Experimental Paradigms

Experiments can be planned with different goals in mind. In particular, considering a Drosophila mutant or wild-type strain, one may be interested in determining, for example, (1) the natural free-running period of the strain, (2) the capacity for such individuals to adapt to different light-dark (LD) cycles, and (3) the phase of the circadian locomotor activity cycles with respect to the zeitgeber (see Note 10) used. In each of the above cases, the experimental design will differ accordingly. For an excellent recent review on various aspects of entrainment see, for example, ref. 19. It is also important to note that the following subheadings contain information that applies equally well to nonphotic entraining stimuli (such as temperature, for example).

LD 12:12

LD 12:12

LD 12:12

LD 12:12

DD

DD

LD 12:12

LD 16:8

LD 16:8

LD 12:12

LD 8:16

LD 8:16

Fig. 2. Schematic representation of different types of light-dark (LD) cycles showing, in each case, examples of how it is possible to change from 12 h:12 h LD to the other regimes. As explained in the text, a typical experiment requiring switching from one LD regime to another, would entail at least 3 d of 12 h:12 h LD, before switching to the new cycle. For brevity, B-D show only the last of 3 d of 12 h:12 h LD. (A) 3 d of 12 h:12 h LD cycles. (B) 1 d of 12 h:12 h LD, followed by 2 d of constant darkness; here the light gray bars represent the subjective daylight and the dark gray bars the subjective night. (C) 1 d of 12 h:12 h LD, followed by 2 d of 16 h:8 h LD; (D) 1 d of 12 h:12 h LD, followed by 2 d of 8 h:16 h LD.

3.3.1. Determining Natural Free-Running Period (t) of a Drosophila Strain

1. Initially entrain, for instance, with 12 h light:12 h dark (12:12 LD) cycles for at least 3 to 4 d.

2. Continue the experiment for at least another 5 to 7 d, keeping the flies in free-running conditions—constant temperature and constant darkness (DD; see Notes 7-9).

3.3.2. Determining Capacity to Adapt to Different LD Cycles

In order to assay the capacity of flies to adapt to different LD cycles, experiments can be conducted in many different ways. The important variable to constantly bear in mind is that whatever changes to the lighting regime are adopted, the new condition should be monitored for at least 3 to 4 d (if not longer) in order to make sure that transients from the previous condition have disappeared, or at least become attenuated. For example, one may be interested in assaying the circadian locomotor activity response of flies to LD cycles consisting of a long day (16 h:8 h LD) or, conversely, a short day (8 h:16 h LD; Fig. 2). In each case, as flies are normally raised in incubators with 12 h:12 h

LD cycles, it is a good idea to start off the experiment by monitoring 3 to 4 d of 12 h:12 h LD cycles, followed by 5 to 7 d with the new LD condition.

3.3.3. Determining Phase-Response Characteristics

This type of experiment aims at obtaining information on the way flies respond to flashes of saturating (25-40 ^mol photons/m2/s) light given at different times of the night or the following subjective day.

1. Treat the flies with a pulse (usually lasting 10-20 min) of saturating white light (25-40 ^mol photons/m2/s). Start from the last night of the LD period and continue on the following first day of DD, at equally spaced intervals, each to be assayed in a separate experiment.

2. Monitor the activity of the flies in DD for at least 3 to 4 d.

3. Analyze the actograms of individual flies to establish whether the morning and evening locomotor activity peaks have remained in phase with the original zeitgeber time (ZT), or whether they have been influenced (phase-shifted) by the flash of light (see Note 11), and if so, by how many minutes or hours.

4. Plot the average (of all flies) of each phase delay/advance produced at each ZT and circadian time (CT) to obtain a phase-response curve. This provides important dynamic information on the state of the circadian clock at different times during the day.

3.4. Data Analysis

The many issues concerning the analysis of data collected during the monitoring of Drosophila circadian locomotor activity have received a great deal of attention from the very outset of circadian rhythms biology. In particular the major issue of concern has nearly always been the choice of the statistical approach used to extract information on the periodicities that would best describe the nature of the experimental data. An in-depth treatment of such issues can be found in Chapter 2. However, in this chapter it may be of some use to consider our personal experience in this respect. In particular, one of us (M. A. Z.) became involved in a project for the implementation of time series analysis software, possibly using platform-independent programming tools. We then chose programming tools that were not only platform-independent, but also open source (for a complete definition of the open source concept, see http:// www.opensource.org/docs/definition_plain.php). In particular, all the routines so far developed have been coded using the Python (www.python.org) scripting environment, enhanced by the Numeric library (Numerical Computing for Python library; (www.numpy.sourceforge.net) and the Tcl/Tk cross-platform graphics library (www.tcl.tk). Currently, routines for producing double-plotted actograms and for performing autocorrelation plots and CLEAN (21,22) spectral analysis have been implemented. CLEAN is further enhanced by us-

