General Models For Circadian Timekeeping In Photoperiodism

Although admittedly based on results from rather few species, evidence of the kind outlined has shown that a circadian oscillator is involved in the photoperiodic processes of plants. Erwin Bunning (1936) was the first to suggest that the measurement of time in photoperiodism is dependent on an endogenous circadian oscillation. His original proposal was that photoperiodic timekeeping involves a regular oscillation of phases (i.e. portions of the rhythm) with different sensitivities to light. He suggested that transfer to light set in motion a 12 h photophile (light-requiring) phase, which is followed by a 12 h skotophile (dark-requiring) phase. Light was required during the photophile phase but flowering was inhibited by light given during the skotophile phase, thus giving rise to the inhibitory night-break effect in SDP and the need for both a critical duration of darkness and for light during the photoperiod, as had previously been demonstrated (see Chapter 1). In order to account for the behaviour of LDP, it was necessary to assume that the beginning of their photophile phase was delayed by about 12 h after transfer to light; light would then fall in the photophile phase only in long photoperiods. This would suggest that the photophile and skotophile phases in LDP are displaced by about 12 h compared with SDP, which seems to accord with some of the experimental data (Figs 2.6 and 2.8). The hypothesis also received support from results with the LDP dill, which was shown to flower better when a period of darkness was given during the postulated skotophile phase. This rather general model has become known as Biinning's hypothesis. It has, however, since been substantially modified from its original form, by both Bunning and others, and more explicit models based on known properties of the circadian system have been developed.

One problem with the original hypothesis was that many LDP did not appear to show a promotion of flowering by a period of darkness during the postulated skoto-phile phase (i.e. the first 12 h of the photoperiod in LDP). Consequently, the scheme for LDP was modified to propose that the difference between LDP and SDP is that light during the skotophile phase promotes flowering in the former, but inhibits flowering in SDP. Thus coincidence or non-coincidence of light with the skotophile phase was considered to be the determining factor in the induction of flowering in photoperiodically sensitive plants, with light having opposite effects in the two response groups (Bunning, 1960). Much of the later work followed the lead set by Bunning in supposing that there is a light-sensitive phase in the photoperiodic rhythm (Biinning's skotophile phase) and that the response is determined by coincidence or non-coincidence of light with this phase (Andrade and La Motte, 1984; Bollig et al., 1976; Fukshansky, 1981; Thomas and Vince-Prue, 1987; Vince-Prue and Lumsden, 1987). This type of scheme is known as external coincidence. It assumes that there is a single photoperiodic rhythm and that light has direct effect to prevent the induction of flowering in SDP (or induce it in LDP) when it is coincident with a particular lightsensitive phase of this rhythm. Biinning's term skotophile relates to a full half-cycle of the rhythm (180°). However, it has been observed that photoperiodic induction is contingent on the illumination of a much smaller fraction of the cycle and that light does not affect induction during many parts of the presumed skotophile phase. Consequently, Pittendrigh (1966) proposed the introduction of the term inducible phase (cpj), as it is more restrictive in time and explicitly distinguishes between the inducing and entraining actions of light in the photoperiodic rhythm. He postulated that 'photoperiodic induction (in the SDP Lemna paucicostata) is contingent on the (non-)coincidence of light and a specific inducible phase, (pi? in the oscillation'. He also emphasised that, in such a model, light has two different kinds of effect, firstly as an entraining or rephasing agent and secondly as a photoperiodic inducer/inhibitor. As discussed for overt circadian rhythms, the entire subjective night (or skotophile half-cycle) is sensitive to light as it affects the phase control of the oscillation but the photoperiodic induction process is sensitive to light only at a particular time.

The crucial factor of any external coincidence mechanism for photoperiodic timekeeping is that induction, or non-induction, depends on the coincidence of an external signal (light) with an internal light-sensitive phase of a circadian rhythm. Much of the evidence obtained with SDP, especially those which can be induced with a single short-day/long-night cycle, is consistent with an external coincidence model of the type originally proposed by Pittendrigh. In this model, it is argued that the oscillation assumes a definite phase relationship with the light cycle so that the light-sensitive inducible phase («pO is, or is not illuminated as the light/dark regime changes. Thus, according to this theory, the annual variations in daylength illuminate (ft at some seasons but not others. The way in which the oscillation may be controlled by the light signals to achieve this result has been the subject of many experiments and is discussed in the next section with reference to SDP. A detailed discussion of timekeeping in LDP is deferred to Chapter 5.

