Clocks, metabolism, evolution – toward an integrative chronobiology

The biannual meeting of the Society for Research on Biological Rhythms happened last week. Unfortunately, I could not attend, so will have to wait another two years for the next opportunity.

I am not sure how this stuff happens, but there was a flurry of new papers in the circadian field just preceding the event. Several of them have already received quite a lot of attention in both old and new media, and rightfully so, but I decided not to cover them one at a time just as the embargo lifted for each one of them.

Instead, I will just very briefly describe and explain the main take-home messages of each one of them, link to the best coverage for those who want more detail (“Cover what you do best. Link to the rest.“), and then try to come up with more of a ‘big picture’ summary of the current state of the field.

I apologize in advance for covering and linking to some of the papers that are not published in Open Access journals. I am not as strict about this policy as some other bloggers are (“if my readers cannot access it, they cannot fact-check me”), and will occasionally cover non-OA papers. Even if most of my readers cannot access them, I gather that a miniscule proportion can access and, if I got something wrong, can alert the other readers in the comments. And speaking of Open Access, I am not one to sign many online petitions, but this one is worth it so please sign if you have not done it already.

So, let’s see what new and exciting in chronobiology these days…



Article: Ben Collins, Elizabeth A. Kane, David C. Reeves, Myles H. Akabas, Justin Blau, Balance of Activity between LNvs and Glutamatergic Dorsal Clock Neurons Promotes Robust Circadian Rhythms in Drosophila, Neuron, Volume 74, Issue 4, 706-718, 24 May 2012

What is it about: Robustness of daily rhythms and their flexible, adaptive responses to the environment, require a feedback loop between a cluster of clock cells in the brain and another cluster of non-clock cells in the brain of Drosophila melanogaster.

What is new: Feedback loops between two or more brain centers (or tissues, or organs) as necessary for either existence of some circadian rhythms, or for the rhythms’ robustness and fine-tuned response to the environment, have been studied mostly in vertebrates, especially birds and lizards, and to some extent mammals. Such feedback loops have been found in the fruitfly as well. This paper finds out a lot of detail about this feedback loop, including the use of glutamate as a neurotransmitter in one half of the loop. As Drosophila is still the lab organism with the most developed techniques for precise genetic manipulations, this is an important advance.

Take-home message: How core clock genes turn each other on and off within a cell over 24 hours is just the beginning, a small part of the story. To work properly, to be adaptive, and to respond well to environmental cues, circadian rhythms require organization at a higher level, with fine-tuned communication among clock-cells and between clock and non-clock tissues.

Some more thoughts: This is a technical tour-de-force. Fruitfly genetics techniques today are so powerful and this paper appears to use them all: inserting, deleting, downregulating and upregulating genes of choice in precisely targeted cells in the brain. As every behavioral biologist knows, once introduced into an experiment animals do whatever the heck they please. The behavior measured in this paper was a simple light-avoidance test – fruitfly larvae (and in the last experiment, also adults) are placed in a petri dish that is half in light, half in darkness, and the movement and position of the larvae is monitored. Considering how messy such behavioral data tend to be, the results in this paper are quite impressive.

The Abstract/Summary, the Introduction, and the (far too short) Discussion are very clear, straighforward and easy to read and understand. They are also upfront and direct about their main take-home message. The many pages in-between, though, are clearly meant to be read only by Drosophila clock geneticists who can actually wade through the essentially endless litany of acronyms in hope of replicating or following up on this study. Clocks are my field, but I am not a geneticist or drosophilist, so much of the Materials & Methods and Results sections in this paper are over my head. Maddeningly, some of the most important stuff is hidden in the Supplemental Materials, including this model for how the whole thing works (shouldn’t this image be up front, on the top of the whole thing?):

Drosophila neuron model

Drosophila neuron model

Note: If I remained in research, I would have done something like this, not necessarily in fruitflies, but definitely looking at neural networks, feedback loops and higher-level organization of the circadian system, within ecological and evolutionary contexts. This may bias me toward liking this paper as much as I do.

Good coverage elsewhere: None that I can find. Only a warmed-up (and not that good) press release at Futurity and ScienceDaily.