Fig. 3. The figure shows the graphical user interface of the actogram plotting program used in our laboratory. The left-hand part of the figure shows the slide controls that allow the assignment of the time of lights on and lights off for each single day, "on the fly." This leads to the coloring of the bars in the two double-plotted actograms (left-hand panel = raw data; right-hand panel = smoothed data) accordingly (i.e., white bars = lights on; dark gray bars = lights off). The smoothed actogram was obtained by applying a two-pole Butterworth filter to the raw data, in order to filter out high-frequency (i.e., <4-h periodicity) "noise."

Fig. 3. The figure shows the graphical user interface of the actogram plotting program used in our laboratory. The left-hand part of the figure shows the slide controls that allow the assignment of the time of lights on and lights off for each single day, "on the fly." This leads to the coloring of the bars in the two double-plotted actograms (left-hand panel = raw data; right-hand panel = smoothed data) accordingly (i.e., white bars = lights on; dark gray bars = lights off). The smoothed actogram was obtained by applying a two-pole Butterworth filter to the raw data, in order to filter out high-frequency (i.e., <4-h periodicity) "noise."

ing a Monte Carlo approach to generate 95 and 99% confidence limits, to be used as an objective criterion for the assessment of the significance of the peaks present in the CLEAN spectrum (see Note 12).

In addition, the actogram plotting routines generate two side-by-side plots consisting of (1) the double plot of the original (raw) locomotor activity data and (2) the double plot of the same data after smoothing (following the application of a two-pole Butterworth filter designed to filter out data with a periodicity below 4 h [24]). The latter plot usually facilitates the observation of the main peaks present in the behavioral data even when these are relatively "noisy" (Fig. 3). The actogram-plotting program provides a graphical user interface with slide controls (see left-hand part of figure) that allow the assign ment of the time of lights on and lights off for each single day, in a dynamic fashion. This leads to the coloring of the bars in the two actograms (raw and smoothed) accordingly (i.e., white bars = lights on; dark gray bars = lights off). An additional tool allows the user to select particular bins in the smoothed actogram. These choices can be saved in a text file, which will then contain information regarding the main phase reference points (i.e., time of lights on and lights off, mean daily activity, as well as the position of the bins selected graphically by the user; see Note 13).

4. Notes

1. Most published reports relate to data obtained only from male flies, the main reason being one of convenience, which stems from the complicating necessity that female flies must be virgin in order to avoid egg deposition in the glass tube in which the fly is housed during the data acquisition period. The eggs and the ensuing larvae can interfere seriously with the infrared locomotor activity detector.

2. The paper towel will absorb atmospheric humidity (which may condense in the Petri dish); this guarantees that the flies will stay asleep for all the time necessary to load them into the tubes, but without risking the irreversible damage (often leading to death) following prolonged periods (>20 min) of CO2 anesthesia.

3. This means that, within a physiologically well-tolerated range of temperatures (i.e., 18-28°C), the general circadian characteristics of the activity patterns will in fact be conserved, although the daily mean of locomotor activity may show subtle changes, and the relative heights and position of the morning and evening activity peaks may also be influenced (16).

4. The "daylight" type tube has a definitely more intense spectral emission in the 400- to 570-nm range, and has a relatively lower emission at wavelengths greater than 570 nm (20). Fluorescent tubes in general show relatively little changes in emission during their normally fairly long lifetime (several months of continuous usage); furthermore, they emit little far red/infrared radiation, which entails the generation of relatively little heat.

5. Recently it has become possible to consider employing high-efficiency LEDs, as these are available in types that can provide full visible spectrum emission (i.e., 400-600 nm) as well as fairly narrow bandwidth monochromatic light (usually available in the colors of blue, green, yellow, orange, and red). The main advantage of such illumination devices is their low energy consumption, limited generation of heat, and, owing to their small size, the ease of placement in the experimental setup (which can even envisage having an array of LEDs so that each fly is directly illuminated by its own personal "sun").

6. It is, however, important to point out that, should one wish to simply monitor the lighting conditions occurring in an incubator during an experiment in which Drosophila circadian locomotor activity is being monitored, it is sufficient to employ a luxmeter (or even a simple phototransitor) to provide information on whether the lighting source is turned on (with an intensity reading as well) or off at a given moment in time.

