Skip to main content

Circadian clock without DNA--History and the power of metaphor

This article was published in Scientific American’s former blog network and reflects the views of the author, not necessarily those of Scientific American


Last week, two intriguing and excellent articles appeared in the journal Nature, demonstrating that the transcription and translation of genes, or even the presence of DNA in the cell, are not necessary for the daily ("circadian") rhythms to occur (O’Neill & Reddy 2011, O’Neill et al., 2011). (Scientific American is part of Nature Publishing Group.)

The two papers received quite a lot of media coverage, and deservedly so, but very few science bloggers attempted to write in-depth blog posts about them, placing them in a broader historical, theoretical and methodological context. I had a feeling that everyone was waiting for me to do so. Which is why you are now reading this. I know it is a long "Explainer" (which is all the rage in science journalism these days) but I hope you have patience for it and that you find it informative and rewarding.

What I intend to do is to, first, briefly describe and explain the research in these two papers, though the press release and media coverage were quite accurate this time. Diana Gitig did the best job of it at Ars Technica - I highly recommend you read her piece for clear background information.


On supporting science journalism

If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.


Then I will try to give you a historical perspective so you can get a feel for the context in which this research was performed. This look at the history will bring into sharp relief how powerful the scientific metaphors are in guiding the questions that researchers try to answer in their laboratories. Finally, a look at the media coverage will show that the lay audience (including journalists) is guided by other metaphors - not always the same ones that are used by researchers.

What did they do?

In each of the two papers, the researchers chose an unusual laboratory model for their study. What is common to both models is that a) they are both Eukaryotic cells and b) there is no DNA transcription or RNA translation going on inside the cells.

In the first study (O’Neill & Reddy 2011), they used human red blood cells (photo left) as these cells have no nucleus, thus no DNA at all.

In the second study (O’Neill et al. 2011), the model of choice was a small protist, Ostreococcus tauri which has an interesting property - when kept in constant darkness, there is no DNA transcription or RNA translation that can be detected.

The starting point of both studies was detection of peroxiredoxins in the cytoplasm. Peroxiredoxins are enzymes (thus chemically proteins) that protect the cells from damage from strongly oxidizing molecules (often refered to as "free radicals"). The process of neutralizing such oxidants temporarily changes the chemical structure of the peroxiredoxin, which then reverts to its native state again - thus the molecule is constantly switching between the two states. This oscillation between the two states follows a daily (~24h) cycle synchronized to the day-night cycle of the environment.

In both studies, peroxiredoxins were detected using antibodies ("immunoblotting"). One of the chemical states of the molecule can be detected with this method, while the other state is indirectly detected by the comparative lack of signal. Thus a circadian rhythm would be seen as an alternating series of rises and falls of the detected signal with a period close to 24 hours.

And this is exactly what they discovered in both cases - there was a clear circadian rhythm of peroxiredoxins state-switching both in cultured red blood cells (above right) and in the cultured Ostreococcus tauri (below).

Furthermore, in the protist study, they used measurement of light emitted by luciferase added to the sea-water solution as a marker of DNA transcription and translation. While the cells were kept in constant darkness no light emitted due to presence of luciferase could be detected.

But at the onset of environmental light, luciferase-induced light measurements indicated that the transcription started at the phase predicted from the state of the clock before the cells were placed in the dark. This means that the circadian rhythm of DNA transcription did not start at some "Phase Zero," triggered by switching on the light, but that it was driven by a clock that was operating all along while the organism was kept in the dark - a clock that does not require DNA transcription and translation.

Detecting a 24-hour rhythm is not sufficient to ensure that the rhythm is actually circadian. For a biological cycle to be considered circadian, it has to satisfy a number of criteria, e.g., it has to be endogeonous (generated inside the cell, not forced onto it by the environment), it has to persist for several cycles, it has to be unaffected by temperature levels (i.e., the period of the rhythm should be the same regardless of the level of environmental temperature kept constant in the laboratory), it has to be entrainable (synchronizable) by imposed cycles in the environment, etc.

In both model systems, the researchers performed (either in these or prior studies) the entire battery of standard experimental protocols to demonstrate that yes, these are indeed circadian rhythms in both laboratory models.

Use of temperature cycles instead of light-dark cycles to demosntrate entrainment in the first experiment makes sense as human red blood cells do not experience (and cannot detect) light, while they are normally exposed to daily fluctuations of body temperature. The difference between the dawn minimum and evening maximum temperature inside the human body can be as large as 1 degree Celsius, more than sufficient for entrainment - some entire cold-blooded animals like lizards, insulated from the environment by skin and scales, can entrain their rhythms to temperature cycles with the high-low difference as small as 2 degrees Celsius and in some individuals as small as 0.1 degrees (Underwood and Calaban 1987).

In the red blood cell paper, the researchers went further. They also detected circadian rhythms in a few other biochemical processes. For example, hemoglobin, the molecule that transports oxygen from the lungs to the cells, and carbon dioxide from the cells to the lungs, can exist in two different forms inside red blood cells. It is a complex protein molecule, built of four almost-identical units. In this form, hemoglobin can perform its normal function. But there is also a two-unit form which cannot perform the function in gas exchange. Furthermore, the two-unit form produces the oxidizing small molecules - exactly the kinds of molecules that peroxiredoxins have evolved to scavenge and neutralize. It is not surprising that the switching between two-unit and four-unit forms of hemoglobin was also seen to be circadian - and in sync with the peroxiredoxin rhythm.

