- Email: [email protected]
Light and temperature cooperate to regulate the circadian locomotor rhythm of wild type and period mutants of Drosophila melanogaster
Journal of Insect Physiology 44 (1998) 587–596
Light and temperature cooperate to regulate the circadian locomotor rhythm of wild type and period mutants of Drosophila melanogaster Kenji Tomioka a
a,*
, Makoto Sakamoto a, Yuka Harui a, Nobutaka Matsumoto a, Akira Matsumoto a,b
Department of Physics, Biology and Informatics, Faculty of Science, Yamaguchi University, Yamaguchi 753, Japan b Department of Biology, Faculty of Science, Kyushu University, Ropponmatsu, Fukuoka 810, Japan Received 24 September 1997; received in revised form 22 December 1997
Abstract In the wild type (Canton-S) and period mutant flies of Drosophila melanogaster, we examined the effects of light and temperature on the circadian locomotor rhythm. Under light dark cycles, the wild type and perS flies were diurnal at 25°C. However, at 30°C, the daytime activity commonly decreased to form a rather nocturnal pattern, and ultradian rhythms of a 2 苲 4 h period were observed more frequently than at 25°C. The change in activity pattern was more clearly observed in per0 flies, suggesting that these temperature dependent changes in activity pattern are mainly attributable to the system other than the circadian clock. In a 12 h 30°C:12 h 25°C temperature cycle (HTLT12:12), per0 flies were active during the thermophase in constant darkness (DD) but during the cryophase in constant light (LL). The results of experiments with per0;eya flies suggest that the compound eye is the main source of the photic information for this reversal. Wild type and per0 flies were synchronized to HTLT12:12 both under LL and DD, while perS and perL flies were synchronized only in LL. This suggests that the circadian clock is entrainable to the temperature cycle, but the entrainability is reduced in the perS and perL flies to this particular thermoperiod length, and that temperature cycle forces the clock to move in LL, where the rhythm is believed to be stopped at constant temperature. 1998 Elsevier Science Ltd. All rights reserved. Keywords: Circadian rhythm; Drosophila melanogaster; period mutants; Light-dark cycle; Temperature cycle
1. Introduction It is well known that ambient temperature profoundly affects the circadian rhythm in many insects. In general, ambient temperature has three distinct effects. First, it affects the period of freerunning rhythms. This effect has been investigated in a wide variety of animals and plants and it was found that temperature changes the freerunning period slightly but significantly in many organisms including insects (Aschoff, 1979). The Q10 is generally very close to 1.0. Second, the phase of the rhythm is often profoundly affected by an ambient temperature level. For example, the cricket Gryllus bimaculatus is nocturnally active at 25°C but becomes diurnal when * Corresponding author. Tel and Fax: + 81-839-33-5714; E-mail: [email protected]
0022–1910 /98 /$19.00 1998 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 2 - 1 9 1 0 ( 9 8 ) 0 0 0 4 6 - 8
exposed to 20°C (Ikeda and Tomioka, 1993). In the third effect, temperature cycle is well known to entrain the circadian rhythm in some animals such as lizards (Hoffman, 1969) and insects (Zimmerman et al., 1968; Lankinen and Riihimaa, 1997; also see Saunders, 1982). However, the mechanism through which temperature affects the rhythm still remains unclear. In Drosophila pseudoobscura temperature plays an important role in setting the phase of the eclosion rhythm. Based on the results of a series of elegant kinetical experiments, Pittendrigh et al. (1958) proposed the famous two oscillator model including a light-sensitive master oscillator and a temperature sensitive slave oscillator. They explained that the eclosion behavior is under direct control of the temperature sensitive oscillator that is in turn entrained by the light sensitive master oscillator. The applicability of the two oscillator model has not been thoroughly investigated in Drosophila melanogaster.
588
K. Tomioka et al. / Journal of Insect Physiology 44 (1998) 587–596
There are three types of period (per, a clock gene) mutant flies of D. melanogaster, i.e. perS, perL and per0, which behave as alleles of a single locus of the X chromosome. perS shortens the period from 24 h to 19 h, perL lengthens it to 29 h, and per0 is arrhythmic (Konopka and Benzer, 1971). It has been reported that perS and perL mutant flies have properties reciprocal to one another with respect to temperature and light intensity dependency of their freerunning periods (Konopka et al., 1989) and that wild type and per0 flies are entrainable to temperature cycles in DD (Wheeler et al., 1993). In the present study, we examined the effects of change in ambient temperature level on the locomotor rhythm and entrainability of the rhythm to temperature cycles under various lighting schedules in wild type and per mutant flies. The results show that the light and temperature have considerable effects on the activity pattern and that the light sensitive clock is also entrainable to temperature cycles, but the entrainability is reduced in perS and perL mutant flies. Since the arrhythmic, per0 flies were most sensitive to both light and temperature, it is suggested that the light and temperature sensitive driven system is normally overcome by the pacemaker in rhythmic flies.
2. Materials and methods 2.1. Rearing of flies The fruit fly, Drosophila melanogaster, was grown on glucose-cornmeal-yeast medium at 25°C under a light cycle with 12 h light to 12 h dark. The strains that we used for the behavioral assay were Canton-S, perS, perL1 and per01. We conventionally call perL1 and per01 as perL and per0, respectively, throughout. We also used per0;so which lacks the compound eyes and ocelli and per0;eya which lacks the compound eyes. 2.2. Recording of locomotor activity rhythms Male flies of five days old were individually housed in transparent acrylic rectangle tubes (3 × 3 × 70 mm). The tube was plugged at one end with agar/glucose medium as food and was sealed with paraffin, and at the other end with a silicon tube filled with damped absorbent cotton as a water source. A moving fly interrupted an infra-red beam and the number of interruptions during each 6 min was recorded using a computerized system (Tomioka et al., 1997). 2.3. Light and temperature conditions Light cycles were given by a cool white fluorescent lamp in the environment controlled room connected to an electric timer. Light intensity at the animals level was
100 to 300 lux and varied with the proximity to the lamp. Temperature cycles were set by a built-in thermostat driven by an electronic timer. Temperature steps up and down were finished within 15 min. 2.4. Estimation of circadian parameters Freerunning periods of the locomotor rhythm were estimated by the chi-square periodogram (Sokolove and Bushell, 1978). Existence of ultradian rhythms was also analyzed by the chi-square periodogram and if peaks of the periodogram appeared above the 0.05 confidence level, the animal was designated as rhythmic. Drosophila melanogaster shows bimodal activity rhythm peaking around lights-on (morning peak) and lights-off (evening peak). The time of peaks was determined from the daily activity average curves based on more than 5 days.
3. Results 3.1. Locomotor rhythms under a light cycle at different ambient temperature levels All the wild type and mutant flies were clearly synchronized to LD12:12 at two different ambient temperature levels of 25°C and 30°C (Fig. 1). perL flies consistently showed nocturnal activity rhythm in both temperature levels. Although night time activity tended to similarly decrease both when transferred from 30°C to 25°C and from 25°C to 30°C (Fig. 2C), the daytime activity often appeared clearly when transferred from 30°C to 25°C, forming a rather diurnal component (Fig. 1C). In 25°C, wild type and perS flies showed clear bimodal rhythms peaking around lights-on and before lightsoff with the activity being concentrated in the light phase, while at 30°C the daytime activity tended to decrease and on the contrary the night time activity remarkably increased, making activity rhythm rather nocturnal (Fig. 1A and B). per0 flies showed the most drastic change in activity pattern (Fig. 1D): They were diurnal in 25°C, while they turned rather nocturnal at 30°C. However, the highest peak occurred at lights-off independent of the temperature level. Statistical analysis revealed that a decrease in daytime activity at a higher temperature was commonly observed in all the strains (Figs 1 and 2). The night time activity was either significantly greater in higher temperature (perS) or remained at a similar level (per0), resulting in the night time activity being greater than daytime activity with the elevation of temperature level (Fig. 2). The phase of morning and evening peaks also slightly changed in association with the temperature level (Fig. 3). The morning peak commonly significantly advanced by a few hours with the elevation of the temperature.