7. The data corresponding only to the time spent by the flies in DD will then be used for the appropriate statistical analyses (see Subheading 3.4.) in order to obtain an estimate of the desired t value. As an extra precaution, the first day of DD data should be excluded from the statistical analysis to avoid, at least in part, the effects of transients (i.e., aftereffects reflecting the fly's previous LD experience). Such transients may take (on average) a day or two to completely disappear (for example, see ref. 17).

8. Because one has gone to the trouble of organizing the experiment, it is not a bad idea to continue collecting data even after the end of the DD period, by providing the flies again with 3 to 4 d of 12 h:12 h LD cycles with the same phase as the LD cycles used at the beginning of the experiment. Inspection of actograms obtained from the data collected during the whole experiment (i.e., 3-4 d LD + 5-7 d DD + 3-4 d LD) can provide useful information on the capacity of the flies to readapt to a zeitgeber once they have been left in free running conditions for some days.

9. As stated in the introduction, the collection of locomotor activity data, among other things, entails the choice of an appropriate sampling frequency. The answer to this question is strictly related to how this choice will affect the type of inferences we wish to be able to make from our data. In particular, as circadian rhythms biologists, our major goal (normally) is the identification of patterns of events occurring with circadian or close to circadian regularity. In general, information and signals analysis theory state that the sampling frequency should be twice that of the highest frequency to be analyzed (in other terms, the interval between samples should be half that of the shortest periodicity of interest), a concept that is also known as the "Nyquist limit" (23). This implies that with a sampling frequency of 0.02 Hz (i.e., one sample every 10 min), we would be limited to the analysis of periodicities no shorter than 20 min. For anyone interested in circadian rhythms this would already be much more than actually required. In fact, most cir-cadian experimental setups use sampling rates included between 0.04 and approx 0.06 Hz (5 and 30 min, respectively). There is also an upper limit to the determination of periodicities in time-series analysis (i.e., the longest periodicity that can be determined can be no longer than half of the length of the series itself), which means that, let us suppose, if we are interested in periodicities in the circadian range (i.e., 24 h) then our data should consist of at least 48 1-h samples (23).

10. By convention, times of the day during 12 h:12 h LD cycles are referred to as ZTs, so that the time at which lights come on is referred to as ZT 0, and lights off is referred to as ZT 12. Thus, daylight hours are included between ZT 0 and ZT 12, whereas hours of darkness are included between ZT 12 and ZT 24 (or ZT 0). During free-running conditions (DD), it is customary to refer to the original ZTs as reference points during the circadian cycle, but in this case the times are referred to as CTs, with the same numeration as above (i.e., CT 0 is the time when lights would have come on if the flies were in an LD cycle; similarly CT 12 is the time when lights would have been turned off in an LD cycle).

11. Typically, in D. melanogaster, flashes of light given between ZT 12 and ZT 18 will tend to produce a phase delay in the flies (i.e., the morning peak of activity, which is normally due at around CT 0, will tend to occur later); flashes of light fil* I PAAJ IIBld- .

Fig. 4. Our data analysis software package currently contains routines that allow on-screen viewing, in a single screen shot, of: (1) upper panel: the CLEAN spectral plot along with the 95 and 99% Monte Carlo confidence limits, while also providing the periods corresponding to the first 10 most significant peaks; (2) lower panel: the autocorrelogram, along with the canonical 95% confidence limits.

Fig. 4. Our data analysis software package currently contains routines that allow on-screen viewing, in a single screen shot, of: (1) upper panel: the CLEAN spectral plot along with the 95 and 99% Monte Carlo confidence limits, while also providing the periods corresponding to the first 10 most significant peaks; (2) lower panel: the autocorrelogram, along with the canonical 95% confidence limits.

given after ZT 18 will tend to produce a phase advance in the flies (i.e., the morning peak of activity, which is normally due at around CT 0, will tend to occur earlier).

12. The algorithms employed for the CLEAN spectral analysis software were ported from original FORTRAN listings by Dr. J. Lehar (MIT, Cambridge, MA). We are also planning to implement the algorithms necessary to perform maximum entropy spectral analysis (25) in the same package. The software package currently contains routines that allow the on-screen viewing of the CLEAN spectral plot along with the 95 and 99% confidence limits (which also provides the periods corresponding to the first 10 most significant peaks) and the autocorrelogram in a single screen shot (Fig. 4).

13. These data form the basis of phase analysis of the locomotor activity data under study. In this respect, we adopted an external statistical package, which we found to be very useful in performing circular statistics as well as cross-correlation analysis (which are both related to phase analysis [25J) — the open source R package (the R Foundation for Statistical Computing; see ref. 25), integrated by the CircSats package from the Comprehensive R Archive Network (CRAN, www.cran.r-project.org).

