosophila clock protein PER, in the basic helix-loop-helix (bHLH)-containing transcription factors ARNT and AHR in mammals, and in SIM in flies (8) (see figure). Other PAS-containing proteins have now been identified (9), and PAS domains have been shown to mediate protein-protein interactions (8) and, in one case, to bind small ligands.

PAS domains line up.

AHR, the mammalian dioxin receptor; MESPHY, an algal phytochrome.

The identities of WC-1 and WC-2 as phototransducers have now been extended in a surprising way by Crosthwaite et al. (1), who have shown that WC-2 is also a clock component, required for the maintenance of expression of frq. Thus, WC-2 appears to be intimately involved in both phototransduction and clock function and represents the first positively acting clock component. Indeed, WC-2 could well be the positive factor regulating FRQ that is negated by the negative feedback of FRQ itself and therefore may provide the first example of “closing” the clock loop in anyorganism. Although the wc-1 gene appears not to encode a clock component, it is essential for light responses, and the clock eventually sputters to a halt in its absence; therefore, WC-1 appears to be required generally for clock progression but is not directly involved in the feedback loop. One would expect several such factors to be found.

The significance of these findings is likely to extend well beyond Neurospora, because PAS is now recognized as a signature motif found in clock proteins from both Neurospora and Drosophila. This provides the first real evidence that clock proteins may have a common evolutionary origin. But where did the exons that constitute PAS domains arise? A wonderful twist in this tale can be found in clues from photoreceptor proteins. Lagarias and colleagues (10) pointed out that the plant photosensory photoreceptors, the phytochromes, contain evidence of homology to the PAS-A and -B domains, and also to almost half of the bacterial blue-light receptor, photoactive yellow protein (PYP). PYP binds its 4-hydroxycinnamyl chromophore in a loop, half of which is contained in the PAS/photoreceptor homologous regions (10, 11). Did some clock genes, bearing PAS domains, arise from ancient photoreceptor proteins? The idea is attractive, given that light perception and clock function are so closely related.

How can this question be resolved? First and foremost, we will need to see sequences of clock genes from a range of organisms. The recent identification of clock mutants in the prokaryotic cyanobacteria (12), and also in higher plants (13), should lead to the cloning of more clock components. Will some of these contain PAS domains, making the use of this motif universal? Second, are proteins such as WC-1 and WC-2 (or even PER) also photoreceptor molecules themselves? Macino and his colleagues have speculated this could be so, on the basis of the homology of the PAS region to part of the chromophore-binding region of PYP (7). Modeling of PAS domains based on the PYP structure may indicate whether the domain is at least a structurally, if not evolutionarily, conserved motif. Whatever the answers one thing is clear: The study of photosensory processes and the the circadian clock must now proceed hand-in-hand.

References and Notes

View Abstract

Navigate This Article