Likewise in the protist paper (see O.tauri in the photo at right), the researchers performed a whole suite of additional experiments designed to eliminate a variety of potential alternative hypotheses. For example, pharmacological suppression of DNA transcription and translation did not eliminate the circadian rhythm in peroxiredoxin chemistry, but instead demonstrated a complex interplay between the clock driven by transcription of genes and the clock driven by spontaneous biochemical reactions in the cytoplasm.

In summary, the two papers are very solid, the experiments are well designed and performed, the results are persuasive, and the names of authors give me confidence that the data can be trusted.

What does this all mean?

The results of both papers demonstrated that transcription of DNA and translation of RNA is not necessary for the generation of circadian rhythms in two different types of eukaryotic cells belonging to evolutionarily very distant relatives - protists and mammals.

In the case of red blood cells, the result is clear - there is no DNA or RNA in these cells. Thus, the circadian rhythms in these cells have to be generated in the cytoplasm.

In the case of O.tauri, the picture is a little bit more complex: the cells have a nucleus which has DNA. There is a clock driven by transcription and translation of canonical "clock genes." Yet, when this mechanism is supressed - either by constant darkness or by chemicals - the cells still exhibit circadian rhythms generated by the molecules residing in the cytoplasm (and some of those molecules, at least during the first day or two, may be strands of RNA transcribed earlier).

Furthermore, the phase at which the DNA-centered clock starts its cycle is determined by the phase of the cytoplasmic clock, not the other way round, i.e., the cytoplasmic clock is dominant over the nuclear clock.

Why is this so exciting?

Depends who you ask!

When the articles were first published I did not yet have time to read them carefully. But I have e-mail notifications set up so every time Google detects a news article or blog post containing the word "circadian" I get a message. Thus I read a number of media articles about these studies before I read the studies themselves.

The media accounts (see someexamples) tended to emphasize two reasons why these studies are important.

The first one was a surprise that both humans and protists have the same molecules doing the same thing. Their surprise was a surprise to me! Circadian clocks are found everywhere. Peroxiredoxins are found in almost all living organisms on Earth. Just like the structure of cell membrane, the processes of DNA transcription and RNA translation, the genetic code, or the use of ATP (adenosine triphosphate) as the energy currency of the cell, peroxiredoxins are ubiqutous molecules in almost all of life on this planet.

Those are life's universals, something that is expected as we have understood the unity of life even before Darwin. These universals are, to a biologist, usually deemed pretty useless and boring - the background. What gets a biologist excited is variation - the exceptions to the universals. If most organisms use a particular molecule for a particular function but one organism does not - now that is exciting! Why is this? How and why did this organism evolve this switch in function? What was the initial mutation, what was the selective pressure? Those are useful questions in biology that help us understand evolution.

My hunch is that the journalists, either by being lay audience themselves or by targeting their articles to the lay audience as they understand it, focus on the universals - the "Unity Of Life" - as something that in their minds is evidence for evolution and a counter-argument against non-scientific ideas about life (e.g. intelligent design creationism). While biologists find surprise and delight in exceptions, which are useful entries into detailed studies of evolutionary mechanisms, many in the lay audience are still surprised by the plain fact of the unity of life as it evolved from a single common ancestor.

The second reason given in MSM (mainstream media) articles as to why these studies are important is their novelty. For example, Chemical & Engineering News states that this "...involves a previously unknown cycle of posttranslational modifications, in addition to the transcription of a well-known handful of clock genes..."

There is a sense, reading all the coverage, that this is so novel, creative and revolutionary, it must have been the very first time anyone has ever thought of this! And even better - this is a great "conflict" story, in which a single study puts into question an entire field! A small group of young geniuses proved the entire old establishment wrong!

Not so fast!

Both of these studies have been done before. Several times.

I quickly went to the PDFs of the two papers and yes, the references to the old studies are there. The authors are aware of the history of the field, the giants on whose shoulders they climbed in order to see further. Of course, these are Nature papers with severe constraints on space and on the number of references. With so many experiments and so much to explain about their methods and results, they could not spend enough time on the history of the idea and on the work of their predecessors. And they could not cite all of the preceeding studies. But they chose the key ones and noted them briefly in the text - not prominently, but they are there.

So why did the MSM articles not pick up on this? First, by stressing the novelty - and the "conflict" story of geniuses proving the establishment wrong - they make the new studies seem more newsworthy, thus more likely to be approved by the editors to get into print.

Second, the MSM articles are limited by space and there is only so much one can put in 500 words. Thus, as usual, it is the context that gets left out and the novelty-factor that remains in the piece.

Finally, even if a dilligent journalist wanted to follow up on the background he or she would bump into the dreaded paywalls.

Nobody expects a journalist to know as much about the field as its practitioners do. I knew exactly what to look for and even I needed a few days to collect all the papers and read them - which is why you are reading this post now instead of last week.

As a member of the Society for Research on Biological Rhythms I have free access to the Journal of Biological Rhythms, the premier journal in the field. Once logged in, I knew exactly which five articles to look for. Those five articles (all cited below) contained the references to all the other papers I was interested in. As none of those are Open Access, I had to go to Twitter and, by using the hashtag #ICanHazPDF ask my followers to find me and send me PDFs of all of these papers. I got most of them (though some are not available or not even digitized yet as many of them are quite old). Then I had to spend some time reading them.