References

1. Konopka, R. J., and Benzer, S. (1971) Clock mutants of Drosophila melanogaster. Proc. Natl. Acad. Sci. USA 68,, 2112-2116.

2. Martin, J-R., Ernst, R., and Heisenberg, M. (1999) Temporal pattern of locomotor activity in Drosophila melanogaster. J. Comp. Physiol. 184, 73-84.

3. Martin, J-R. (2003) Locomotor activity: a complex behavioural trait to unravel. Behavioural Processes 24, 145-160.

4. Shaw, P. J., Cirelli, C., Greenspan, R., and Tononi, G. (2000) Correlates of sleep and waking in Drosophila melanogaster. Science 287, 1834-1837.

5. Sehgal, A., Price, J. L., Man, B., and Young, M.W. (1994) Loss of circadian behavioural rhythms and per RNA oscillations in the Drosophila mutant timeless. Science 263, 1603-1606.

6. Yang, Z., and Sehgal, A. (2001) Role of molecular oscillations in generating behavioural rhythms in Drosophila. Neuron 29, 453-467.

7. Emery, P., So, W. V., Kaneko, M., Hall, J. C., and Rosbash, M. (1998) CRY, a Drosophila clock and light-regulated cryptochrome, is a major contributor to circadian rhythm resetting and photosensitivity. Cell 95, 669-679.

8. Hamblen, M. J., White, N. E., Emery, P. T., Kaiser, K., and Hall, J. C. (1998) Molecular and behavioral analysis of four period mutants in Drosophila melanogaster encompassing extreme short, novel long, and unorthodox arrhythmic types. Genetics 149, 165-178.

9. Allada, R., White, N. E., So, W. V., Hall, J. C., and Rosbash, M. (1998) A mutant Drosophila homolog of mammalian clock disrupts circadian rhythms and transcription of period and timeless. Cell 93, 791-804.

10. Ashburner, M. (1989) Drosophila. A Laboratory Handbook. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

11. Roberts, D. B. (1998) Drosophila: A Practical Approach (2nd Ed.). The practical approach series (Hames, B .D., series ed.). Oxford University Press, Oxford, UK.

12. Helfrich-Forster, C. (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 15, 135-154.

13. Gatti, S., Ferveur, J-F., and Martin, J-R. (2000) Genetic identification of neurons controlling a sexually dimorphic behaviour. Curr. Biol. 10, 667-670.

14. Sawyer, L. A., Hennessy, J. M., Peixoto, A. A., et al. (1997) Natural variation in a Drosophila clock gene and temperature compensation. Science 278, 2117-21120.

15. Rensing, L., Mohsenzadeh, S., Ruoff, P., and Meyer, U. (1997) Temperature compensation of the circadian period length — a special case among general homeo-static mechanisms of gene expression? Chronobiol. Int. 14, 481-498.

16. Majercak, J., Sidote, D., Hardin, P. E., and Edery, I. (1999) How a circadian clock adapts to seasonal decreases in temperature and day length. Neuron 24, 219-230.

17. Zordan, M. A., Osterwalder, N., Rosato, E., and Costa, R. (2001) Evidence for extraocular and red light-mediated photic entrainment in Drosophila melanogaster. J. Neurogenet. 15, 1-20.

18. Zordan, M. A., Rosato, E., Piccin, A., and Foster, R. (2001) Photic entrainment of the circadian clock: from Drosophila to mammals. Semin. Cell Devel. Biol. 12, 317-328.

19. Johnson, C. H., Elliott, J. A., and Foster, R. (2003) Entrainment of circadian programs. Chronobiol. Int. 20, 741-774.

20. Ryer, A. D. (1996) Light Measurement Handbook [On-line.] Available at www.intl-light.com/customer/handbook/. Last accessed: June 12, 2006.

21. Robert, D. H., Lehar, J., and Dreher, J. W. (1987) Time series analysis with CLEAN. I. Derivation of spectra. Astron. J. 93, 968-989.

22. Negi, J. G., Tiwari, R. K., and Rao, K. N. N. (1996) Clean periodicity in secular variations of dolomite abundance in deep marine sediments. Marine Geology 133, 113-121.

23. Taylor, F. J. (1994) Principles of Signals and Systems. McGraw-Hill, Singapore.

24. Levine, J. D., Funes, P., Dowse, H. B., and Hall, J. C. (2002) Signal analysis of behavioural and molecular cycles. BMC Neuroscience 3, 1.

25. R Development Core Team. (2003) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-00-3, http://www.r-project.org. Las accessed: June 12, 2006.

Was this article helpful?

0 0

Post a comment