Article: Rachel S. Edgar et al., Peroxiredoxins are conserved markers of circadian rhythms, Nature, Published online 16 May 2012, doi:10.1038/nature11088

What is it about: A protein (peroxiredoxin) that is found in almost all living organisms has two states/conformations that cycle at approximately 24 hours. Presence and proper function of core circadian clock genes is not necessary for the cycling of this protein. An Archea species that does not live on the surface does not have the clock and does not have this protein.

What is new: This finding in human red blood cells and a Protist (O.tauri) was published last year. This paper adds similar data for a whole bunch of other organisms: cyanobacteria, archaea, fungi, plants, insects and vertebrates.

Take-home message: There are really two take-home messages, one physiological, one evolutionary. First, this demonstrates that circadian clocks are properties of the entire cells (or assemblages of cells in case of multicellular organisms), not just the transcription/translation loops of core clock genes.

Second, the protein in question, the peroxiredoxin, could be akin to “scaffolding”, something that allows a cell to keep cycling while genes come and go, mutate and change and duplicate, while being fine-tuned by natural selection. Over billions of years, this can result in major groups of organisms (e.g., animals vs. plants vs. fungi vs. bacteria, vs. several different groups of protists) having entirely different circadian genes, yet all of them using the same “logic” (transcription/translation feedback loops).

Both the peroxiredoxins and the circadian clocks are thought to have originated as defense from UV radiation of the early oceanic surfaces (Pittendrigh, C.S., 1967. Circadian rhythms, space research, and manned space flight. In: Life Sciences and Space Research 5:122-134. North-Holland, Amsterdam.), or defense against other kinds of demage, including that from oxidation, so it makes sense that they co-originated and co-evolved only once in the history of the planet, perhaps around 2.5 million years ago when photosynthetis bacteria introduced lots of molecular oxygen into the Earth’s atmosphere. It also makes sense that they both are missing in organisms that have never lived close to the oceanic or terrestrial surface (e.g., many Archaea).

Essential reading: When the two papers (on red blood cells and the protists) came out last year, I wrote a very comprehensive post that places this research direction into historical, philosophical, methodological and even media context. There is not much in that post that would change with the publication of this new paper apart from additional confirmation in several new species. So just go and read it again.

Some more thoughts: Peroxiredoxins cycle in all kinds of different cells with an approximately 24 hour period. This makes them, almost by definition, circadian clocks. Last year’s papers also show that the peroxiredoxin clock dominates its phase over the clock driven by core circadian genes. But there is something still to find out, and it is important: can the peroxiredoxin clock drive any other rhythms? For it to work as a biological pacemaker, it is not sufficient for it to cycle itself, it also need to drive timing of other events in the cell (and the entire organism). I am assuming that this research group will look at this problem next.

Mammals have six peroxiredoxin genes. In an experiment (Zhang et al., Cell , 2009), human cell lines were engineered in such a way that each culture had a different peroxiredoxin gene knocked out. None of the knock-outs had any effect on the regular circadian expression of the core clock gene Bmal1. Of course, having six genes indicates redundancy in function. One would need to knock out all six simultaneously in order to see an effect on other rhythms in the cell, but there is a question if cells with all peroxiredoxins knocked out can survive at all. Someone should try this.

Also, that experiment was done in mammalian cells. Mammals are probably the worst model system for studying this question. Vertebrates have undergone several events of gene (and genome) duplication, and mammals got at least another one. If you look at mammals, every clock gene exists in multiples (e.g,. Per1, Per2, Per3). Poor peroxiredoxin probably cannot do much in such a massively genetically determined system.

Gene duplication allows for evolutionary experimentation. As long as one copy of the gene keeps working, the others are free to mutate. Some mutations will be selected against (e.g., if they mess up the clock function) and others will be selected for (e.g., if they fine-tune the circadian function, making it more flexible and adaptable, or start performing some other valuable function instead). This means that functions formerly in the domain of higher-level organization or the domain of phenotypic plasticity, are now under control of genes. This process is called Genetic assimilation (and sometimes Baldwin effect, though that term is usually reserved for genetic assimilation of learned behaviors). So it is quite possible that the clock in mammals is over-determined by genes, making it useless for the study of peroxiredoxins as scaffolding for circadian evolution.