This all takes time, and I had the advantage of knowing where to start...perhaps even the advantage of being aware, to begin with, that such studies exist. Even Allison Brager who is in the field of chronobiology, did not note on her blog Dormivigilia the existence of prior research and exclaimed with excitement that "....the clinical and scientific relevance of this work are HUGE!!!!...." It is not always stressed hard enough to graduate students how important it is to read the historical literature of one's field.

So, let's quickly go through some of the aspects of the history of circadian research that are most relevant to the understanding of the context in which these two papers appeared.

Brief history of clock research

While the observations of daily rhythms in plants and animals go back to the antiquity, the first experiment in the field was performed in 1729 by Jean-Jacques d'Ortous de Mairan who observed the rise and fall of leaves of a Mimosa plant kept in constant darkness. Much early research was done in the 19th century, mostly on plants, but also a little bit on insects and humans.

In the early 20th century a number of people started studying rhythmic phenomena in living organisms. They came from very different scientific disciplines, e.g., botany (Bunning), ecology (DeCoursey), animal behavior (Kramer, Beling, Sauer), protozoology (Hastings), evolutionary biology (Pittendrigh), mammalian physiology (Richter), human biology and medicine (Aschoff, Halberg), and agriculture (Garner and Allard). It took them a few decades to discover each other's work and to recognize that they are dealing with the same phenomenon regardless of the organism they were studying, be it fiddler crabs, starlings, tobacco or humans.

The founding moment of the field was the 1960 meeting at Cold Spring Harbor. The book of Proceedings from the Meeting (Symposia on Quantitative Biology, Vol.XXV) is a founding document of the field: I own three copies, strategically placed in three different spots in the house so at least one copy has a chance to be saved in case of fire. And you can bet I have read it over and over again during my 10 years in grad school (and after).

In 1960, structure of DNA was very new, and comparatively little was known about the inner workings of a cell. But most researchers were eager to break into the 'black box' and start investigating how the circadian clock ticks inside of the cells. A number of conceptual models were proposed, some focusing on cell membranes, others on DNA. Thus, experiments were started to separate the two parts of the cell and to test the role of nucleus and DNA in the clock mechanism.

Roenneberg and Merrow (2005) provide an excellent timeline of the research since then, especially the emergence of molecular and genetic techniques and subsequent findings. But briefly, over the next two decades or so, circadian mutants were discovered in a protist (Chlamidomonas), insect (Drosophila) and mammal (Golden hamster). The first canonical clock gene - the Drosophila period gene - was sequenced shrtly after.

Around 1995, clock genetics explodes. New clock genes were discovered left and right in several different organisms, from cyanobacteria to humans. There was a sudden influx of people into the field from other areas of genetics, and they required a few years to catch up on the field's history and theory before they stopped making beginners's mistakes in their experimental designs (though see Dunlap 2008 for different perspective on that history). But there were many of them, they had plenty of funding, they were excited and creative, had powerful new techniques, and they worked fast, so every week, or so it seemed at the time, there was a new discovery of genes involved in circadian rhytms.

Here is a timeline of the key events in this aspect of the history of the field:

 

What emerged from all of this activity is the transcription/translation model (Hardin, Hall & Rosbash 1990) for the circadian mechanism: a suite of canonical clock genes get transcribed and translated, and their protein products, after some delay, inhibit the transcription and translation of those same genes. Day in and day out. Those genes and their products then also regulate expression of all the other genes that the cell uses in its daily function.

But not all were happy with it.

If you read Dunlap 2008 you will certainly detect a tension between researchers who studied whole organisms (always with evolution in the backs of their minds) and treated the clock as a 'black box' for decades before the genetic explosion, and the geneticists who came into the field in the mid-1990s. The former regarded the latter as arrogant and simplistic in their complete focus on DNA. The latter regarded the former as out-dated holists who don't understand the magnificient power of DNA and treated them like Lysenko by those who do not understand Lysenko.

You need to remember that this was in the middle of the Human Genome Project hype, when there was a huge overselling of crude genocentric and gene-deterministic ideas, many of which were fully embraced by the geneticists at the time (they have learned better since then).

I do not want to exaggerate the tension - it was mostly muted. Geneticists were mainly welcome into the field. After all, they were bringing in their tools and skills to do what the field was hoping to do all along - crack open the "black box" and peer inside. The two groups treated each other with respect, and soon started collaborating. The old guard of chronobiology was impressed by the speed and capability of the genetics labs, the rate at which new techniques were developed and improved, and the rapidity of discoveries. They learned (or sent their students to learn) the techniques, and started thinking how to employ them to study circadian phenomena that prior research has already shown occured at higher levels of organization.

The hope was that the genetics and molecular approaches will quickly discover all the core clock genes and the way they interract with each other so the focus can shift back to explaining the things that really matter - properties of ensembles of clock cells and the behaviors of whole organisms (after all, DNA is invisible to selection - phenotype of the whole organism is what ecology and evolution can see to act upon).