K. Tomioka et al. / Journal of Insect Physiology 44 (1998) 587–596
589
Fig. 4 shows representative records of locomotor activity of the four strains of Drosophila under DD. per0 flies were arrhythmic both under 25°C and 30°C: activity equally dispersed over 24 h but ultradian rhythmicity with periods of about 3 h was evident particularly in 30°C. Flies of all other strains showed clear freerunning rhythms under DD. The change in temperature level did not significantly affect the phase of the rhythm but slightly the freerunning period (Table 2). The rhythms of wild type files showed a very slight change of freerunning period in association with temperature change (P > 0.27, t-test). However, in the two mutant flies, perS and perL, the freerunning period changed significantly with temperature change. The two mutant alleles responded to a decrease in temperature in reciprocal ways: perS lengthened its period, whereas perL shortened (Fig. 4, Table 2). Under constant light, all strains were arrhythmic at 25°C, and the arrhythmic pattern persisted in the majority of flies even when temperature was stepped up by 5°C (Fig. 5). 3.3. Locomotor rhythms under temperature cycles in constant lighting conditions
Fig. 1. Locomotor activity of wild type (A) and per mutant flies (B, perS; C, perL; D, per0) under a 12 h light to 12 h dark cycle at two constant temperature levels (30°C and 25°C). Both temperatures step up to 30°C and step down to 25°C were made at 18:00 on day 7. All flies were entrained to the LD but with the peak changing with the change of temperature level. With higher temperatures the daytime activity decreased and the night time activity increased in wild type (A) and perS strains (B). perL flies (C) were rather consistently nocturnal. per0 flies (D) switched from diurnally-active to nocturnally-active. Ultradian rhythms became prominent on transfer from 25°C to 30°C (A, B and D). Black and white bars indicate the dark and light portions, respectively.
However, the evening peak either significantly delayed (Canton-S and perS), advanced (perL) or unchanged (per0). It is notable that the chi-square periodogram analysis revealed ultradian rhythms with periods of 2-4 h in some flies. In perS and per0 flies, the number of flies showing the ultradian rhythm were significantly higher in 30°C than in 25°C (Table 1, P ⬍ 0.05, 2-test). 3.2. Rhythms under constant darkness and constant light at different temperature levels We then examined the effects of temperature on the locomotor rhythms under constant lighting conditions.
To examine whether the flies can be entrained to temperature cycles, we recorded the locomotor rhythm under a temperature cycle of 12 h high temperature (30°C) and 12 h low temperature (25°C)(HTLT12:12). Under LL conditions, flies of the four strains were synchronized to the temperature cycle. In wild-type and per0 flies the activity pattern drastically changed with the change in temperature level. Locomotor activity of wild type flies peaked twice per cycle around temperature step-up and before step-down. These two peaks apparently correspond to the morning and evening peaks under LD. The morning peak was very faint but the evening peak continued into early cryophase (Fig. 6A). The activity of per0 was concentrated in cryophase (Fig. 6D). perS and perL flies were also synchronized to the HTLT cycle under LL (Fig. 6B, Fig. 6C). The perS flies showed a trimodal activity with peaks at temperature step-up, down, and middle of the thermophase, while perL flies peaked twice, at the beginning of both thermophase and cryophase. In constant darkness, the locomotor rhythms of wild type and per0 flies clearly synchronized with the HTLT12:12 (Fig. 6). The pattern of synchronized locomotor rhythms is quite similar to that reported by Wheeler et al. (1993). In the wild type flies, the morning peak occurred several hours prior to the temperature step-up. When the temperature cycle was phase advanced or delayed by 6h by shortening or lengthening the thermophase, the rhythm was reentrained to the shifted cycle with several transient cycles (Fig. 7). The transient cycles were 5.5 ± 1.0 days and 5.1 ± 1.2 days
590
K. Tomioka et al. / Journal of Insect Physiology 44 (1998) 587–596
Fig. 2. Activity levels of day (open square) and night (closed square) in LD12:12 at different temperature levels in wild type (A) and per mutant flies (B, perS; C, perL; D, per0). Each point represents the mean of 10 to 49 animals. Vertical bars indicate SEM. Arrows indicate the direction of animal transfer from 25°C to 30°C or from 30°C to 25°C. Examples of rhythms are shown in Fig. 1.
for phase delay and advance, respectively. per0 flies showed a very clear rhythm with activity concentrated in the thermophase (Fig. 6D). A major peak occurred just after the temperature rose. When the temperature cycle was shifted, the activity phase almost immediately
shifted to synchronize with the new thermophase (Fig. 7). In perL and perS flies, however, the rhythms did not synchronize to the temperature cycle in DD but freeran with masking effects of temperature: The activity level
K. Tomioka et al. / Journal of Insect Physiology 44 (1998) 587–596
591
Fig. 3. Phases of morning peak (open circles) and evening peak (closed circles) of wild type (A) and per mutant flies (B, perS; C, perL; D, per0) in LD12:12 at different temperature levels. Each point represents the mean of 42 to 81 animals. Horizontal bars show SD. Asterisks show the value significantly different from that in 25°C (**P ⬍ 0.01, Student’s t-test). Examples of rhythms are shown in Fig. 1.
Strain
No. of animals with ultradian rhythms 30°C 25°C
per0 flies, the occurrence of the ultradian rhythm was significantly higher in DD where the activity was concentrated in the thermophase (2-test, P ⬍ 0.05 and P ⬍ 0.01, respectively). In contrast, it was higher in LL in perL (2-test, P ⬍ 0.01).
Canton-S perS perL perO
11 10* 6 20*
3.4. Locomotor rhythms under temperature cycles in constant lighting conditions in per0 flies lacking external photoreceptors
Table 1 Occurrence of ultradian rhythms at different temperature levels at LD12:12
(81) (77) (55) (69)
3 2 8 8
(63) (72) (42) (63)
Numbers in parenthesis indicate number of animals used. *P ⬍ 0.05 (2-test).
of perS and perL was higher when the activity occurred in the thermophase and cryophase, respectively (Fig. 6B, Fig. 6C). The freerunning component of perL and perS flies corresponded to the activity component occurring at temperature step-down and at the middle of the thermophase, respectively, in the previous LL. The chi-square periodogram analysis revealed ultradian rhythms in some flies (Table 3). In wild-type and
It is well known that Drosophila has an extraretinal photoreceptor for photic entrainment of the circadian rhythm (Helfrich and Engelmann, 1983; Wheeler et al., 1993). To examine the photoreceptor responsible for the phase regulation in per0 flies, we recorded locomotor activity in eya and so flies that lack the compound eyes and both compound eyes and ocelli, respectively, and their double mutant with per0, i.e., per0;eya and per0;so, under temperature cycles with different lighting conditions. In both constant light and constant darkness, locomotor rhythms of all these single and double mutant
592
K. Tomioka et al. / Journal of Insect Physiology 44 (1998) 587–596
Fig. 4. Locomotor activity of wild type (A) and per mutant flies (B, perS; C, perL; D, per0) under constant darkness at two constant temperature levels (30°C and 25°C). With lower temperature the freerunning period of the rhythm was lengthened in wild type and perS strains, but shortened in perL. Although per0 fly was consistently arrhythmic, ultradian rhythm was prominent in 30°C. Table 2 Effects of ambient temperature on freerunning period in DD Strain
Freerunning period (mean ± SD h) 25°C 30°C
Canton-S perS perL perO
24.28 ± 0.45 (5) 19.17 ± 0.41 (9) 31.2 ± 0.67 (7) arrythmic
24.05 ± 0.80 (11) 18.83 ± 0.30* (14) 32.22 ± 0.55** (6) arrhythmic
Numbers in parenthesis indicate number of animals used. *P ⬍ 0.05; **P ⬍ 0.01(t-test) compared with 25°C.
flies well synchronized to the temperature cycle, peaking twice, at the beginning of the warm phase and at the beginning of the cold phase (Fig. 8). The activity pattern of eya and so flies showed little change associated with lighting conditions. Similarly, the pattern was little affected by the lighting conditions in both double mutant strains: only at most a two-fold increase in thermophase activity was observed in DD (Fig. 8B, Fig. 8D), while in per0 flies thermophase activity was about 9-fold increased and cryophase activity was reduced to less than 40% on transfer from LL to DD (Fig. 8E). It is thus suggested that the compound eye is the main source of
Fig. 5. Locomotor activity of wild type (A) and per mutant flies (B, perS; C, perL; D, per0) under LL at two constant temperature levels (30°C and 25°C). All flies were arrhythmic. The temperature step up to 30°C was made at 06:00 on day 7.
light information for the photic modulation of the rhythm under temperature cycles.