If I was doing this research, I would stay away from these vertebrate oddballs for at least the next five or ten years, and focus my time, funds and energy on the study of bacteria, archaea, protists, fungi and perhaps some plants – smaller the genome the better.

Good coverage elsewhere: Great coverage by Ed Yong, Megan Scudellari and Ewen Callaway. Was also covered by Debora MacKenzie.

Additional reading: Whence Clocks? and Clock Evolution


Article: Faure, S., Turner, A.S., Gruszka, D., Christodoulou, V., Davis, S.J., von Korff, M. & Laurie, D.A. Mutation at the circadian clock gene EARLY MATURITY 8 adapts domesticated barley (Hordeum vulgare) to short growing seasons, Proceedings of the National Academy of Sciences, DOI: 10.1073/pnas.1120496109

What is it about: In barley cultivars from Northern Europe a mutation in a gene responsible for flowering feeds back on the circadian clock genes, greatly reducing the amplitude of the gene cycling, effectively shutting down the clock. Without proper clock function, barley does not use the clock to measure seasonal changes in daylength (photoperiodism) but instead matures at the fastest rate its development permits. This allows barley to mature and flower early in the season, as well as to photosynthetise throughout the long days of summer in the North.

What is new: Yet another organism in which some of the clock function is temporarily or permanently eliminated. Good news: unlike the other such organisms which tend to be not-well-studied inhabitants of extreme environments, barley is a domesticated plant, well researched and easy to use in the lab.

Take-home message: one has to be careful with interpreting studies like these – just because an organism does not show a couple of well-studied rhythms in physiology and behavior, and does not show cycling in expression of core clock genes does not mean that all circadian function is gone. Ensembles of cells, or feedback loops between tissues, or cytoplasmic factors like peroxiredoxin may still be working in the organism, it’s just that this cannot be detected with the techniques used in the study.

Good coverage elsewhere: As far as I know, I am the only one who covered this paper.



Article: Summa KC, Vitaterna MH, Turek FW (2012) Environmental Perturbation of the Circadian Clock Disrupts Pregnancy in the Mouse. PLoS ONE 7(5): e37668. doi:10.1371/journal.pone.0037668 (Open Access)

What is it about: Female mice kept in 24 cycles get and remain pregnant easily. Female mice kept in rotating shifts (either advances or delays of the 24 hour cycle by 6 hours every several days) do not. Difference is striking!

What is new: Interactions between circadian rhythms and reproductive cycles have been studied for decades in many different organisms. Last year, a study with already pregnant mice moved to rotating shifts did not result in spontaneous losses of pregnancy. This study suggests that rotating shifts prevent pregnancy to begin in the first place, probably by interfering with implantation of the egg in the uterus.

Take-home message: Another example how clock is not just about timing of downstream events, but also plays part in them more directly. While this study was done in rodents, it works well together with epidemiological data from humans working on rotating shifts. As the light-dark cycle is shifted, the brain clock resets itself pretty quickly, over a period of a couple of days. But peripheral clocks in all the other organs will reset slower, each at its own rate. This include the ovaries, uterus etc, which may not be ready for egg implantation at the time of day when the brain send the relevant signal. In essence, internal desynchronization of clocks prevents all the parts of the system to work in synchrony – this is the main negative effect of jet-lag, and it applies to reproduction as much as it does to digestion and other functions.

circadian phaseshift pregnancy

Effect of entrainment on success of pregnancy

Some more thoughts: Both the paper and the media coverage are clear, straightforward and readable. But if one wanted to explain these data by building a formal/conceptual or mathematical model this could easily get mind-bogglingly complicated: one would have to take into account multiple feedback loops between repeatedly desynchronized oscillators, plus potential effects of photoperiodism.

Good coverage elsewhere: Sarah Fecht did a great job. Most of the rest of the media just regurgitated the press release. Oh, this was even covered by Daily Mail 😉

See also: Oxytocin and Childbirth. Or not.