The discovery of clock genes required the use of a limited number of laboratory models that are amenable to genetic dissection: mouse as the model for all vertebrates, fruitfly representing all invertebrates, Neurospora crassa standing in for all fungi, Arabidopsis being "the plant" and Synechococcus being used as the only bacteirum known (at the time) to possess a circadian clock. In the early years, a few protists were also used - Paramecium, Euglena, Gonyalax, Acetabularia and Chlamydomonas, but they were later largely abandoned, while new models, like Xenopus and zebrafish entered the arena.

This was a highly unusual state for the field - chronobiology was always extremely comparative, with thousands of species of organisms being studied over the years. The hope was that, once the genes are discovered in model organisms, the findings can be applied to other creatures for a more comprehensive and comparative research program.

Likewise, the focus on genes was also seen as temporary, something to be "waited out" until the findings can be applied to other levels of organization in a more integrative approach. Thus, entire lines of research were reduced or have essentially stopped - tidal, lunar and circannual rhythms, Sun-compass orientation, photoperiodism, development, ecology and evolution - waiting for new techniques and new findings that will enable them to re-start.

And that is exactly what happened - a decade or so later, the field concluded that all the core clock genes were discovered and that the transcription/translation feedback loop model is good enough. The study of previously semi-abandoned topics (and organisms) started with new zeal and gusto.

Persisting problems

At the time of the Cold Spring Harbor symposium in 1960, there were two main lines of thinking about the cellular mechanism of the circadian clock. One focused on the nucleus and the DNA (Ehret and Trucco 1966). The other focused on the cell membrane (Njus et al. 1974).

How does one go about figuring out which one of the two models is right, using techniques available at the time?

One approach is to use cells that do not normally possess a nucleus or any DNA - like mammalian red blood cells - to see if they have circadian rhythms. If yes - nuclues is not important, membrane (or cytoplasm) is. Studies were difficult and results not always clear, but most could detect rhythms in red blood cells (Cornelius and Rensing 1976, Mabood et al. 1978, Ohm-Schradera et al. 1980, Peleg et al. 1990a,b)

Second approach is to use very large cells that can survive long enough once the nucleus is removed - in comes the protist Acetabularia (Sweeney and Haxo 1961, Schweiger et al. 1964, Vanden Driesche 1966,Terborgh and McLeod 1967, Vanden Driesche and Bonotto 1969, Sweeney 1974, Mergenhagen and Schweiger 1975a,b, Hartwig et al. 1985, Woolum 1991, Runft, Linda and Mandoli 1996). These studies showed that clock operates after the nucleus is removed, and, once the nucleus is reintroduced, it is the clock in the cytoplasm that determines the phase, entraining the nuclear clock.

The third approach is to pharmacologically block DNA transcription and RNA translation. This was, over the years, performed in a number of organisms, including Acetabularia (photo on the right) and, much more recently, the sea-slug Bulla gouldiana (Page 2000). Again, rhythms persisted in the absence of DNA transcription.

Fourth approach is to find single-cell organisms that reproduce or divide more often than once a day and see if the circadian phase is preserved during the process - there is no DNA transcription during cell division. This was initially done in the protist Paramecium (Barnett 1966), but later it was cyanobacteria that were used in this approach (Mori et al. 1996, Kondo et al. 1997). Circadian phase is preserved during reproduction in Paramecium and cell-division in bacteria.

Fifth approach is to find organisms that have circadian rhythms but do not have clock genes. Yeast (Saccharomyces cerevisiae) is one such organism. In the nematode Caenorhabditis elegans, which shows circadian rhythms, the genes usually used for circadian timing are instead used for developmental timing (so-called heterochronic genes).

Sixth approach is to study the rhythms in either the cell membranes (for example in the protist Gonyalax polyedra, Adamich et al. 1976, or fruiftly, Nitabach et al. 2005) or elements of the cytoplasm directly, in a dish (using bacterial clock proteins, Tomita et al. 2004, Mehra et al. 2006, Mori et al, 2007). Again, the isolated cell membrane cycles, and blocking the membrane processes also blocks overt rhythms in whole organisms. Bacterial clock proteins (not DNA) kaiA, kaiB and kaiC, when placed in a test tube, spontaneously oscillate in a circadian fashion.

Finally, one can genetically affect the clock: mutating, deleting, shutting down or overexpressing (forcing expression at high levels at all times with no cycling) canonical clock genes and see if any residual rhytmicity remains. This was done in the fruitfly (Helfrich-Förster 2000), where morning peak of activity is eliminated when the clock gene cycling is stopped, but the evening bout of activity in male flies persists nonetheless. Sometimes a genetics paper would triumphantly state that a deletion of a gene rendered half of the flies arrhythmic, just to be met with a question "so, how do the other half of the flies still cycle without it?"

This research program started with enthusiasm immediatelly after the symposium, yielding troves of interesting data over the years. But, once the geneticists entered the fray, these results were forgotten or ignored. They did not conform to the DNA-based model. The easiest way to make a circadian geneticist in the mid-1990s angry at a conference was to utter the word "Acetabularia" - this was "noise" to be ignored and swept under the rug.

You can see the timeline of the history of this "shadow research program" here:

 

Why did this research persist despite the victorius run of the transcription/translation model?

The earliest studies in this area were a direct outgrowth of the ideas discussed at Cold Spring Harbor. They all yielded the data suggesting that DNA is not the only part of the clock mechanism. Yet, once genetics work took off, these results were ignored. At least some of the people in the field were worried that genetic work is ignoring something potentially important.