An alternative type of scheme, called internal coincidence, ascribes photoperiodic responses to the interaction of two rhythms, with induction occurring only when critical phase points in the two rhythms coincide. An example would be a rhythm of enzyme activity together with a rhythm of substrate availability. In such a mechanism, light would not interact directly with a circadian rhythm. The inhibition of flowering in unfavourable cycles would then be due to the rephasing of one oscillation so that it is no longer in phase with another. In this way the critical phase points would only coincide under particular photoperiods to give rise to short- or long-day responses. This approach has not yet been explored in any detail in plants.

CIRCADIAN TIMEKEEPING IN SHORT-DAY PLANTS

Any model for circadian timekeeping in SDP must be capable of explaining how an oscillation with a 24 h periodicity can measure the critical duration of darkness which determines their photoperiodic response. A number of explicit models have been developed, based mainly on results with single-cycle SDP such as Pharbitis, Xanthium and Chenopodium. Pharbitis is a particularly useful subject for studying the behaviour of the photoperiodic rhythm because dark-grown seedlings will respond to a single light/dark cycle. This means that the relationship between photoperiod duration and dark-timekeeping can be studied without the complicating problem of entrainment to non-inductive light/dark cycles during growth of the plant to experimental size. However, this problem can be partly overcome in other plants (e.g. Xanthium-, Papenfuss and Salisbury, 1967) by exposing them to a 'neutral', non-inductive dark period before giving the single, experimental cycle.

Timing the Night-Break

One approach to understanding how the circadian rhythm operates to control photoperiodic timekeeping has been to examine the time of maximum sensitivity to a night-break (NBmax) following a single photoperiod of varying duration. Dark-grown seedlings of Pharbitis will flower in response to a 5 min pulse of R followed by an inductive dark period before transfer to LL, provided that the seedlings are sprayed with benzyladenine at the time of giving the brief photoperiod (Ogawa and King, 1979a). Under these conditions, a circadian rhythm of responsiveness to a night-break occurred with the first NBmax at approximately 15 h after the initial exposure to light (King et al„ 1982; Lumsden et al, 1982; Lee et al, 1987). The time of the first NBmax varied between 10 and 18 h (in different experiments and seed lots) but the second NBmax always occurred 24 h after the first. Rhythm phasing was established by the initial exposure to R, irrespective of the age of the dark-grown seedlings at that time. When the duration of the photoperiod was varied, it was found that the first NBmax occurred at a constant time (approximately 15 h) from the beginning of the photoperiod when this was less than about 6 h long. When the photoperiod was longer than 6 h, the time of sensitivity to light was delayed and NBmax always occurred at a constant time after the end of the photoperiod (Fig. 2.11). With longer photoperiods the rhythm was gradually damped and, after 24 h, there was essentially only a single NBmax response at 25°C (Fig. 2.7).

These results with Pharbitis have been interpreted in the following way (Fig. 2.12).

FIG. 2.11. The effect of photoperiod duration on the time of night-break sensitivity in dark-grown seedlings of Pharbitis. Dark-grown seedlings received a photoperiod of various durations as indicated on the figure, followed by an inductive dark period interrupted at various times by a 10 minute night-break. Plants were returned to continuous white light 72 h after the beginning of the photoperiod. The time of the maximum sensitivity to a night-break (NBmax) was constant (15 h) from light-on with photoperiods of < 6 h duration; with photoperiods of > 6 h, NBmax was constant (9 h) from light-off. After Lumsden et ah, 1982.

NBmax i h

FIG. 2.11. The effect of photoperiod duration on the time of night-break sensitivity in dark-grown seedlings of Pharbitis. Dark-grown seedlings received a photoperiod of various durations as indicated on the figure, followed by an inductive dark period interrupted at various times by a 10 minute night-break. Plants were returned to continuous white light 72 h after the beginning of the photoperiod. The time of the maximum sensitivity to a night-break (NBmax) was constant (15 h) from light-on with photoperiods of < 6 h duration; with photoperiods of > 6 h, NBmax was constant (9 h) from light-off. After Lumsden et ah, 1982.