4. Discussion 4.1. Entrainment to temperature cycles It is well known that the circadian rhythms synchronize not only to light cycles but also to temperature cycles (Aschoff, 1981; Saunders, 1982). The locomotor rhythm of D. melanogaster is also known to be entrained by temperature cycle in DD (Wheeler et al., 1993). The results presented here not only confirmed the fact but also demonstrated for the first time that the entrainability to the temperature cycle was substantially affected by per mutations. Wild type flies were entrained by the temperature cycle in all lighting conditions. However, perL and perS flies were entrainable only under constant light: in DD they freeran. One may interprete the fact that those mutations in period locus reduce the entrainability to temperature cycle. However, there is another possibility that the 24 h period lenghth can be outside of their limits of entrainment because of the large period difference between the experimental and their own period. It is necessary to examine their entrainability in thermoperiod
K. Tomioka et al. / Journal of Insect Physiology 44 (1998) 587–596
Fig. 6. Thermoperiodic entrainment of the locomotor rhythm in wild type (A) and per mutant flies (B, perS; C, perL; D, per0) in constant light and darkness. LL to DD transition was performed at 18:00 on day 7. Wild type and per0 flies were entrained to a cycle of 12 h 30°C: 12h 25°C (thermophase:06:00-18:00), but the activity phase changed with the change in lighting conditions: per0 flies switched from cryoactive to thermo-active on transfer from LL to DD. Ultradian rhythms were evident in per0 fly under LL. perS and perL flies were entrained to the temperature cycle in LL but freeran in DD with weak masking effects. Rectangular frames on the left half of actograms indicate thermophase.
lengths of 19 and 29 h to draw a conclusion for this issue. In constant light, the Drosophila clock is believed to be stopped, since PER protein level does not show any rhythmic change in LL (Price et al., 1995). However, temperature cycles induced the rhythm even in LL (Fig. 6). The thermoperiodically induced rhythms in LL reflect the oscillation of the clock, since the release to DD made the rhythms freerun with peaks occurring at the same phase. A similar observation has been reported for Drosophila eclosion rhythms (Lankinen, 1993). It is now explained that LL reduces the level of TIMELESS protein and the reduced level of the protein prohibits the cycling of per expression (Lee et al., 1996; Myers et al., 1996). In this context, thermoperiodic synchronization of the rhythm under LL suggests that temperature cycles have an ability to force the autoregulatory feedback sys-
593
Fig. 7. Reentrainment of the locomotor rhythm to a 6h phase advanced (a) and a 6 h phase delayed (b) temperature cycle of 12 h 25°C:12 h 30°C in DD. Wild type flies showed several transient cycles until resynchronized to the new temperature cycle, while per0 flies appeared to resynchronize immediately after the phase shifts of the temperature cycle. The portions indicated by black frames show thermophase (30°C). Table 3 Occurrence of ultradian rhythms in thermoperiodic cycles Strain
No. of animals with ultradian rhythms DD LL
Canton-S perS perL perO
5* 1 0** 19**
(19) (36) (49) (59)
0 2 7 0
(18) (38) (38) (55)
Numbers in parenthesis indicate number of animals used.*P ⬍ 0.05; ** P ⬍ 0.01(2-test) compared with LL.
tem of per to work through some unknown process. Our preliminary results of western blotting support this hypothesis: The PER level was quite reduced in LL but still changed rhythmically with peaks in the early cryophase (wild type and perS) or from the late thermophase to early cryophase (perL)(A. Matsumoto and K. Tomioka, unpublished data).
594
K. Tomioka et al. / Journal of Insect Physiology 44 (1998) 587–596
Fig. 8. Average daily activity profiles of the locomotor rhythm in eya (A), per0;eya (B), so (C), per0;so (D), and per0 (E) under 12 h 30°C: 12 h 25°C (thermophase:06:00-18:00). per0 flies switched from cryophase active to thermophase active on transfer from LL to DD, while only a little increase in thermophase activity was observed under DD in per0;eya, and per0;so flies. Vertical bars show SEM. n, number of flies used.
4.2. Effects of temperature on circadian rhythms The effects of temperature on the freerunning period have been extensively studied in many animals (Aschoff, 1979). The freerunning period is generally temperature compensated so that the period changes only slightly with the change in temperature level, and in insects, it has been empirically shown that the higher the ambient temperature is, the more the freerunning period is lengthened (Aschoff, 1979). The present study revealed that the freerunning rhythm of perS mutant does not obey the empirical rule. In the higher temperature, the freerunning period was shortened in perS flies, while it was lengthened in perL flies. The reciprocal effect of temperature on perS and perL is consistent with the previous report by Konopka et al. (1989). Ambient temperature level has profound effects not only on the phase angle relationship of the rhythm to the light cycle but also on the distribution of locomotor activity. The temperature related change in phase has
been reported for mosquitoes (Chiba et al., 1982) and crickets (Ikeda and Tomioka, 1993), and it is generally explained to be the result of changes in freerunning periods, since it has been also shown that the phase angle relationship between the rhythm and the zeitgeber depends on the freerunning period of the rhythm (Aschoff, 1981). However, although perS and perL showed the reciprocal change in freerunning periods, their morning peak commonly significantly advanced when temperature increased (Fig. 3). The fact is inconsistent with the previous explanation and awaits an appropriate explanation. The wild type and perS flies showed a decrease of daytime activity in LD at 30°C (Fig. 2). This decrease in daytime activity is more robust in nonsense mutant per0 flies that do not have a functional clock mechanism (Wheeler et al., 1993). It is thus suggested that the temperature directly exerts its effect on the center for locomotor activity to produce rather nocturnal pattern when ambient temperature is higher. There must be a pathway
K. Tomioka et al. / Journal of Insect Physiology 44 (1998) 587–596
through which photic information is input to the center, since in the thermoperiodic conditions the activity phase depends on the lighting conditions (Fig. 6). These direct effects of temperature and light on the locomotor activity may cause masking effects in rhythmic flies. The arrhythmic mutant per0 flies are a useful model to reveal the direct effects of light and temperature on the activity by-passing the circadian clock. In constant darkness at constant temperature, per0 flies are continuously active with bouts of activity dispersed over the whole day (Fig. 4D), while in temperature cycles or light cycles, locomotor activity is concentrated in the certain phase of the cycle and the active phase is highly dependent upon lighting or temperature conditions (Fig. 1D and Fig. 6D). This means that temperature and light cooperate to suppress the activity in a certain phase and induce activity in the opposite phase: at warmer temperatures the flies tend to be more active in the dark phase, while at cooler temperatures in the daytime. The cooperation of the two environmental factors seems not to be a simple sum of their effects (Wheeler et al., 1993) but to be mediated by a certain integration mechanism. The compound eye seems to be the photoreceptor for this modulation of activity, since genetic ablation of the compound eye prohibited the photic modulation of activity pattern of per0 flies under temperature cycles (Fig. 8). 4.3. Ultradian rhythms At 30°C, many flies showed clear ultradian rhythms. This is most prominent in per0 flies. The ultradian rhythm is evidently endogenous since it persists in constant darkness (Fig. 4D). This fact suggests that Drosophila has an ultradian oscillator(s) of which expression is fairly temperature dependent. It has been suggested that the fruit flies have multiple ultradian oscillators (Dowse et al., 1987; Dowse and Ringo, 1987). On the basis of the data in which the strength of ultradian oscillation is inversely proportional to that of the circadian rhythm, Dowse and Ringo (1987) proposed that the circadian rhythm may be a result of coupling of the ultradian oscillators. However, in our data, the appearance of the ultradian rhythms is dependent on temperature even in per0 flies. The most likely explanation may be that there is an ultradian oscillator(s) which couples to activity at higher temperature but the coupling may be also substantially inhibited by the circadian clock, since the proportions of animals with ultradian rhythms are fairly low in rhythmic strains. 4.4. Underlying oscillator mechanism With elegant experiments Pittendrigh et al. (1958) proposed that the Drosophila pseudoobscura eclosion rhythm is regulated by two oscillators: one is light-sensi-