Article: Johnni Hansen, Christina F Lassen, Nested case–control study of night shift work and breast cancer risk among women in the Danish military. Occup Environ Med doi:10.1136/oemed-2011-100240

What is it about: A large-scale study of Danish female soldiers found a higher incidence of breast cancer in those who had to work night shifts. Longer the period of night-shift work, greater the incidence. Also, early risers were more susceptible to this negative effect of night shift work.

What is new: Earlier studies were mostly done in nurses in the USA. This provides a much larger data-set of women followed over a long period of time in a different profession in a different country. Military environment also controls for many other aspects of life (food, quality of medical care, physical fitness, etc.) which tends to be more uniform than in the general population.

Take-home message: Prolonged night shift work, especially if you are an early bird, may be bad for your health.

Some more thoughts: Internal desynchronization between various body clocks, especially long-term, is bound to have negative consequences. Suffering from jet lag occasionally when traveling is fine. But getting jet-lagged every day for years is seriously impairing all sorts of body functions (see reproduction above).

Good coverage elsewhere: Steven Reinberg



Article: Yuta Fuse, Akiko Hirao, Hiroaki Kuroda, Makiko Otsuka, Yu Tahara and Shigenobu Shibata, Differential roles of breakfast only (one meal per day) and a bigger breakfast with a small dinner (two meals per day) in mice fed a high-fat diet with regard to induced obesity and lipid metabolism. Journal of Circadian Rhythms 2012, 10:4 doi:10.1186/1740-3391-10-4 (Open Access)

What is it about: Three groups of mice were fed a high-fat diet, each group getting exactly the same amount of food each day (and eating it all up each day). One group had free access to food at all times (and ate all the time). The second group was given food in a limited time regimen: a large breakfast and a small dinner. The third group was given all the food to gobble up in one large brekfast. The group that got only a large breakfast got obese, had other metabolic problems and had a disrupted expression of circadian genes.

What is new: Another interesting paper showing that timing of meals determines how the food is processed by the body.

Take-home message: Eating one big meal per day is bad for your health – spread it out a little.

Some more thoughts: The paper is interesting as the data suggest something different from most of the other papers in this line of research (see the next two papers below).

Good coverage elsewhere: I could not find any.


Article: Megumi Hatori, Christopher Vollmers, Amir Zarrinpar, Luciano DiTacchio, Eric A. Bushong, Shubhroz Gill, Mathias Leblanc, Amandine Chaix, Matthew Joens, James A.J. Fitzpatrick, Mark H. Ellisman, Satchidananda Panda, Time-Restricted Feeding without Reducing Caloric Intake Prevents Metabolic Diseases in Mice Fed a High-Fat Diet. Cell Metabolism, 17 May 2012 doi:10.1016/j.cmet.2012.04.019

What is it about: Groups of mice were fed either normal or high-fat diet either with continuous free access to food or with feeding time limited to 8 hours during the night (remember that mice are nocturnal – this is their active period, i.e., their “day”). The results appear to be opposite from the paper above (by Fuse at al.) – it is the mice with unlimited feeding that got obese and developed metabolic problems, as well as reduced amplitude of the circadian gene expression.

What is new: Hmmm, which one of the two papers is “more right” than the other? The devil is in the details, so we’ll have to look there.

There are two obvious differences between the two papers. The Hatori paper gave full volume of the normal daily intake of food, while the Fuse paper gave mice only 80% of the normal food quantities per day – which is calory restriction in itself. This may explain why the free-feeding mice in Hatori paper got obese and developed problems, while the mice in the Fuse paper did not.

Second, there is a difference in timing of meals in time-restricted groups. There is “time-restricted feeding” and then there is “time-restricted feeding”! The Hatori paper restricted feeding to an 8-hour period starting one hour after lights-off and ending three hours before lights-on. The Fuse group gave breakfast at the moment of lights-off (the paper does not say for how long – presumably with reduced diet the hungry mice just ate it all very fast) and a smaller “dinner” at the moment of lights-on. These are very different timing schedules!

In many ways, the two-meal schedule of the Fuse paper is similar to the time-restricted schedule of the Hatori paper. Note that these two schedules did the best in regard to obesity and metabolism. Both the free-feeding (especially with the full diet in the Hatori paper) and extremely restricted feeding (brief but huge breakfast in the Fuse paper) resulted in bad metabolic effects. One can perhaps conclude that extremes are bad – one huge meal is bad as is continuous grazing, but that the spread of feeding over two or more smaller meals does better.