Who in the field was worried about this depended on their own background and experience? First, people who worked on organisms that yielded unusual experimental data throughout the history of circadian research, including the fungi and the protists (especially Gonyaulax polyedra, recently renamed Lingulodinium polyedrum but you are unlikely to find many circadian papers using the new name, and the systematics may still be in flux) were one such group.

People working on non-mammalian vertebrates (fish, amphibians, reptiles and birds) were cognisant of the complexity of circadian organization - same clock genes, expressed in different tissues, resulted in clocks of different properties. The clock in the pineal organ, the clock in the retina, the clock in the SCN (suprachiasmatic nucleus of the hypothalamus), the peripheral clocks in all the other tissues - each of those behaved differently despite using the exact same molecular machinery. So the properties must have been modified by something else in the cell, or by the interactions between cells in the tissue.

On top of that, many phenomena, e.g., photoperiodism or sleep, are not properties of individual cells but of interactions between ensembles of cells in the tissues, or even interactions between the organism and its environment. The simplistic "this gene is for clocks" model just could not explain the complexity of observed reality.

Once all the clock genes were deemed discovered, the critiques started popping up (Roenneberg and Merrow 1998, 1999, 2005, Lakin-Thomas 2000, 2006), trying to move the circadian research up the levels of organization to the interplay between cells, tissues, organs and organisms. Most of these calls for the return to the organism were reviews of all the studies showing that DNA is not enough - somewhat like this article is. The two Nature articles last week are just the latest research results in this tradition.

The power of metaphor

Where does this fundamental misunderstanding between molecular and organismal biologists come from? They are both biologists, right? So they should be expected to operate from the same basic principles.

But they don't. Geneticists come from a tradition starting with Schroedinger's 1944 book What is Life? This is a linear, hierarchical view of life, with upward causation: genes cause and control everything else. Also, gene is the only level on which natural selection acts (Dawkins 1976). The reigning metaphor of this worldview is the "program."

On the other hand, biologists coming from the study of evolution, ecology and animal behavior have a "systems" view of life in which many interacting elements, none of them with a primacy, determine the behavior of the entire system. There is no single element in control. The phenomena are a result of interactions, not of dominance of any particular actor.

The causation is downward (natural selection). DNA is just one of the elements in the system. Selection acts simultanously at several levels, including whole organisms and groups (Brandon 1996, Gannett 1976, Godfrey-Smith 1999, Griffiths and Gray 1994, Hubbard and Wald 1993, Keller 1995, Lewontin 1992, Nijhout 1990, Nelkin and Lindee 2004, Kitcher, P. 1999, Rose et al. 1990, are just a tip of the iceberg of the literature analyzing and criticizing the hierarchical DNA-first worldview). The reigning metaphor of this worldview is "the tangled bank".

Circadian field is not the only area of biology in which these two worldviews clashed. But it is worth noting here that the studies of clock genes ignored everything else, while the studies that questioned DNA supremacy never just shifted the control to some other element - all of those studies say that DNA is not sufficient, not that it is replaced by another controller.

Let's look at the "program" as a metaphor. A program is a term from information theory. It is a deterministic algorithm leading to a particular result. But what is reading that program? What is the "computer" that runs it? The cell?

And where is the person using the computer, the one who decides to run the program and decides if the program is useful or not? Where is natural selection?

Look at all the terminology of molecular biology: transcription, translation...those are all terms from information theory, which is linear, deterministic and hierarchical - there is a cause that controls the effect.

Even the "News and Views" article accompanying last week's two papers (Bass and Takahashi 2011) re-frames the results of the papers into information theory metaphor. All the stuff that is happening in the cytoplasm is referred to as "post-translational" as if it was just something more that DNA "caused," perhaps a little further downstream than usual.

But it is not. The cycles in the cytoplasm are not caused by anything any piece of DNA did. When in sync, the genetic feedback loops and cytoplasmic clocks work synergistically. But when placed in opposition, the cytoplasmic clock dominates (e.g., determines the phase, period, etc.).

The centrality of the gene in much of biological thinking led to another error that these two papers in Nature just fixed. Because different kingdoms of life (bacteria, protista, plants, fungi and animals) have different clock genes, it was assumed, despite the identical mechanistic logic of the mechanism, that the clock evolved independently several times. Identity of players trumped the mechanism of interaction between them.

But if, as the papers show, all organisms have cytoplasmic clocks based on anti-oxidant enzymes, then this cytoplasmic clock is the scaffolding, the base which allows evolution and replacement of all sorts of clock genes in different groups. As clock genes come and go, they can always latch onto the ever-present cytoplasmic clock. And the organism can keep on ticking regardless of the evolving stage in which any particular clock gene may be. This argues for a single origin of the circadian clock, due to universally adaptive nature of the clock, as postulated by Colin Pittendrigh decades ago.

The clock metaphor

The theory of biological rhythms has benefited immensely from the use of the clock as metaphor. Thinking of biological rhythms in terms of oscillatory theory (borrowed from physics) has allowed us to understand how the biological clock works, how it gets synchronized with the environment (entrainment), and how systems with multiple clocks can act together to produce higher-order phenomena (e.g., photoperiodism - measurement of seasonal changes in daylength).

The clock metaphor was also a key for understanding the mechanism as a collection of interacting cogs and wheels. This was crucial for the discovery of clock genes later on.