595
tive master oscillation (A-oscillator) which is instantaneously reset by light pulses and the other is a temperature sensitive slave oscillator (B-oscillator) which regulates the phase of eclosion under the control of Aoscillator. In their model, the A-oscillator unidirectionally regulates the phase of the B-oscillator and there is no feedback from the B-oscillator to the A-oscillator. This model was based on solid data and believed to underlie the eclosion rhythm of D. pseudoobscura. Our present data did not reveal that there is a temperature sensitive B-oscillator in D. melanogaster. On the contrary, the data rather suggest that the photo-entrainable clock is also temperature sensitive, since the rhythm can be entrained with the temperature cycle and it freeruns from the entrained phase. If there is the separate temperature sensitive B-oscillator then there must be a strong feedback loop to entrain the A-oscillator. The possibility of involvement of the B-oscillator should be addressed in the future work, since the temperature sensitive secondary oscillator has been reported in cockroaches (Page, 1985) and crickets (Rence and Loher, 1975; Abe et al., 1997) and there are lines of evidence showing that many cells outside the putative pacemaker neurons exhibit rhythmic expression of per (Hege et al., 1997; Emery et al., 1997). In addition, Gillanders and Saunders (1992) proposed a two oscillator model for the photoperiodic time measurement in D. melanogaster; The proposed model is constituted by a light sensitive pacemaker and a photoperiodic slave oscillation which is a damping oscillation and receiving entrainment from two sources, from the pacemaker and also from the light. It should be also addressed in the future work whether the proposed photoperiodic slave oscillation has any role in regulation of circadian locomotor rhythm.
Acknowledgements This work was partly supported by the grants from the Ministry of Education, Science, Sports and Culture of Japan.
References Abe, Y., Ushirogawa, H., Tomioka, K., 1997. Circadian locomotor rhythm of the cricket Gryllodes sigillatus. I. Localization of the pacemaker and the photoreceptor. Zoological Science 14, 719–727. Aschoff, J., 1979. Circadian rhythms: Influences of internal and external factors on the period measured in constant conditions. Zeitschrift fu¨r Tierpsychologie 49, 225–249. Aschoff, J., 1981. Entrainment. In: Aschoff, J., (Ed.), Behavioral Neurobiology, Vol. 4 – Biological Rhythms. Plenum Press, New York and London, pp. 81–93. Chiba, Y., Kubota, M., Nakamura, Y., 1982. Differential effects of temperature upon evening and morning peaks in the circadian activity of mosquitoes, Culex pipiens pallens and C. pipiens molestus. Journal of Interdisciplinary Cycle Research 13, 55–60.
596
K. Tomioka et al. / Journal of Insect Physiology 44 (1998) 587–596
Dowse, H., Hall, J.C., Ringo, J., 1987. Circadian and ultradian rhythms in period mutants of Drosophila melanogaster. Behavior Genetics 19, 529–542. Dowse, H., Ringo, J., 1987. Further evidence that the circadian clock in Drosophila is a population of coupled ultradian oscillators. Journal of Biological Rhythms 2, 65–76. Emery, I.F., Noveral, J.M., Jamison, C.F., Siwicki, K.K., 1997. Rhythms of Drosophila period gene in culture. Proceedings of National Academy of Sciences of the United States of America 94, 4092–4096. Gillanders, S.W., Saunders, D.S., 1992. A coupled pacemaker-slave model for the insect photoperiodic clock: interpretation of ovarian diapause data in Drosophila melanogaster. Biological Cybernetics 67, 451–459. Hege, D.M., Stanewsky, R., Hall, J.C., Giebultowicz, J.M., 1997. Rhythmic expression of a PER-reporter in the malpigian tubules of decapitated Drosophila: evidence for a brain-independent circadian clock. Journal of Biological Rhythms 12, 300–308. Helfrich, C., Engelmann, W., 1983. Circadian rhythm of the locomotor activity in Drosophila melanogaster and its mutants ’sine oculis’ and ’small optic lobes’. Physiological Entomology 8, 257–272. Hoffman, K., 1969. Einfluss der Zeitgebersta¨rke auf der synchronisierten circadianen Periodik. Zeitschrift fu¨r vergleichende Physiologie 62, 93–110. Ikeda, M., Tomioka, K., 1993. Temperature dependency of the circadian locomotor rhythm in the cricket Gryllus bimaculatus. Zoological Science 10, 597–604. Konopka, R.J., Benzer, S., 1971. Clock mutants of Drosophila melanogaster. Proceedings of National Academy of Sciences of the United States of America 68, 2112–2116. Konopka, R.J., Pittendrigh, C.S., Orr, D., 1989. Reciprocal behaviour associated with altered homeostasis and photosensitivity of Drosophila clock mutants. Journal of Neurogenetics 6, 1–10. Lankinen, P., 1993. Characterization of linne, a new autosomal eclosion rhythm mutant in Drosophila subobscura. Behavior Genetics 23, 359–367.
Lankinen, P., Riihimaa, A., 1997. Effects of temperature on weak circadian eclosion rhythmicity in Chymomyza costata (Diptera: Drosophilidae). Journal of Insect Physiology 43, 251–260. Lee, C., Parikh, V., Itsukaichi, T., Bae, K., Edery, I., 1996. Resetting the Drosophila clock by photic regulation of PER and a PER-TIM complex. Science 271, 1740–1744. Myers, M.P., Wager-Smith, K., Rothenfluh-Hilfiker, A., Young, M.W., 1996. Light-induced degradation of TIMELESS and entrainment of the Drosophila circadian clock. Science 271, 1736–1740. Page, T.L., 1985. Circadian organization in cockroaches: effects of temperature cycles on locomotor activity. Journal of Insect Physiology 31, 235–243. Pittendrigh, C.S., Bruce, V.G., Kaus, P., 1958. On the significance of transients in daily rhythms. Proceedings of National Academy of Sciences of the United States of America 44, 965–973. Price, J.C., Dembinska, M.E., Young, M.W., Roshbash, M., 1995. Suppression of PERIOD protein abundance and circadian cycling by the Drosophila clock mutation timeless. The EMBO Journal 14, 4044–4049. Rence, B.G., Loher, W., 1975. Arrhythmically singing crickets: thermoperiodic reentrainment after bilobectomy. Science 190, 385–387. Saunders, D.S., 1982. Insect Clocks, 2nd edn. Pergamon Press, Oxford. Sokolove, P.G., Bushell, W.N., 1978. The chi square periodogram: its utility for analysis of circadian rhythm. Journal of Theoretical Biology 72, 131–160. Tomioka, K., Uwozumi, K., Matsumoto, N., 1997. Light cycles given during development affect freerunning period of circadian locomotor rhythm of period mutants in Drosophila melanogaster. Journal of Insect Physiology 43, 297–305. Wheeler, D.A., Hamblen-Coyle, M.J., Dushay, M.S., Hall, J.C., 1993. Behavior in light-dark cycles of Drosophila mutants that are arrhythmic, blind, or both. Journal of Biological Rhythms 8, 67–94. Zimmerman, W.F., Pittendrigh, C.S., Pavlidis, T., 1968. Temperature compensation of the circadian oscillation in Drosophila pseudoobscura and its entrainment by temperature cycles. Journal of Insect Physiology 14, 669–684.
