I wrote this post back on January 23, 2005. It explains how clock biologists think and how they design their experiments:
So, are you ready to do chronobiological research? If so, here are some of the tips – the thought process that goes into starting one’s research in chronobiology.
First, you need to pick a question. Are you interested in doing science out of sheer curiosity to discover stuff that nobody knew before (a very noble, but hard-to-fund pursuit)? Or would you prefer your work to be applicable to human medicine or health policy, veterinary medicine, conservation biology, or agriculture?
Do you want to figure out some nitty-gritty details about the molecular basis of circadian rhythms, or neural connections between pacemakers in a mammalian brain? Perhaps you want to know if a cave animal still has a functioning clock, or if a microorganism has a clock at all. Or perhaps, you are interested in tests of adaptive function of biological rhythms and want to do your research out in the field. You may be looking for an organism to evaluate if it can become a new model, or you already know which of the current model organisms you are going to use.
Are you a kind of person who revels in a competitive area, rushing to publish as many papers as quickly as possible, each paper certain to add just a little bit to the current knowledge, all the papers accumulating reputation for you in the long run? Or are you a kind of person who prefers to tackle a difficult and risky project with no guarantee of success but a high return if it does succeed in the end, a project that may take years to accomplish while you live in obscurity, but can potentially result in a minor revolution in the thinking within the field once it is done? Think hard about the kind of personality you are and how thick is your skin.
No matter what your project is going to be, if you are in chronobiology, you have to be able to continuously monitor biological rhythms in your organisms for at least several cycles. Rare are the experiments in which you can make do without it. For instance, if you are looking at circadian rhythms of gene expression, you still need to know the phase of the cycle at which you are taking your samples. You can only know the phase if you are monitoring some kind of output of the circadian system.
The most widely used overt rhythms in laboratory research are behavioral rhythms, e.g., gross locomotor activity, wheel-running activity in rodents (and cockroaches), perch-hopping rhythms in passerine birds (songbirds), tube-running behavior in fruitflies, feeding or drinking rhythms in some other animals, etc. The advantages of behavioral rhythms are the ease and low cost of monitoring them (a LED diode, or a simple switch will do the trick). The disadvantages are the sensitivity of behaviors to various environmental events. For instance light (or darkness) may directly stimulate (or inhibit) behavior regardless of the phase of the clock. Mice are much more light-shy than rats, for instance, though both are nocturnal rodents. Darkness may inhibit feeding in some diurnal animals. The ability of an environmental cue to directly induce changes in the measured output is called masking, and one needs to be aware of this in one’s model animal, either from published literature, or by running a set of experiments to determine the appearance and/or intensity of masking. If masking effects make the project impossible, one should lookk at other possible outputs, e.g., body temperature, heart-rate, blood-pressure, oxygen-consumption, blood levels of a hormone (e.g., melatonin). The best thing to do, if technically feasible, is to simultaneously monitor two or more overt rhythms in each animal.
Here is an example of a laboratory setup for studying circadian rhythms in rodents. A hamster, in this case, is housed in a cage that contains a running wheel. The running wheel has a switch that registers every revolution of the wheel and sends that information to a computer. The computer puts a time stamp on each data point, and collects the data over long periods of time.
For visual analysis of the data, computer software was developed that presents the data in a graphical format called an actograph. As you can see in the figure above, an actograph has 24 hours of the day plotted on the X axis. The data from the first day are plotted on the top, the second day is plotted immediatelly below the first day, the third day data below the second, etc. Each time point (e.g., in 10-minute bins) is depicted either as white or black. White denotes times when wheel was not moving. Black denotes times when the hamster was running in the wheel.
Once the whole actograph is printed out, one can see long-term patterns. In this example, the hamster was kept in a light-dark cycle (LD) with 12 hours of light (from 7am till 7pm) and 12 hours of darkness (LD 12:12). From the actograph, we can see that the circadian clock driving wheel-running in this hamster has entrained (synchronized) to the LD cycle – the hamster was running in the wheel at the beginning of each night. However, at his point, we are still not certain that what we see is the real output of the circadian clock, as the same pattern would emerge if light exerted a masking effect on behavior by inhibiting wheel-running. Thus, at this point in our research project, the rhythm is properly called a diurnal rhythm, not a circadian rhythm (the insect folks like to call this diel rhythm).
How do we know if the observed rhythm is really circadian? By releasing the animal into constant conditions, usually constant darkness (DD). In some animals, DD inhibits feeding, thus we have to use very dim constant light (dLL), often of a single wavelength (e.g., green). In photosynthetizing organisms, like plants, we usually use constant light (LL). Here is an example of a diurnal animal (a gallinaceous bird) kept initially in LD cycles, then released into DD. The measured overt rhythm here is core body temperature (measured by radiotransmitters implanted into its abdominal cavity, with a receiver and a computer translating transmitted radiofrequqncy into degrees Celsius). White represents body temperature below the mean temperature of that particular day, while black represents times (in 10-minute bins) when the temprature was above the daily mean.
Notice how the temperature rhythm entrains to the LD cycle but, after release into DD (after about two weeks), continues to cycle indefinitely. This is called a freerunning rhythm. The morning rise of body temperature occurs a little bit earlier every day, thus the inherent, endogenous, genetically determined period of the freerunning rhythm of this bird is shorter than 24 hours. Actually, in this example, it is about 22.5 hours.
Another way to plot data is a strip chart. This method allows one to plot only a rather small number of days/cycles, but has an advantage of showing the amplitude and the exact shape of the rhythm. For instance, this is a strip chart of human body temperature over just a single cycle.
Actually it is more than a simple strip chart, as it is a composite of four groups of humans. Each data point is an average of measurements taken at that particular time point from the whole group. Further, each group has been held at a different ambient temperature. Comparison of the four strip charts all plotted together tells us the amount of masking that environmetal temperature can exert on the clock-controlled rhythm of core body temperature.