But once the number of clock genes was determined to be very small, and the interlocking feedback loops between them became the dominant paradigm for the mechanism, the meaning of the clock metaphor shifted - instead of looking at all the potential cogs and wheels, only those made of or from DNA counted. There is nothing wrong with counting everything - genes, and cytoplasmic elements, and the cell membrane, and the interactions between clock cells in a tissue - as cogs and wheels of the biological clock, but somehow, somewhere, we forgot that and settled for a DNA-only view.

Every metaphor that scientists invent has a heuristic value. The information theoretical thinking about genes sped up the research in genetics and molecular biology. The clock metaphor sped up the circadian research.

But it is always a good idea to sometimes step back and consider if the dominant metaphor is constraining in some ways, if it limits the imagination. I have argued before that an occasional switch to a different circadian metaphor - perhaps player-piano, or endless tape recorder, or Rube-Goldberg Machine, or camshaft, or Moebius strip - can be a good way to look at the problem from a new angle. This can be a very productive endeavor, opening one's eyes to new angles, starting new avenues of research. Every field of science has its metaphors, and it is always a good idea to sometimes analyze them, and sometimes replace them once they outlive their usefulness.

What metaphors are used by lay audience and the media?

There is a difference between metaphors used by scientists to guide their research programs, and metaphors used by journalists to explain research to lay audiences.

The clock metaphor, for example, means 'interlocking cogs and wheels to study' for researchers, but 'timepiece in your brain that tells you when to wake up and when to fall asleep' for the audience.

Likewise, the gene-control metaphor is something that is easy to understand for the audience that may be used to a hierachical worldview of top-down control (in society, family, religion, politics, or simplistic mechanics of everyday life). A systems-worldview requires a little bit more tolerance for ambiguity (which not everyone has) and a little more sophisticated understanding as to how complex systems work (i.e., how complex behaviors emerge out of interactions between multiple elements, in which the nature of interactions is more important than the identity and behavior of individual elements).

This is probably why the media reports could not capture the complexity of the findings. It provided an or option instead of an and option - the lay reader is probably going to think that DNA has nothing to do with the clock, instead of understanding that both DNA and other elements of the cell are partners. Still, in the media saturated with "gene for X" stories, an occasional "not in our genes" story is a positive event.

On the other hand, since the early 2000s (once the hype over Human Genome Project died down a little bit), the geneticists have moved away from the gene-control metaphor to some extent. Yes, they still sometimes slip up to old habits of mind (and their terminology shows it) - like when they use the term "post-translational" for everything that does not involve DNA in the cell - but the results of their own studies, from quantitative genetics to bumping into walls in some areas of research, have moved them to a more systems-like thinking. They are reinventing Physiology and calling it Systems Biology. And we are all better off for it. It is a more complete Physiology, with the 'black box' now wide open.

Another way that gene-primacy seeps into coverage of science is when new studies using molecular techniques are said to have confirmed the old studies using more traditional methods. For a recent example, see how the hypothesis of butterflymigration and speciation by Nabokov was said to have been confirmed by a recent molecular study. But molecular techniques are new, still being tested, calibrated and evaluated. The Nabokov story is really about well done work from the past using tried and tested old reliable techniques, that was strengthened by the new study and in turn validated the molecular method. Comparative anatomy is what validated the genetic method, not the other way round.

Likewise, in this example in clocks, it is very nice that new techniques repeated the old results. Each strengthens the other. The new study does not confirm the old as much as they all confirm each other. But for those enamored with molecules (or those who always think that new is better than old), this duo of papers will seal the deal if the old papers did not.

Conclusion

To summarize, the publication of these two studies in Naturelast week is, in my opinion, quite a milestone in the field. First, it showed how it was possible for the clock to originate only once on Earth yet evolve a number of different molecular elements - the cytoplasmic clock was there all along, keeping time while the genes swapped.

Second, it re-framed the discussion of the mechanism. It forcefully demonstrated what many prior studies did in small increments, but this time with modern techniques we love and with enormous power. By reminding the people in the field that DNA is an important but not sufficient element of the clock, it will hopefully guide future research in a new direction, with a more complete view of the clock, and perhaps may even allow some people to venture out and try other productive metaphors instead.

References:

ADAMICH, MARINA, PHILIP C. LARIS & BEATRICE M. SWEENEY (1976) In vivo evidence for a circadian rhythm in membranes of Gonyaulax, Nature 261, 583 - 585, doi:10.1038/261583a0

Barnett, Audrey (1966) A circadian rhythm of mating type reversals in Paramecium multimicronucleatum, syngen 2, and its genetic control, Journal of Cellular Physiology, Volume 67, Issue 2, pages 239–270

Bass, Joseph and Joseph S. Takahashi (2011) Circadian rhythms: Redox redux, Nature 469: 476–478, doi:10.1038/469476a

Brandon, R.N. (1996) Concepts and Methods in Evolutionary Biology. Cambridge University Press, Cambridge, UK.

Cornelius, Gerd and Ludger Rensing (1976) Daily rhythmic changes in Mg2+-dependent ATPase activity in human red blood cell membranes in vitro, Biochemical and Biophysical Research Communications. Volume 71, Issue 4, 23 August 1976, Pages 1269-1272

Dawkins R (1976) The selfish gene. Oxford University Press.