Light and temperature cooperate to regulate the circadian locomotor rhythm of wild type and period mutants of Drosophila melanogaster Kenji Tomioka a
a,*
, Makoto Sakamoto a, Yuka Harui a, Nobutaka Matsumoto a, Akira Matsumoto a,b
Department of Physics, Biology and Informatics, Faculty of Science, Yamaguchi University, Yamaguchi 753, Japan b Department of Biology, Faculty of Science, Kyushu University, Ropponmatsu, Fukuoka 810, Japan Received 24 September 1997; received in revised form 22 December 1997
Abstract In the wild type (Canton-S) and period mutant flies of Drosophila melanogaster, we examined the effects of light and temperature on the circadian locomotor rhythm. Under light dark cycles, the wild type and perS flies were diurnal at 25°C. However, at 30°C, the daytime activity commonly decreased to form a rather nocturnal pattern, and ultradian rhythms of a 2 苲 4 h period were observed more frequently than at 25°C. The change in activity pattern was more clearly observed in per0 flies, suggesting that these temperature dependent changes in activity pattern are mainly attributable to the system other than the circadian clock. In a 12 h 30°C:12 h 25°C temperature cycle (HTLT12:12), per0 flies were active during the thermophase in constant darkness (DD) but during the cryophase in constant light (LL). The results of experiments with per0;eya flies suggest that the compound eye is the main source of the photic information for this reversal. Wild type and per0 flies were synchronized to HTLT12:12 both under LL and DD, while perS and perL flies were synchronized only in LL. This suggests that the circadian clock is entrainable to the temperature cycle, but the entrainability is reduced in the perS and perL flies to this particular thermoperiod length, and that temperature cycle forces the clock to move in LL, where the rhythm is believed to be stopped at constant temperature. 1998 Elsevier Science Ltd. All rights reserved. Keywords: Circadian rhythm; Drosophila melanogaster; period mutants; Light-dark cycle; Temperature cycle
1. Introduction It is well known that ambient temperature profoundly affects the circadian rhythm in many insects. In general, ambient temperature has three distinct effects. First, it affects the period of freerunning rhythms. This effect has been investigated in a wide variety of animals and plants and it was found that temperature changes the freerunning period slightly but significantly in many organisms including insects (Aschoff, 1979). The Q10 is generally very close to 1.0. Second, the phase of the rhythm is often profoundly affected by an ambient temperature level. For example, the cricket Gryllus bimaculatus is nocturnally active at 25°C but becomes diurnal when * Corresponding author. Tel and Fax: + 81-839-33-5714; E-mail: [email protected]
0022–1910 /98 /$19.00 1998 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 2 - 1 9 1 0 ( 9 8 ) 0 0 0 4 6 - 8
exposed to 20°C (Ikeda and Tomioka, 1993). In the third effect, temperature cycle is well known to entrain the circadian rhythm in some animals such as lizards (Hoffman, 1969) and insects (Zimmerman et al., 1968; Lankinen and Riihimaa, 1997; also see Saunders, 1982). However, the mechanism through which temperature affects the rhythm still remains unclear. In Drosophila pseudoobscura temperature plays an important role in setting the phase of the eclosion rhythm. Based on the results of a series of elegant kinetical experiments, Pittendrigh et al. (1958) proposed the famous two oscillator model including a light-sensitive master oscillator and a temperature sensitive slave oscillator. They explained that the eclosion behavior is under direct control of the temperature sensitive oscillator that is in turn entrained by the light sensitive master oscillator. The applicability of the two oscillator model has not been thoroughly investigated in Drosophila melanogaster.
588
K. Tomioka et al. / Journal of Insect Physiology 44 (1998) 587–596
There are three types of period (per, a clock gene) mutant flies of D. melanogaster, i.e. perS, perL and per0, which behave as alleles of a single locus of the X chromosome. perS shortens the period from 24 h to 19 h, perL lengthens it to 29 h, and per0 is arrhythmic (Konopka and Benzer, 1971). It has been reported that perS and perL mutant flies have properties reciprocal to one another with respect to temperature and light intensity dependency of their freerunning periods (Konopka et al., 1989) and that wild type and per0 flies are entrainable to temperature cycles in DD (Wheeler et al., 1993). In the present study, we examined the effects of change in ambient temperature level on the locomotor rhythm and entrainability of the rhythm to temperature cycles under various lighting schedules in wild type and per mutant flies. The results show that the light and temperature have considerable effects on the activity pattern and that the light sensitive clock is also entrainable to temperature cycles, but the entrainability is reduced in perS and perL mutant flies. Since the arrhythmic, per0 flies were most sensitive to both light and temperature, it is suggested that the light and temperature sensitive driven system is normally overcome by the pacemaker in rhythmic flies.
2. Materials and methods 2.1. Rearing of flies The fruit fly, Drosophila melanogaster, was grown on glucose-cornmeal-yeast medium at 25°C under a light cycle with 12 h light to 12 h dark. The strains that we used for the behavioral assay were Canton-S, perS, perL1 and per01. We conventionally call perL1 and per01 as perL and per0, respectively, throughout. We also used per0;so which lacks the compound eyes and ocelli and per0;eya which lacks the compound eyes. 2.2. Recording of locomotor activity rhythms Male flies of five days old were individually housed in transparent acrylic rectangle tubes (3 × 3 × 70 mm). The tube was plugged at one end with agar/glucose medium as food and was sealed with paraffin, and at the other end with a silicon tube filled with damped absorbent cotton as a water source. A moving fly interrupted an infra-red beam and the number of interruptions during each 6 min was recorded using a computerized system (Tomioka et al., 1997). 2.3. Light and temperature conditions Light cycles were given by a cool white fluorescent lamp in the environment controlled room connected to an electric timer. Light intensity at the animals level was
100 to 300 lux and varied with the proximity to the lamp. Temperature cycles were set by a built-in thermostat driven by an electronic timer. Temperature steps up and down were finished within 15 min. 2.4. Estimation of circadian parameters Freerunning periods of the locomotor rhythm were estimated by the chi-square periodogram (Sokolove and Bushell, 1978). Existence of ultradian rhythms was also analyzed by the chi-square periodogram and if peaks of the periodogram appeared above the 0.05 confidence level, the animal was designated as rhythmic. Drosophila melanogaster shows bimodal activity rhythm peaking around lights-on (morning peak) and lights-off (evening peak). The time of peaks was determined from the daily activity average curves based on more than 5 days.
3. Results 3.1. Locomotor rhythms under a light cycle at different ambient temperature levels All the wild type and mutant flies were clearly synchronized to LD12:12 at two different ambient temperature levels of 25°C and 30°C (Fig. 1). perL flies consistently showed nocturnal activity rhythm in both temperature levels. Although night time activity tended to similarly decrease both when transferred from 30°C to 25°C and from 25°C to 30°C (Fig. 2C), the daytime activity often appeared clearly when transferred from 30°C to 25°C, forming a rather diurnal component (Fig. 1C). In 25°C, wild type and perS flies showed clear bimodal rhythms peaking around lights-on and before lightsoff with the activity being concentrated in the light phase, while at 30°C the daytime activity tended to decrease and on the contrary the night time activity remarkably increased, making activity rhythm rather nocturnal (Fig. 1A and B). per0 flies showed the most drastic change in activity pattern (Fig. 1D): They were diurnal in 25°C, while they turned rather nocturnal at 30°C. However, the highest peak occurred at lights-off independent of the temperature level. Statistical analysis revealed that a decrease in daytime activity at a higher temperature was commonly observed in all the strains (Figs 1 and 2). The night time activity was either significantly greater in higher temperature (perS) or remained at a similar level (per0), resulting in the night time activity being greater than daytime activity with the elevation of temperature level (Fig. 2). The phase of morning and evening peaks also slightly changed in association with the temperature level (Fig. 3). The morning peak commonly significantly advanced by a few hours with the elevation of the temperature.
