CIRCADIAN SYSTEM PROPERTIES

Jürgen Aschoff,     Max-Planck-Institut für Verhaltensphysiologie, Andechs, FRG

Publisher Summary

This chapter discusses several basic properties of the circadian system. Circadian rhythms are characterized by their capability to free-run in constant conditions like self-sustaining oscillations and by the way in which they are synchronized—entrained—by periodic factors in the environment, the Zeitgebers. Circadian rhythms are only entrainable by Zeitgebers within a limited range of periods. Within this range of entrainment, the rhythm changes its phase-relationship to the Zeitgeber: the rhythm phase leads a Zeitgeber having a relatively long period, and phase lags a Zeitgeber with a short period. Moreover, when exposed to an appropriate Zeitgeber, for example, a light-dark cycle complemented by regular gong signals, human circadian rhythms can be entrained—within limits—to periods other than 24 h.

Én nem csalódom. Minden szervem ora, mely csillagokhoz igazitva jár.

(I am not misled. All my organs are clocks which run adjusted to the stars.)

Attila József, 1937

When Attila József wrote his poem ‘Majd emlékezni jó lesz’ (It will be good to remember) in which he refers to his inner clocks, the biological basis of such time measuring devices had yet to be discovered. As a true poet, József envisaged what was shown to be true about 25 years later: that there is a biological clock in man which measures time of day, and that it consists of a multiplicity of oscillating units, located in various organs and coacting as a complex entity called ‘the circadian system’. In this lecture, I am attempting to introduce a few basic features of the circadian system to those who have become interested in the subject only recently. To a large extent, the survey is based on data which have been collected by my co-workers E. Gwinner, K. Hoffmann, H. Pohl, U. von Saint Paul, and R. Wever over the last 15 years.

1 FREERUNNING AND ENTRAINED RHYTHMS IN ANIMALS

By evolutionary adaptation to the temporal program of day and night, eucariothic organisms have developed endogenous periodic processes whose natural frequency approximates that of the earth’s rotation and which persist in the absence of any periodic input to the organism. Since the period of the rhythm slightly deviates from 24 h under artificially constant conditions, the prefix circa has been introduced by Halberg (1959). Circadian rhythms, then, are characterized a) by their capability to freerun in constant conditions like self-sustaining oscillations, and b) by the way in which they are synchronized (entrained) by periodic factors in the environment, the Zeitgebers. The two examples provided in Fig. 1 show rhythms in oxygen uptake of two chaffinches, kept initially in light-dark cycles of 12 h light and 12 h darkness (LD 12:12) and thereafter in conditions of constant dim illumination (LL). As dayactive animals, the birds have a high level of oxygen uptake during L, and a low one during D. In LL, the rhythm persists undamped with a period, τ, which is longer than 24 h in the upper record, and shorter than 24 h in the lower record. This difference indicates that there can be a substantial interindividual variability in τ. In addition, τ is known to depend on the physiological state of the organism e.g. with regard to its reproductive functions, as well as on external factors such as intensity of illumination or ambient temperature. The effects of external factors show systematic differences between dayactive and nightactive species. (For a review, cf. Aschoff 1979a).

Fig. 1 Circadian rhythms of oxygen uptake in two chaffinches, Fringilla coelebs, kept first in a light-dark cycle (LD), thereafter in constant dim illumination (LD). τ = Mean circadian period. Shaded area: darkness. (From Pohl, published in Aschoff et al. 1980)

When entrained by a Zeitgeber, the circadian rhythm maintains a distinct phase-relationship with the entraining signals. This phase-angle difference ψ might be measured between an arbitrary phase of the rhythm, e.g. a minimal value, and the time of ‘light-on’ of a LD-cycle. If in case of the records reproduced in Fig. 1 one takes as phase reference the point where oxygen starts to increase from its low D-level, it becomes evident that ψ has a small positive value (i.e. a leading phase) in the upper record, and a large positive value in the lower record. This difference in ψ is correlated with the difference in the τ-values (24.8 h in the upper record, 23.1 h in the lower record) and reflects the general rule that ψ depends on the ratio between the τ of a rhythm (as measured in constant conditions) and the period T of the entraining Zeitgeber. In consequence of this rule, ψ also changes when a rhythm becomes entrained by Zeitgebers with periods other than 24 h (cf. Fig. 3). (For a discussion of these rules, cf. Aschoff 1965a, 1981a).