Q Light initiates and suspends rhythm: this restarts on transfer to dark Light inhibits flowering at <P i (CT15)

FIG. 2.12. Scheme showing the actions of light in the photoperiodic control of flowering in Pharbitis nil. tPi represents a specific light-sensitive phase of the circadian photoperiodic rhythm when light acts directly to inhibit flowering. The action of light to initiate and suspend the rhythm is considered to be on the underlying circadian oscillator to which the rhythm in light sensitivity is coupled.

A single photoperiodic rhythm of sensitivity to light is initiated by transfer to light at dawn (i.e. CT = 0 at light-on) and the light-sensitive phase (<Pi) of this rhythm occurs about 15 h after the light-on signal (i.e. at CT 15). The rhythm initially continues to run in continuous light so that, in real time, NBmax always occurs 15 h after light-on. A light-on signal is thus sufficient to induce flowering under these conditions unless a second pulse of light is given at (ft. After about 6 h in continuous light (i.e. at CT 6), the rhythm appears to become 'suspended' and remains at CT 6 for as long as the plant stays in continuous light. It is then released by a light/dark transition. Since the rhythm is suspended at CT 6, the time of NBmax (in real time) is always about 9 h after transfer to darkness (i.e. at CT 15), as would be predicted from the original light-on rhythm. At temperatures sufficiently high to maintain growth, the daylength under natural conditions would always be longer than 6 h so that the flowering rhythm would go into suspension during the photoperiod and be released by the light/dark transition at the end of the day; this would result in NBmax always occurring at a constant time from dusk. Induction would then depend on whether or not (ft is reached before the dawn signal is experienced.

This model for Pharbitis is based on results obtained with dark-grown seedlings where the first exposure to light initiates the circadian rhythm of light sensitivity. Under natural conditions, plants are repeatedly exposed to photoperiods of several hours duration and, according to this model, would have a dominant rhythm phased from the dusk signal. One can question, therefore, whether the mechanism operates under these conditions and also whether it operates in other SDP. In Pharbitis, although the results were slightly different from those obtained in dark-grown plants, there is evidence that the model is applicable to light-grown seedlings. Seedlings given a preliminary photoperiod of 24 h, followed by 8 h dark, showed a phase shift (advance) of the light-off rhythm following a second photoperiod from 10 min to 2 h in duration, while light for more than between 2 and 6 h resulted in the first NBmax at a constant time from the new light-off (Fig. 2.13). Earlier experiments with seedlings grown for 3-4 weeks in LL or long days produced similar results: following a 6 h exposure to R, NBmax always occurred at 9-9.5 h from light-off, whereas a 2 h exposure resulted in a phase shift (Bollig, 1977).

In Pharbitis, there is clear evidence of a rhythmic sensitivity to night-break light, at least when the photoperiods are short (Fig. 2.7). In Xanthium, in contrast, attempts to demonstrate the existence of a rhythm by giving light perturbations at different times during a long dark period have been wholly unsuccessful. Nevertheless, it seems likely that photoperiodic timekeeping in Xanthium operates in a way wholly comparable to that proposed for Pharbitis. Following a neutral, non-inductive dark period of 7.5 h, NBmax occurred at a constant time (14.0 h) from the beginning of the photoperiod when this was less than 5 h long. After longer photoperiods, NBmax occurred at a constant time (about 8.5 h) after the end of the photoperiod (Fig. 2.14). These results are consistent with the concept that a rhythm is initiated or rephased (at CT 0) by the light-on transition at dawn with a light-sensitive phase, (pi5 occurring 14.0 h later (at approximately CT 14). After 5 h in the light, this rhythm goes into suspension and the time of NBmax occurs about 8.5 h after transfer to darkness in real time (i.e. close to CT 14) as predicted from the original light-on rhythm).

The results with Xanthium (Fig. 2.14) were obtained when a photoperiod of varying duration was given after a dark period of 7.5 h. In other experiments, it was found that

NBmax/h c o

Prediction from dark-grown plants

FIG. 2.13. Effects of light to phase shift the photoperiodic rhythm in light-grown plants of Pharbitis. Seedlings were grown in darkness and exposed to a 24 hour photoperiod before transfer to a 72 hour dark period. A second photoperiod in red light of various durations was given beginning at the 6th hour of darkness. The time of maximum sensitivity to a subsequent 10 minute night-break (NBmax) is shown. Photoperiods of 10 minutes and 40 minutes caused a phase shift in the position of NBmax, and a photoperiod of 6 hours or longer was sufficient to give a new light-off signal with NBmax occurring 8 h later. After Lumsden and Furuya, 1986.