K. Tomioka et al. / Journal of Insect Physiology 44 (1998) 587–596
589
Fig. 4 shows representative records of locomotor activity of the four strains of Drosophila under DD. per0 flies were arrhythmic both under 25°C and 30°C: activity equally dispersed over 24 h but ultradian rhythmicity with periods of about 3 h was evident particularly in 30°C. Flies of all other strains showed clear freerunning rhythms under DD. The change in temperature level did not significantly affect the phase of the rhythm but slightly the freerunning period (Table 2). The rhythms of wild type files showed a very slight change of freerunning period in association with temperature change (P > 0.27, t-test). However, in the two mutant flies, perS and perL, the freerunning period changed significantly with temperature change. The two mutant alleles responded to a decrease in temperature in reciprocal ways: perS lengthened its period, whereas perL shortened (Fig. 4, Table 2). Under constant light, all strains were arrhythmic at 25°C, and the arrhythmic pattern persisted in the majority of flies even when temperature was stepped up by 5°C (Fig. 5). 3.3. Locomotor rhythms under temperature cycles in constant lighting conditions
Fig. 1. Locomotor activity of wild type (A) and per mutant flies (B, perS; C, perL; D, per0) under a 12 h light to 12 h dark cycle at two constant temperature levels (30°C and 25°C). Both temperatures step up to 30°C and step down to 25°C were made at 18:00 on day 7. All flies were entrained to the LD but with the peak changing with the change of temperature level. With higher temperatures the daytime activity decreased and the night time activity increased in wild type (A) and perS strains (B). perL flies (C) were rather consistently nocturnal. per0 flies (D) switched from diurnally-active to nocturnally-active. Ultradian rhythms became prominent on transfer from 25°C to 30°C (A, B and D). Black and white bars indicate the dark and light portions, respectively.
However, the evening peak either significantly delayed (Canton-S and perS), advanced (perL) or unchanged (per0). It is notable that the chi-square periodogram analysis revealed ultradian rhythms with periods of 2-4 h in some flies. In perS and per0 flies, the number of flies showing the ultradian rhythm were significantly higher in 30°C than in 25°C (Table 1, P ⬍ 0.05, 2-test). 3.2. Rhythms under constant darkness and constant light at different temperature levels We then examined the effects of temperature on the locomotor rhythms under constant lighting conditions.
To examine whether the flies can be entrained to temperature cycles, we recorded the locomotor rhythm under a temperature cycle of 12 h high temperature (30°C) and 12 h low temperature (25°C)(HTLT12:12). Under LL conditions, flies of the four strains were synchronized to the temperature cycle. In wild-type and per0 flies the activity pattern drastically changed with the change in temperature level. Locomotor activity of wild type flies peaked twice per cycle around temperature step-up and before step-down. These two peaks apparently correspond to the morning and evening peaks under LD. The morning peak was very faint but the evening peak continued into early cryophase (Fig. 6A). The activity of per0 was concentrated in cryophase (Fig. 6D). perS and perL flies were also synchronized to the HTLT cycle under LL (Fig. 6B, Fig. 6C). The perS flies showed a trimodal activity with peaks at temperature step-up, down, and middle of the thermophase, while perL flies peaked twice, at the beginning of both thermophase and cryophase. In constant darkness, the locomotor rhythms of wild type and per0 flies clearly synchronized with the HTLT12:12 (Fig. 6). The pattern of synchronized locomotor rhythms is quite similar to that reported by Wheeler et al. (1993). In the wild type flies, the morning peak occurred several hours prior to the temperature step-up. When the temperature cycle was phase advanced or delayed by 6h by shortening or lengthening the thermophase, the rhythm was reentrained to the shifted cycle with several transient cycles (Fig. 7). The transient cycles were 5.5 ± 1.0 days and 5.1 ± 1.2 days
590
K. Tomioka et al. / Journal of Insect Physiology 44 (1998) 587–596
Fig. 2. Activity levels of day (open square) and night (closed square) in LD12:12 at different temperature levels in wild type (A) and per mutant flies (B, perS; C, perL; D, per0). Each point represents the mean of 10 to 49 animals. Vertical bars indicate SEM. Arrows indicate the direction of animal transfer from 25°C to 30°C or from 30°C to 25°C. Examples of rhythms are shown in Fig. 1.
for phase delay and advance, respectively. per0 flies showed a very clear rhythm with activity concentrated in the thermophase (Fig. 6D). A major peak occurred just after the temperature rose. When the temperature cycle was shifted, the activity phase almost immediately
shifted to synchronize with the new thermophase (Fig. 7). In perL and perS flies, however, the rhythms did not synchronize to the temperature cycle in DD but freeran with masking effects of temperature: The activity level
K. Tomioka et al. / Journal of Insect Physiology 44 (1998) 587–596
591
Fig. 3. Phases of morning peak (open circles) and evening peak (closed circles) of wild type (A) and per mutant flies (B, perS; C, perL; D, per0) in LD12:12 at different temperature levels. Each point represents the mean of 42 to 81 animals. Horizontal bars show SD. Asterisks show the value significantly different from that in 25°C (**P ⬍ 0.01, Student’s t-test). Examples of rhythms are shown in Fig. 1.
Strain
No. of animals with ultradian rhythms 30°C 25°C
per0 flies, the occurrence of the ultradian rhythm was significantly higher in DD where the activity was concentrated in the thermophase (2-test, P ⬍ 0.05 and P ⬍ 0.01, respectively). In contrast, it was higher in LL in perL (2-test, P ⬍ 0.01).
Canton-S perS perL perO
11 10* 6 20*
3.4. Locomotor rhythms under temperature cycles in constant lighting conditions in per0 flies lacking external photoreceptors
Table 1 Occurrence of ultradian rhythms at different temperature levels at LD12:12
(81) (77) (55) (69)
3 2 8 8
(63) (72) (42) (63)
Numbers in parenthesis indicate number of animals used. *P ⬍ 0.05 (2-test).
of perS and perL was higher when the activity occurred in the thermophase and cryophase, respectively (Fig. 6B, Fig. 6C). The freerunning component of perL and perS flies corresponded to the activity component occurring at temperature step-down and at the middle of the thermophase, respectively, in the previous LL. The chi-square periodogram analysis revealed ultradian rhythms in some flies (Table 3). In wild-type and
It is well known that Drosophila has an extraretinal photoreceptor for photic entrainment of the circadian rhythm (Helfrich and Engelmann, 1983; Wheeler et al., 1993). To examine the photoreceptor responsible for the phase regulation in per0 flies, we recorded locomotor activity in eya and so flies that lack the compound eyes and both compound eyes and ocelli, respectively, and their double mutant with per0, i.e., per0;eya and per0;so, under temperature cycles with different lighting conditions. In both constant light and constant darkness, locomotor rhythms of all these single and double mutant
592
K. Tomioka et al. / Journal of Insect Physiology 44 (1998) 587–596
Fig. 4. Locomotor activity of wild type (A) and per mutant flies (B, perS; C, perL; D, per0) under constant darkness at two constant temperature levels (30°C and 25°C). With lower temperature the freerunning period of the rhythm was lengthened in wild type and perS strains, but shortened in perL. Although per0 fly was consistently arrhythmic, ultradian rhythm was prominent in 30°C. Table 2 Effects of ambient temperature on freerunning period in DD Strain
Freerunning period (mean ± SD h) 25°C 30°C
Canton-S perS perL perO
24.28 ± 0.45 (5) 19.17 ± 0.41 (9) 31.2 ± 0.67 (7) arrythmic
24.05 ± 0.80 (11) 18.83 ± 0.30* (14) 32.22 ± 0.55** (6) arrhythmic
Numbers in parenthesis indicate number of animals used. *P ⬍ 0.05; **P ⬍ 0.01(t-test) compared with 25°C.
flies well synchronized to the temperature cycle, peaking twice, at the beginning of the warm phase and at the beginning of the cold phase (Fig. 8). The activity pattern of eya and so flies showed little change associated with lighting conditions. Similarly, the pattern was little affected by the lighting conditions in both double mutant strains: only at most a two-fold increase in thermophase activity was observed in DD (Fig. 8B, Fig. 8D), while in per0 flies thermophase activity was about 9-fold increased and cryophase activity was reduced to less than 40% on transfer from LL to DD (Fig. 8E). It is thus suggested that the compound eye is the main source of
Fig. 5. Locomotor activity of wild type (A) and per mutant flies (B, perS; C, perL; D, per0) under LL at two constant temperature levels (30°C and 25°C). All flies were arrhythmic. The temperature step up to 30°C was made at 06:00 on day 7.
light information for the photic modulation of the rhythm under temperature cycles.