Take-home message: The perennial “more research is needed”….until then it is wise to eat your meals at normal times, more than once per day, no grazing in-between, and no midnight snacks…

Hatori image

Effect of feeding regimen on body weight and metabolism

Some more thoughts: Hatori paper is….overwhelming! There is so much work done. As I was reading it all, my thought was that ten pages in the middle of the paper could be completely cut out of the paper and the result would still remain exactly the same – just weigh the mice! All sorts of things were measured in a variety of ways, from gene expression patterns, to standard metabolic tests to histology. All that work strengthens the notion that obesity and metabolic problems are correlated, so the work is definitely not for nothing, and is very impressive. It certainly adds a lot of information to the notion that circadian clock is not just a timer, but intimately involved in regulation of metabolism.

Good coverage elsewhere: Peter Janiszewski, Ph.D. and Garth Sundem. Also see Michael Coston. There was plenty of coverage in mainstream media, mostly OK, some bad…


Article: Till Roenneberg, Karla V. Allebrandt, Martha Merrow, Céline Vetter, Social Jetlag and Obesity. Current Biology, Volume 22, Issue 10, 939-943, 10 May 2012. doi: 10.1016/j.cub.2012.03.038

What is it about: A huge number of Europeans from various ages, latitudes and longitudes were assessed for a variety of circadian, sleep and health parameters. It turns out that most suffer from “social jet-lag” – an internal desynchronization between various body clocks as a result of continuous mismatch between the natural body rhythms and societally and culturally imposed wake-up, school, work and bed-time schedules. During the school/work week, people are massively sleep deprived. They then make up for it, “paying off the sleep debt”, over the weekends. The difference between the amount of sleep one gets on work/school nights and the amount of sleep one gets during the weekend is especially stunning in adolescents, whose internal clocks are naturally phase-delayed and thus most dramatically out of sync of what the society is forcing them to do:

roenneberg image

Average sleep time by age and gender (top) and discrepancy between workday and weekend sleep (bottom)

One of the most remarkable results is that people with the largest workday/weekend discrepancy, the most socially jet-lagged individuals, are also most prone to smoking, drinking, obesity and other health problems. Naturally thin people (B) did not get obese if they were more sleep deprived, but those somewhat prone to obesity became obese if they were also sleep deprived due to social jet-lag:

roenneberg image2

Effect of social jetlag on obesity in obesity-prone people (top) and thin people (bottom)

What is new: Bits and pieces of this were known for a while. But nobody has ever done such a tremendous study on such a large number of people, controlling for so many factors and measuring so many parameters and so many outcomes. This is definitely a tour-de-force paper of the year in this field and is appropriately matched with a brand new book by the lead author – Internal Time: Chronotypes, Social Jet Lag, and Why You’re So Tired – which I am currently reading.

Take-home message: Socially imposed school and work schedules are messing with our health. Political will is needed to change the mindset, change the culture, and change the way we use time in the society.

Some more thoughts: This paper goes together well with the several papers described above – living a life outside of synchrony with the light-dark cycle of the natural environment (and no, artificial indoor light cannot match it) has serious health consequences, leading to metabolic, physiological, reproductive and psychological problems that negatively affect billions of people on Earth, and cost the society billions of dollars of lost productivity, unnecessary medical care, and loss of educational potential in teenage students. Watch the video:

Note: Again, and infuriatingly so, most of the interesting stuff is in the Supplemental Materials.

Good coverage elsewhere: Maria Popova, Jamie Condliffe, Robert T. Gonzalez, Allison Aubrey, Kate Southam and many others.

Additional readings: When Should Schools Start in the morning? and Everything You Always Wanted To Know About Sleep (But Were Too Afraid To Ask) and Sun Time is the Real Time.



Genes are, for the most part, invisible to natural (and sexual) selection.