S2 14

16

"O

o

14

d>

Q.

g

12

O ■C

Q.

<0

£

10

o

■o

8

c a>

a>

£

6

E

o

4

(0 i—

O

JZ

2

m

Duration of photoperiod / h

FIG. 2.14. Time of maximum sensitivity to a night-break in Xanthium as affected by the duration of the preceding photoperiod. Plants were given a 'phasing' dark period of 7.5 hours followed by a photoperiod of various durations. After this an inductive dark period of 14 hours was interrupted by a short night-break at various times. The figure shows the times at which this night-break was most effective (NBmax) in terms of the number of hours from (A) the beginning or (B) the end of the photoperiod. The time of a short exposure to light did not act as a dawn signal until more than 6 h of darkness had elapsed. Thus, in Xanthium, it appears that dawn (light-on) will rephase to CT 0 only after at least 6 h of darkness. Similarly, in Pharbitis, the effect of a light pulse on the rhythm phase of light-grown seedlings varied with the time from light-off (Lumsden and Furuya, 1986).

A crucial feature of an external coincidence model is that light has two actions: one action is on the clock to set the phase of the photoperiodic rhythm and the other is a direct action to inhibit (in SDP) flowering at a specific inducible phase of the rhythm. However, in interpreting the results with Xanthium, it was initially proposed that light may only have one action, namely on the phasing of the clock (Papenfuss and Salisbury, 1967). In this model a night-break always acts to rephase the photoperiodic rhythm and the inhibition of flowering results only when it is rephased in such a way that the particular phase of the rhythm which is essential for floral induction is never reached. In the absence of experiments designed specifically to address the question, it is not possible to differentiate between the two concepts since, in both cases, dawn would rephase the rhythm (to CT 0) and flowering would occur only when an inducible phase for flowering is reached before the next dawn occurs. However, in Pharbitis, there is good evidence for two actions of light, since they can be discriminated on the basis of their dose-response characteristics (Lumsden and Furuya, 1986). For example, at the 8th hour of an inductive dark period following transfer from 24 h LL (when light is strongly inhibitory to flowering), a marked reduction of flowering was obtained with an exposure which had no effect on the phase of the night-break rhythm (Fig. 2.15). At other times, for example at the 6th hour, the rhythm could be phase shifted with no effect on flowering. Using a different protocol with a brief photoperiod, a considerably greater exposure of R (312 jimol m~2) was needed to phase shift the rhythm at the 16th hour of darkness, than the exposure (25-30 (irnol m~2) shown by other workers to inhibit flowering at that time (Lee et al., 1987). These results argue strongly that the inhibition of flowering by light is not a consequence of a shift in the phase of the rhythm and afford strong support for an external coincidence mechanism for photoperiodic control in Pharbitis. It is unfortunate that there is not more direct experimental evidence relating to this question, which has important implications for understanding the actions of light to control flowering in SDP (see Chapter 4) as well as for understanding how photoperiodic timekeeping operates.

The apparent suspension of the rhythm at a particular phase point in continuous light is not peculiar to photoperiodism. In many overt rhythms, rhythmicity is abolished after about 12 h in LL and is resumed at CT 12 when the organism is returned to darkness (Lumsden, 1991). The most straightforward explanation is that the circadian oscillator is arrested at a characteristic phase in LL and is restarted (or reset) following transfer to darkness. However, when transferred to darkness after photoperiods which were longer or shorter than 12 h, the eclosion rhythm in the flesh fly Sarcophaga agyrostoma showed residual circadian fluctuations, indicating that the oscillation in LL continued but with only a slight variation around the apparently 'suspended' phase point (Peterson and Saunders, 1980). Based on these results, it was suggested that the apparent suspension in continuous light may result from a change in the dynamics of the pacemaker such that it oscillates within a much reduced cycle (a light-limit cycle), effectively occupying a small time domain. Following transfer to darkness, the circadian oscillator moves to a dark-limit cycle taking essentially the