4. Discussion 4.1. Entrainment to temperature cycles It is well known that the circadian rhythms synchronize not only to light cycles but also to temperature cycles (Aschoff, 1981; Saunders, 1982). The locomotor rhythm of D. melanogaster is also known to be entrained by temperature cycle in DD (Wheeler et al., 1993). The results presented here not only confirmed the fact but also demonstrated for the first time that the entrainability to the temperature cycle was substantially affected by per mutations. Wild type flies were entrained by the temperature cycle in all lighting conditions. However, perL and perS flies were entrainable only under constant light: in DD they freeran. One may interprete the fact that those mutations in period locus reduce the entrainability to temperature cycle. However, there is another possibility that the 24 h period lenghth can be outside of their limits of entrainment because of the large period difference between the experimental and their own period. It is necessary to examine their entrainability in thermoperiod
K. Tomioka et al. / Journal of Insect Physiology 44 (1998) 587–596
Fig. 6. Thermoperiodic entrainment of the locomotor rhythm in wild type (A) and per mutant flies (B, perS; C, perL; D, per0) in constant light and darkness. LL to DD transition was performed at 18:00 on day 7. Wild type and per0 flies were entrained to a cycle of 12 h 30°C: 12h 25°C (thermophase:06:00-18:00), but the activity phase changed with the change in lighting conditions: per0 flies switched from cryoactive to thermo-active on transfer from LL to DD. Ultradian rhythms were evident in per0 fly under LL. perS and perL flies were entrained to the temperature cycle in LL but freeran in DD with weak masking effects. Rectangular frames on the left half of actograms indicate thermophase.
lengths of 19 and 29 h to draw a conclusion for this issue. In constant light, the Drosophila clock is believed to be stopped, since PER protein level does not show any rhythmic change in LL (Price et al., 1995). However, temperature cycles induced the rhythm even in LL (Fig. 6). The thermoperiodically induced rhythms in LL reflect the oscillation of the clock, since the release to DD made the rhythms freerun with peaks occurring at the same phase. A similar observation has been reported for Drosophila eclosion rhythms (Lankinen, 1993). It is now explained that LL reduces the level of TIMELESS protein and the reduced level of the protein prohibits the cycling of per expression (Lee et al., 1996; Myers et al., 1996). In this context, thermoperiodic synchronization of the rhythm under LL suggests that temperature cycles have an ability to force the autoregulatory feedback sys-
593
Fig. 7. Reentrainment of the locomotor rhythm to a 6h phase advanced (a) and a 6 h phase delayed (b) temperature cycle of 12 h 25°C:12 h 30°C in DD. Wild type flies showed several transient cycles until resynchronized to the new temperature cycle, while per0 flies appeared to resynchronize immediately after the phase shifts of the temperature cycle. The portions indicated by black frames show thermophase (30°C). Table 3 Occurrence of ultradian rhythms in thermoperiodic cycles Strain
No. of animals with ultradian rhythms DD LL
Canton-S perS perL perO
5* 1 0** 19**
(19) (36) (49) (59)
0 2 7 0
(18) (38) (38) (55)
Numbers in parenthesis indicate number of animals used.*P ⬍ 0.05; ** P ⬍ 0.01(2-test) compared with LL.
tem of per to work through some unknown process. Our preliminary results of western blotting support this hypothesis: The PER level was quite reduced in LL but still changed rhythmically with peaks in the early cryophase (wild type and perS) or from the late thermophase to early cryophase (perL)(A. Matsumoto and K. Tomioka, unpublished data).
594
K. Tomioka et al. / Journal of Insect Physiology 44 (1998) 587–596
Fig. 8. Average daily activity profiles of the locomotor rhythm in eya (A), per0;eya (B), so (C), per0;so (D), and per0 (E) under 12 h 30°C: 12 h 25°C (thermophase:06:00-18:00). per0 flies switched from cryophase active to thermophase active on transfer from LL to DD, while only a little increase in thermophase activity was observed under DD in per0;eya, and per0;so flies. Vertical bars show SEM. n, number of flies used.
4.2. Effects of temperature on circadian rhythms The effects of temperature on the freerunning period have been extensively studied in many animals (Aschoff, 1979). The freerunning period is generally temperature compensated so that the period changes only slightly with the change in temperature level, and in insects, it has been empirically shown that the higher the ambient temperature is, the more the freerunning period is lengthened (Aschoff, 1979). The present study revealed that the freerunning rhythm of perS mutant does not obey the empirical rule. In the higher temperature, the freerunning period was shortened in perS flies, while it was lengthened in perL flies. The reciprocal effect of temperature on perS and perL is consistent with the previous report by Konopka et al. (1989). Ambient temperature level has profound effects not only on the phase angle relationship of the rhythm to the light cycle but also on the distribution of locomotor activity. The temperature related change in phase has
been reported for mosquitoes (Chiba et al., 1982) and crickets (Ikeda and Tomioka, 1993), and it is generally explained to be the result of changes in freerunning periods, since it has been also shown that the phase angle relationship between the rhythm and the zeitgeber depends on the freerunning period of the rhythm (Aschoff, 1981). However, although perS and perL showed the reciprocal change in freerunning periods, their morning peak commonly significantly advanced when temperature increased (Fig. 3). The fact is inconsistent with the previous explanation and awaits an appropriate explanation. The wild type and perS flies showed a decrease of daytime activity in LD at 30°C (Fig. 2). This decrease in daytime activity is more robust in nonsense mutant per0 flies that do not have a functional clock mechanism (Wheeler et al., 1993). It is thus suggested that the temperature directly exerts its effect on the center for locomotor activity to produce rather nocturnal pattern when ambient temperature is higher. There must be a pathway
K. Tomioka et al. / Journal of Insect Physiology 44 (1998) 587–596
through which photic information is input to the center, since in the thermoperiodic conditions the activity phase depends on the lighting conditions (Fig. 6). These direct effects of temperature and light on the locomotor activity may cause masking effects in rhythmic flies. The arrhythmic mutant per0 flies are a useful model to reveal the direct effects of light and temperature on the activity by-passing the circadian clock. In constant darkness at constant temperature, per0 flies are continuously active with bouts of activity dispersed over the whole day (Fig. 4D), while in temperature cycles or light cycles, locomotor activity is concentrated in the certain phase of the cycle and the active phase is highly dependent upon lighting or temperature conditions (Fig. 1D and Fig. 6D). This means that temperature and light cooperate to suppress the activity in a certain phase and induce activity in the opposite phase: at warmer temperatures the flies tend to be more active in the dark phase, while at cooler temperatures in the daytime. The cooperation of the two environmental factors seems not to be a simple sum of their effects (Wheeler et al., 1993) but to be mediated by a certain integration mechanism. The compound eye seems to be the photoreceptor for this modulation of activity, since genetic ablation of the compound eye prohibited the photic modulation of activity pattern of per0 flies under temperature cycles (Fig. 8). 4.3. Ultradian rhythms At 30°C, many flies showed clear ultradian rhythms. This is most prominent in per0 flies. The ultradian rhythm is evidently endogenous since it persists in constant darkness (Fig. 4D). This fact suggests that Drosophila has an ultradian oscillator(s) of which expression is fairly temperature dependent. It has been suggested that the fruit flies have multiple ultradian oscillators (Dowse et al., 1987; Dowse and Ringo, 1987). On the basis of the data in which the strength of ultradian oscillation is inversely proportional to that of the circadian rhythm, Dowse and Ringo (1987) proposed that the circadian rhythm may be a result of coupling of the ultradian oscillators. However, in our data, the appearance of the ultradian rhythms is dependent on temperature even in per0 flies. The most likely explanation may be that there is an ultradian oscillator(s) which couples to activity at higher temperature but the coupling may be also substantially inhibited by the circadian clock, since the proportions of animals with ultradian rhythms are fairly low in rhythmic strains. 4.4. Underlying oscillator mechanism With elegant experiments Pittendrigh et al. (1958) proposed that the Drosophila pseudoobscura eclosion rhythm is regulated by two oscillators: one is light-sensi-
595
tive master oscillation (A-oscillator) which is instantaneously reset by light pulses and the other is a temperature sensitive slave oscillator (B-oscillator) which regulates the phase of eclosion under the control of Aoscillator. In their model, the A-oscillator unidirectionally regulates the phase of the B-oscillator and there is no feedback from the B-oscillator to the A-oscillator. This model was based on solid data and believed to underlie the eclosion rhythm of D. pseudoobscura. Our present data did not reveal that there is a temperature sensitive B-oscillator in D. melanogaster. On the contrary, the data rather suggest that the photo-entrainable clock is also temperature sensitive, since the rhythm can be entrained with the temperature cycle and it freeruns from the entrained phase. If there is the separate temperature sensitive B-oscillator then there must be a strong feedback loop to entrain the A-oscillator. The possibility of involvement of the B-oscillator should be addressed in the future work, since the temperature sensitive secondary oscillator has been reported in cockroaches (Page, 1985) and crickets (Rence and Loher, 1975; Abe et al., 1997) and there are lines of evidence showing that many cells outside the putative pacemaker neurons exhibit rhythmic expression of per (Hege et al., 1997; Emery et al., 1997). In addition, Gillanders and Saunders (1992) proposed a two oscillator model for the photoperiodic time measurement in D. melanogaster; The proposed model is constituted by a light sensitive pacemaker and a photoperiodic slave oscillation which is a damping oscillation and receiving entrainment from two sources, from the pacemaker and also from the light. It should be also addressed in the future work whether the proposed photoperiodic slave oscillation has any role in regulation of circadian locomotor rhythm.