What evolution can work on are phenotypes – composites of anatomical, physiological and behavioral traits as they change during the development and lifetime of an organism. Genes are selected for, indirectly, inasmuch as they contribute to the phenotype. While in different organisms and in different cases selection may act on a number of different levels – genes, embryos, cells, organs, organisms, groups, species – usually the most important unit of selection is the individual organism.

How a gene contributes to the phenotype is affected by many factors – multiples ways of splicing it, multiple ways of post-translational modification, which other genes are present, where and when is it expressed during development, interactions between cells, tissues and organs, and interactions between the organism and its environment. As genes, chromosomes and genomes sometimes get duplicated, this provides more opportunities for traits, previously resulting from higher-level interactions, to get incorporated into the genetic instructions via the process of genetic assimilation.

The basic unit of life is the cell. A single-celled organism has to have all of its internal processes coordinated in order to display adaptive responses to the environment in order to survive and reproduce. In multi-celled organisms, each cell type has to behave properly, communicate properly with the other cells, and coordinate its activities with all the other cells in the body for the organism to display adaptive responses to the environment in order to survive and reproduce.

Molecules that do all, or almost all of the work are proteins. They build structures, they catalyze reactions, break down food, store and use energy, control communication between cells, regulate the expression of the genes, and more.

But proteins are hard to study! Nucleic acids are much easier – they are stable, inert, the 3D conformation does not matter, and laboratory techniques have been developed to discover, sequence and study snippets of DNA and RNA in ways that today can be done by middle-schoolers or in DIY science projects. When you have the hammer, everything is a nail. When you have genetic tools, your graduate students are instructed to use them. And then sometimes they forget why they are doing this in the first place…

We study nucleic acids because they are markers, or proxies, for proteins. We locate genes and hope that the processes of transcription, splicing and translation do not confound too much what the resulting proteins are and what they do. This means that study of genes – their sequences and patterns of expression – is a reasonable first step in studying a biological phenomenon as it provides us with tools and information needed to study proteins, cells and higher-order phenomena that are evolutionarily relevant.

Of the Big Four – anatomy, biochemistry, physiology, behavior – it is behavioral traits that are the furthest removed from the underlying genes. It is really difficult to find genes directly involved in behavior. Big screens for genes for behaviors usually come up with genes for kinases, neurotransmitter receptors, neuronal development factors and other generalized components of the nervous system.

Circadian clock is an exception. While one may argue that the clock is not actually a behavior but a physiological mechanism that regulates many other behaviors, it is still the closest to a behavioral trait we ever got in discovering underlying genetics. Most people in the field agree that all major genes involved in clock function have been discovered (mostly during the 1990s) and that it is not productive to search for more.

Sure, the Abstract book of last week’s SRBR meeting still contains some posters and blitz-sessions by students with a detailed genetics work (well, the students need to learn the techniques, right?), but most of the Big Honchos of the field have moved on – to properties of the entire cells, neural networks, properties of multi-clock systems, and interactions with the environment. In a sense, after a detour of the 1990s when all the focus and energy was placed into gene discovery (“opening the black box”), the field of chronobiology is going back to its roots – a historically incredibly comparative and integrative subdiscipline of biology.

For a long time we have thought of circadian clocks as simply timers – something that determines when other downstream functions happen. But evidence over the last couple of decades has accumulated that clocks are much more intimately involved in some of those functions, beyond just timing.

Ten years ago, when I was exiting the field, the interaction between clocks and metabolism was just starting to be explored. We learned that fruitfly clock-gene timeless is involved in cocaine addiction. We learned that mutations in clock genes that change circadian period also change other aspects of timing, from frequencies of fruitfly courtship songs, to developmental timing in nematodes, to photoperiodic responses, to reproductive cycles.

Today, involvement of the circadian clock in many aspects of metabolism is probably the most exciting and most heavily studied area in the field. And this is shown by all the papers I highlighted today. The focus away from identification of genes and moving on to proteins, cells, neural networks, multi-oscillatory systems, and interactions with the environment are making the field as exciting as ever, and in vanguard of much of the rest of biology which is still overly focused on DNA. And the findings have obvious and stark implications not just for our better understanding of Life, but also for understanding of adaptation of organisms to the changing climate, and for understanding the consequences on human health.


Comments are closed.