Dunlap, Jay C. (2008) Salad Days in the Rhythms Trade, Genetics, Vol. 178, 1-13, doi:10.1534/genetics.104.86496

Ehret, C. F. and E. Trucco (1966) Molecular models for the circadian clock : I. The chronon concept, Journal of Theoretical Biology, Volume 15, Issue 2, Pages 240-262, doi:10.1016/0022-5193(67)90206-8

Gannett, L. (1999). What's in a cause?: the pragmatic dimensions of genetic explanations. Biology and Philosophy 14: 349-374.

Godfrey-Smith, P. (1999). Genes ands Codes: Lessons from the Philosophy of Mind? In: Hardcastle VG, ed. Where Biology Meets Psychology. pp.305-331. The MIT Press.

Griffiths, P. and Gray, R. (1994). Developmental systems and evolutionary explanation. Journal of Philosophy 91:227-304.

Hardin, P. E., Hall, J. C. & Rosbash, M. (1990).Feedback of the Drosophila period gene product on circadian cycling of its messenger RNA levels. Nature 343, 536–540

HARTWIG, R., M. SCHWEIGER, R. SCHWEIGER AND H. G. SCHWEIGER (1985) Identification of a high molecular weight polypeptide that may be part of the circadian clockwork in Acetabularia, Proc. Nadl. Acad. Sci. USA, Vol. 82, pp. 6899-6902, October 1985

Hastings, Michael H., Elizabeth S. Maywood and John S. O’Neill (2008) Cellular Circadian Pacemaking and the Role of Cytosolic Rhythms, Current Biology 18, R805–R815, DOI 10.1016/j.cub.2008.07.021

Helfrich-Förster, Charlotte (2000) Differential Control of Morning and Evening Components in the Activity Rhythm of Drosophila melanogaster—Sex-Specific Differences Suggest a Different Quality of Activity, J Biol Rhythms, vol. 15 no. 2 135-154, doi: 10.1177/074873040001500208

Hubbard and Wald, (1993), Exploding the gene myth, Beacon Press.

Kitcher, P. 1999. The Hegemony of Molecular Biology. Biology and Philosophy 14: 195-210.

Keller, E.F. (1995). Refiguring Life: Metaphors of Twentieth-Century Biology. Columbia University Press, New York.

Kondo, T., Mori, T., Lebedeva, N.V., Aoki, S., Ishiura, M., and Golden, S.S. (1997). Circadian rhythms in rapidly dividing cyanobacteria. Science 275, 224–227.

Lakin-Thomas, Patricia L., (2006) Transcriptional Feedback Oscillators: Maybe, Maybe Not..., J Biol Rhythms, vol. 21, 2: pp. 83-92.

Lakin-Thomas, Patricia L. (2000) Circadian rhythms: new functions for old clock genes?Trends Genet 16:135-142.

Lewontin, R. (1992). The Dream of the Human Genome. New York Review of Books, May 28:31-40.

Loudon,Andrew S.I., Andrei G. Semikhodskii and Susan K. Crosthwaite (2000) A brief history of circadian time, Trends in Genetics, Volume 16, Issue 11, 477-481, doi:10.1016/S0168-9525(00)02122-3

Lowrey, Phillip L., Kazuhiro Shimomura, Marina P. Antoch, Shin Yamazaki, Peter D. Zemenides, Martin R. Ralph, Michael Menaker and Joseph S. Takahashi (2000) Positional Syntenic Cloning and Functional Characterization of the Mammalian Circadian Mutation tau, Science, Vol. 288 no. 5465 pp. 483-491, DOI: 10.1126/science.288.5465.483

Mabood SF, Newman PF, Nimmo IA. (1978) Circadian rhythms in the activity of acetylcholinesterase of human erythrocytes incubated in vitro, Biochem Soc Trans. 1978;6(1):305-8.

Mehra A, Hong CI, Shi M, Loros JJ, Dunlap JC, et al. (2006) Circadian Rhythmicity by Autocatalysis. PLoS Comput Biol 2(7): e96. doi:10.1371/journal.pcbi.0020096

MERGENHAGEN, D. and H. G. SCHWEIGER, (1975a) CIRCADIAN RHYTHM OF OXYGEN EVOLUTION IN CELL FRAGMENTS OF ACETABULARIA MEDITERRANEA, Experimental Cell Research 92:127-130

MERGENHAGEN, D. and H. G. SCHWEIGER, (1975b) THE EFFECT OF DIFFERENT INHIBITORS OF TRANSCRIPTION AND TRANSLATION ON THE EXPRESSION AND CONTROL OF CIRCADIAN RHYTHM IN INDIVIDUAL CELLS OF ACETABULARIA. Experimental Cell Research 94: 321-326

Mori, T., Binder, B., and Johnson, C.H. (1996) Circadian gating of cell division in cyanobacteria growing with average doubling times of less than 24 hours. Proc. Natl. Acad. Sci. USA 93, 10183–10188.

Mori T, Williams DR, Byrne MO, Qin X, Egli M, et al. (2007) Elucidating the Ticking of an In Vitro Circadian Clockwork. PLoS Biol 5(4): e93. doi:10.1371/journal.pbio.0050093

Nelkin and Lindee (2004) The DNA Mystique, University of Michigan Press.

Nijhout, H.F. (1990). Metaphors and the Role of Genes in Development. BioEssays 12:441-446.