Acknowledgements This work was partly supported by the grants from the Ministry of Education, Science, Sports and Culture of Japan.
References Abe, Y., Ushirogawa, H., Tomioka, K., 1997. Circadian locomotor rhythm of the cricket Gryllodes sigillatus. I. Localization of the pacemaker and the photoreceptor. Zoological Science 14, 719–727. Aschoff, J., 1979. Circadian rhythms: Influences of internal and external factors on the period measured in constant conditions. Zeitschrift fu¨r Tierpsychologie 49, 225–249. Aschoff, J., 1981. Entrainment. In: Aschoff, J., (Ed.), Behavioral Neurobiology, Vol. 4 – Biological Rhythms. Plenum Press, New York and London, pp. 81–93. Chiba, Y., Kubota, M., Nakamura, Y., 1982. Differential effects of temperature upon evening and morning peaks in the circadian activity of mosquitoes, Culex pipiens pallens and C. pipiens molestus. Journal of Interdisciplinary Cycle Research 13, 55–60.
596
K. Tomioka et al. / Journal of Insect Physiology 44 (1998) 587–596
Dowse, H., Hall, J.C., Ringo, J., 1987. Circadian and ultradian rhythms in period mutants of Drosophila melanogaster. Behavior Genetics 19, 529–542. Dowse, H., Ringo, J., 1987. Further evidence that the circadian clock in Drosophila is a population of coupled ultradian oscillators. Journal of Biological Rhythms 2, 65–76. Emery, I.F., Noveral, J.M., Jamison, C.F., Siwicki, K.K., 1997. Rhythms of Drosophila period gene in culture. Proceedings of National Academy of Sciences of the United States of America 94, 4092–4096. Gillanders, S.W., Saunders, D.S., 1992. A coupled pacemaker-slave model for the insect photoperiodic clock: interpretation of ovarian diapause data in Drosophila melanogaster. Biological Cybernetics 67, 451–459. Hege, D.M., Stanewsky, R., Hall, J.C., Giebultowicz, J.M., 1997. Rhythmic expression of a PER-reporter in the malpigian tubules of decapitated Drosophila: evidence for a brain-independent circadian clock. Journal of Biological Rhythms 12, 300–308. Helfrich, C., Engelmann, W., 1983. Circadian rhythm of the locomotor activity in Drosophila melanogaster and its mutants ’sine oculis’ and ’small optic lobes’. Physiological Entomology 8, 257–272. Hoffman, K., 1969. Einfluss der Zeitgebersta¨rke auf der synchronisierten circadianen Periodik. Zeitschrift fu¨r vergleichende Physiologie 62, 93–110. Ikeda, M., Tomioka, K., 1993. Temperature dependency of the circadian locomotor rhythm in the cricket Gryllus bimaculatus. Zoological Science 10, 597–604. Konopka, R.J., Benzer, S., 1971. Clock mutants of Drosophila melanogaster. Proceedings of National Academy of Sciences of the United States of America 68, 2112–2116. Konopka, R.J., Pittendrigh, C.S., Orr, D., 1989. Reciprocal behaviour associated with altered homeostasis and photosensitivity of Drosophila clock mutants. Journal of Neurogenetics 6, 1–10. Lankinen, P., 1993. Characterization of linne, a new autosomal eclosion rhythm mutant in Drosophila subobscura. Behavior Genetics 23, 359–367.
Lankinen, P., Riihimaa, A., 1997. Effects of temperature on weak circadian eclosion rhythmicity in Chymomyza costata (Diptera: Drosophilidae). Journal of Insect Physiology 43, 251–260. Lee, C., Parikh, V., Itsukaichi, T., Bae, K., Edery, I., 1996. Resetting the Drosophila clock by photic regulation of PER and a PER-TIM complex. Science 271, 1740–1744. Myers, M.P., Wager-Smith, K., Rothenfluh-Hilfiker, A., Young, M.W., 1996. Light-induced degradation of TIMELESS and entrainment of the Drosophila circadian clock. Science 271, 1736–1740. Page, T.L., 1985. Circadian organization in cockroaches: effects of temperature cycles on locomotor activity. Journal of Insect Physiology 31, 235–243. Pittendrigh, C.S., Bruce, V.G., Kaus, P., 1958. On the significance of transients in daily rhythms. Proceedings of National Academy of Sciences of the United States of America 44, 965–973. Price, J.C., Dembinska, M.E., Young, M.W., Roshbash, M., 1995. Suppression of PERIOD protein abundance and circadian cycling by the Drosophila clock mutation timeless. The EMBO Journal 14, 4044–4049. Rence, B.G., Loher, W., 1975. Arrhythmically singing crickets: thermoperiodic reentrainment after bilobectomy. Science 190, 385–387. Saunders, D.S., 1982. Insect Clocks, 2nd edn. Pergamon Press, Oxford. Sokolove, P.G., Bushell, W.N., 1978. The chi square periodogram: its utility for analysis of circadian rhythm. Journal of Theoretical Biology 72, 131–160. Tomioka, K., Uwozumi, K., Matsumoto, N., 1997. Light cycles given during development affect freerunning period of circadian locomotor rhythm of period mutants in Drosophila melanogaster. Journal of Insect Physiology 43, 297–305. Wheeler, D.A., Hamblen-Coyle, M.J., Dushay, M.S., Hall, J.C., 1993. Behavior in light-dark cycles of Drosophila mutants that are arrhythmic, blind, or both. Journal of Biological Rhythms 8, 67–94. Zimmerman, W.F., Pittendrigh, C.S., Pavlidis, T., 1968. Temperature compensation of the circadian oscillation in Drosophila pseudoobscura and its entrainment by temperature cycles. Journal of Insect Physiology 14, 669–684.