Fig. 3 Circadian activity rhythms of chaffinches, Fringilla coelebs, kept in light-dark cycles of various period length (T). τ = mean circadian period. Horizontal bars: activity time. Shaded areas: darkness. (Wever, unpubl.)

Next to a LD-cycle which is the prime Zeitgeber for most organisms, a cycle of low and high temperature can entrain circadian rhythms at least in poikilothermic animals. In lizards, a cycle with a range of only 0.9°C suffices to entrain the activity rhythm of 1/3 of the animals tested (Hoffmann 1969). Homeiothermic animals are less easily entrained by temperature cycles. In squirrel monkeys, a range of 17°C has been found to be effective in about 50% of the animals tested (Fig. 2, left diagram). If entrainment is not achieved, the rhythm continues to freerun with a mean period which is continuously modulated by the signals from the Zeitgeber. This relative coordination (Enright 1965) between the rhythm and a Zeitgeber of insufficient strength is illustrated in the right diagram of Fig. 2. The period of the non-entrained activity rhythm is close to that of the Zeitgeber when onset of activity coincides with the warm half of the temperature cycle, and it is lengthened when the onset of activity falls in the cold half of the cycle. Relative coordination also plays its role when various components of a freerunning circadian system loose their mutual coupling as in the case of internal desynchronization (cf. section 4, Fig. 9 and 10).

Fig. 2 Circadian activity rhythms in two squirrel monkeys, Saimiri sciureus, kept alternatively in constant temperature and in a temperature cycle. Original record duplicated along the abscissa. Within rectangles: time of higher temperature. (Tokura and Aschoff, unpubl.)

Fig. 9 Freerunning circadian rhythms of a singly isolated human subject, showing spontaneous internal desynchronization on day 14 of the experiment. White bars: sleep time. Triangles: minima of rectal temperature. Data plotted twice along the abscissa. (Data from Wever 1979)

Fig. 10 Interaction between the rhythm of rectal temperature and the sleep-wake cycle during internal desynchronization. Range of temperature (upper ordinate) and duration of sleep (lower ordinate) drawn as a function of the time at which sleep onset occurs in the temperature cycle. The abscissa represents a full circadian cycle of rectal temperature with a mean period of 25 h. (Data from Zulley 1980 and Zulley et al. 1981)

Circadian rhythms are only entrainable by Zeitgebers within a limited range of periods. Within this range of entrainment the rhythm changes its phase-relationship to the Zeitgeber according to the rule mentioned above: the rhythm phase leads a Zeitgeber which has a relatively long period, and phase lags a Zeitgeber with a short period. This is illustrated in Fig. 3 by the activity records from chaffinches. When kept in a LD-cycle with a period T = 24 h, the bird is dayactive but a late riser (onset of activity about 2 h after light-on). When T is changed to 25 h, the bird becomes an early riser, while in T = 23 h the normally dayactive bird becomes mainly nightactive. When T is lengthened to 26 h, or shortened to 22 h, entrainment is lost and the freerunning rhythm only shows relative coordination. It should be noted that a LD-cycle with 5 lux in L and 1 lux in D has been used in these experiments. Such a small range in intensity of illumination provides only a weak Zeitgeber; hence the range of entrainment is very small, and the changes in ψ with changing T are large (cf. Aschoff and Pohl 1978).

2 A SYSTEM OF COUPLED OSCILLATORS

So far, circadian rhythms have been treated as being controlled by one basic oscillator, a single circadian clock. A growing body of data indicates that such a one-oscillator model does not account for all the facts. Some...