Nitabach, Michael N., Todd C. Holmes and Justin Blau (2005) Membranes, Ions, and Clocks: Testing the Njus–Sulzman–Hastings Model of the Circadian Oscillator, Methods in Enzymology, Volume 393, 2005, Pages 682-693, doi:10.1016/S0076-6879(05)93036-X

Njus, David, Frank M. Sulzman & J. W. Hastings (1974) Membrane model for the circadian clock, Nature. Vol. 248, pp. 116-120, doi:10.1038/248116a0

Ohm-Schradera, L., G. Holzapfela and R. Hardeland (1980) Circadian rhythms in human erythrocytes in vitro not confirmed, Biological Rhythm Research, Volume 11, Issue 3 November 1980 , pages 199 - 207

O'Neill JS, & Reddy AB (2011). Circadian clocks in human red blood cells. Nature, 469 (7331), 498-503 PMID: 21270888

O’Neill JS., Gerben van Ooijen, Laura E. Dixon, Carl Troein, Florence Corellou, François-Yves Bouget, Akhilesh B. Reddy and Andrew J. Millar, (2011). Circadian rhythms persist without transcription in a eukaryote, Nature, 469 (554–558), doi:10.1038/nature09654

Page, Terry (2000) A Novel Mechanism for the Control of Circadian Clock Period by Light, J Biol Rhythms April 2000 vol. 15 no. 2 155-162, doi: 10.1177/074873040001500209

Peleg, L., A. Dotan, P. Luzato and I. E. Ashkenazi, (1990a) Hexose monophosphate shunt activities in human erythrocytes during oxidative damage induced by hydrogen peroxide, In Vitro Cellular & Developmental Biology - Plant, Volume 26, Number 10, 978-982, DOI: 10.1007/BF02624472

Peleg, L., A. Dotan, P. Luzato and I. E. Ashkenazi (1990b) Long ultradian rhythms in red blood cells and ghost suspensions: Possible involvement of cell membrane , In Vitro Cellular & Developmental Biology - Plant, Volume 26, Number 10, 978-982, DOI: 10.1007/BF02624472

Roenneberg,Till and Merrow, Martha, (1998) Molecular Circadian Oscillators: An Alternative Hypothesis, J Biol Rhythms, vol. 13, 2: pp. 167-179.

Roenneberg,Till and Merrow, Martha, (1999) Circadian Systems and Metabolism, J Biol Rhythms, vol. 14, 6: pp. 449-459.

Roenneberg,Till and Merrow, Martha, (2005) Circadian clocks — the fall and rise of physiology. NATURE REVIEWS MOLECULAR CELL BIOLOGY, VOLUME 6: 965-971

Rose, Steven, Richard C. Lewontin, Leon J. Kamin (1990) Not in our genes: biology, ideology and human nature, Penguin Books

Runft, Linda L., and Dina F. Mandoli (1996) Coordination of cellular events that precede reproductive onset in Acetabularia acetabulum: evidence for a ‘loop’ in development. Development 122, 1187-1194

Schroedinger, E. (1944). What is Life? Cambridge University Press, Cambridge, UK.

SCHWEIGER, E., H. G. WALRAFF, and H. G. SCHWEIGER (1964b) Endogenous circadian rhythm in cytoplasm of Acetabularia: Influence of the nucleus. Science 61548-665:9.

SWEENEY, B. M. (1974) A physiological model for circadian rhythms from the Acetabularia rhythm paradoxes. Int. J. Chronobiol. 2: 25-33.

SWEENEY, B. M., and F. T. HAXO (1961) Persistence of a photosynthetic rhythm in enucleated Acetabularia. Science 134: 1361-1363.

TERBORGH, J., and G. D. McLEOD (1967) The photosynthetic rhythm of Acetabularia crenulata: I. Continuous measurements of oxygen exchange in alternating light-dark regimes and in constant light of different intensities. Biol. Bull. 133: 659-669.

Tomita, Jun, Masato Nakajima, Takao Kondo and Hideo Iwasaki (2004) No Transcription-Translation Feedback in Circadian Rhythm of KaiC Phosphorylation, Science. Vol. 307 no. 5707 pp. 251-254, DOI: 10.1126/science.1102540

Underwood, Herbert and Michael Calaban (1987) Pineal Melatonin Rhythms in the Lizard Anolis carolinensis: I. Response to Light and Temperature Cycles, J Biol Rhythms, vol. 2 no. 3 179-193, doi: 10.1177/074873048700200302

VANDEN DRIESCHE, T. (1966) The role of the nucleus in the circadian rhythms of Acetabularia mediterranea. Biochim. Biophys. Acta 126: 456-470.

VANDEN DRIESSCHE, Therese, and SILVANO BONOTTO, (1969) THE CIRCADIAN RHYTHM IN RNA SYNTHESIS IN ACETABULARIA MEDITERRANEA, Biochim. Biophys. Acta, 179: 58-66

Woolum JC (1991) A re-examination of the role of the nucleus in generating the circadian rhythm in Acetabularia. J Biol Rhythms 6:129-136.Acta 126: 456-470.

Image credits: Red blood cells, Wikimedia Commons; Ostreococcus tauri, The Joint Genome Institute; Acetabularia crenulata, The College of Exploration; graphs - from O’Neill JS, & Reddy AB (2011) and O’Neill JS. et al. 2011; and Transcription-translation Feedback Loop model, Nature Reviews Neuroscience