0 6 20 60 200 600

0 6 20 60 200 600

0 6 20 60 200 600

Duration of R exposure / s

FIG. 2.15. Dose-response curves for the action of light to inhibit flowering and to phase shift the night-break rhythm in Pharbitis. Dark-grown seedlings were given a 24 hour photoperiod followed by a 72 hour dark period. Various durations of light as shown on the abscissa were given 8 hours after the beginning of the dark period, (a) shows the direct inhibition of flowering and (b) the magnitude of the phase shift in NBmax obtained by scanning the subsequent dark period with a second light pulse. After 8 hours in darkness a strong inhibition of flowering occurred with light exposures of 200 s, which had no effect on the phase of the rhythm. After Lumsden and Furuya, 1986.

same time to reach it regardless of the starting point in the light-limit cycle. The rhythm would then appear to resume from a more or less constant phase point irrespective of the duration of the preceding photoperiod.

The fact that, in Xanthium, the time to NBmax is not precisely constant but varies between 7 and 8.5 h (Fig. 2.14) can be interpreted in terms of the light-limit cycle concept, with a small oscillation such that the time taken to reach the dark-limit cycle varies slightly according to the position in the light-limit cycle when the plant is transferred to darkness. However, the results have been interpreted differently by Salisbury (Fig. 2.16), who proposed that the clock begins to oscillate into the 'night-mode' after about 9 h in continuous light, going as far as it can after about 12 h. If plants are placed in darkness at this time, the clock would continue to oscillate and NBmax would occur earlier (after about 7 h). However, if the plants are left in the light, the clock is forced back into the suspended 'day mode', with dusk required to restart the oscillation and NBmax occurring 8.5 h later. It is not clear, however, why more than 12 h of light should be required to suspend the clock; it appears already to be suspended after 5 h in the light since NBmax is a constant time (8.5 h) from light-off after a 5 h photoperiod (Fig. 2.14). Overall, the results with Xanthium seem to be more consistent with the model presented for Pharbitis, with a light-on oscillation which requires about 5 h of light to move into a light-limit cycle; the time taken to move into a dark-limit cycle will vary slightly according to the position in the light-limit cycle that has been reached when plants are transferred to darkness, leading to a small residual oscillation in the time to NBmax, as observed in Fig. 2.14. In Pharbitis, the time to NBmax was essentially constant after a 6 h photoperiod with no clear evidence

Dawn

Dawn

(b) Suspending and restarting

FIG. 2.16. A model for photoperiodic timekeeping in Xanthium. It is asssumed that the rhythm is started by transfer to light and begins to oscillate into the 'dark' phase after about 9 hours. If plants are transferred to darkness after 12 hours in the light (a), the clock continues to oscillate. Consequently the phase point which allows flowering (<p0 is reached after about 7 h (based on the data of Fig. 2.17 for the time of NBmax). After 12-14 hours in the light, the clock is forced into a state of suspension (b); the oscillation is then restarted by transfer to darkness and <p; is reached 8.5 hours later. Based on Salisbury, 1990, and data of Fig. 2.14.

(b) Suspending and restarting

FIG. 2.16. A model for photoperiodic timekeeping in Xanthium. It is asssumed that the rhythm is started by transfer to light and begins to oscillate into the 'dark' phase after about 9 hours. If plants are transferred to darkness after 12 hours in the light (a), the clock continues to oscillate. Consequently the phase point which allows flowering (<p0 is reached after about 7 h (based on the data of Fig. 2.17 for the time of NBmax). After 12-14 hours in the light, the clock is forced into a state of suspension (b); the oscillation is then restarted by transfer to darkness and <p; is reached 8.5 hours later. Based on Salisbury, 1990, and data of Fig. 2.14.

for an oscillation during the photoperiod. However, a small oscillation in flower number was revealed when seedlings were transferred from continuous light at different times after sowing and tested with a dark period of constant duration, which was only slightly longer than the CNL (Spector and Paraska, 1973): it was thought that the initial zeitgeber in this case was the time of seedling emergence into LL. If the NBmax and CNL are timed in the same way (see below), these results could also be interpreted in terms of a light-limit cycle such that the CNL is influenced by the position in the light-limit cycle at the time of transfer to darkness.

Was this article helpful?

0 0

Post a comment