Anisomycin induces phase shifts of circadian pacemaker in primary cultures of rat suprachiasmatic nucleus

BRAIN RESEARCH ELSEVIER

Brain Research684 (1995) 179-184

Research report

Anisomycin induces phase shifts of circadian pacemaker in primary cultures of rat suprachiasmatic nucleus Kazuto Watanabe a, Tatsuya Katagai b, Norio Ishida b, Sadao Yamaoka a, * a Departmentof Physiology, Dokkyo UniversitySchool of Medicine, Mibu, Tochigi321-02, Japan b Cell Biology laboratory, National Institute of Bioscience and Human Technology,AIST, M1TI, TsukubaScience City, Ibaraki 305, Japan

Accepted 14 March 1995

Abstract

We have developed a cell culture system for the rat suprachiasmatic nucleus, in which a clear circadian oscillation of vasopressin release was observed. Using thJis culture system, the effect of anisomycin, an inhibitor of protein synthesis, on the circadian rhythm was studied. A phase-delay of more than 15 h could be produced by a 6-h anisomycin pulse. The magnitude of the phase-shift was dependent on the circadian time of the drug treatment and on its dose. The phase-response curve was similar to the response curves that have been measured for protein synthesis inhibitors in other organisms. These results strongly suggest that protein synthesis may be involved in the generation of circadian rhythm,; in mammals. Keywords: Circadian rhythm; Suprachiasmatic nucleus; Pacemaker; Protein synthesis; Anisomycin; Cell culture; Antisense; Vasopressin

1. Introduction

Circadian rhythmicities have been documented at every level of organization. However, the cellular and molecular mechanisms of the circadian oscillation, which might be essentially common, yet remain unknown. Changes in circadian rhythmicity by administration of various metabolic inhibitors have been studied in an attempt to clarify the mechanism of rhythm generation [2,3,7,14]. In microorganisms, invertebrates and birds, inhibitors of protein synthesis on 80 S ribosomes, e.g. anisomycin, cycloheximide and puromycin, produce changes in the period length or phase of circadian rhythms [1,4,6,8,9,13,17]. It has been also reported that an injection of protein synthesis inhibitors into hamsters induce phase-shifts in the circadian rhythm of locomotor activity [5,11,16,19]. Although the results obtained in hanlsters are similar to those in other organisms, inhibitors (even when they are injected into the suprachiasmatic nucleus (SCN) region) induce only one or two hours of phase-shift in hamsters, whereas they induce several hours of phase-shifts in most of other organisms. Since the large phase-shifts induced by the inhibitors in other organisms have all been observed in

* Correspondingauthor. Fax: (81L)(282) 86-7835. 0006-8993/95/$09.50 © 1995 El,~evierScienceB.V. All fights reserved SSDI 0006-8993(95)00414-9

vitro, it might be possible to induce large phase-shifts in mammalian pacemaker using cell culture system. In mammals, the main pacemaker of circadian rhythms has been considered to be located in the SCN of anterior hypothalamus [10]. Recently, we developed a cell culture system for the SCN [17,18]. In this culture system, (1) the release of arginine vasopressin (AVP) shows a clear circadian oscillation, (2) the circadian phase of AVP release is almost the same among the replicated cultures (peaks at subjective day), and (3) the circadian oscillation persists over one month with a free-running period close to 24 h. In the present study, we examined the effects of anisomycin on the mammalian circadian pacemaker using this cell culture system for the SCN.

2. Materials and methods 2.1. Cell isolation and culture

The SCN cells were isolated from 4- to 6-day-old Sprague-Dawley rats which were maintained on a 14:10-h light/dark cycle (light 05:00-19:00). Methods of cell isolation and culture were essentially the same as those of Murakami et al. [12]. The detailed procedure has been described elsewhere [17,18]. In static culture, cells were

180

K. Watanabe et al. /Brain Research 684 (1995) 179-184

cultured on a cloning plate (Greiner, Frickenhausen, Germany) coated with poly-L-ornithine. In perfusion culture, a 96-well plate (Falcon) coated with poly-L-ornithine was used. A pair of neighboring wells was connected by perforating the wall in advance. The cells were plated on one of the paired wells. In both cultures the cells were maintained for at least 1 week, and then the measurement of circadian oscillation was started. 2.2. Measurement o f circadian oscillation

In static culture, the medium was changed at 6-h intervals (at 04:00, 10:00, 16:00, 22:00) and the amount of A V P in the medium was measured by radioimmunoassay using A V P - R I A kit (Mitsubishi Petrochemical Co., Tokyo) as described previously [17]. To avGid giving temperature pulses, the fresh medium which had been kept in the same incubator was used. The conditions of the fresh medium, such as temperature and pH (CO2-saturation), are nearly the same as those of the used one. Anisomycin pulses of 6 h were applied at selected circadian times. The applied dosages were in general 50 /xM, and otherwise 2, 5 or 10 /xM. At the end of an anisomycin pulse, the medium was changed twice. When cultures received two anisomycin administrations, they were separated by more than 10 days. In peffusion culture, the pair of wells were set with silastic stoppers connected tubing, and medium was perfused from one of the paired wells to another one by a infusion pump. The flow rate was about 170 /xl/h. The medium was collected at 4-h intervals with fraction collector and the amount of AVP in the medium was measured. At selected circadian times, 50 /~1 of 1 mM anisomycin was injected into each well. After the anisomycin pulse, the reagent was left to be washed out slowly at the same flow rate. While the perfusion cultures were kept in constant darkness during the measurement, the static cultures were not lightcontrolled.

and best '~b' were determined). In other cases, we assumed the cultures had a period of 24 h (the best '~b' was obtained from the fixed 'to'), since the sampling period is too short to estimate a meaningful free-running period and the free-running period in our system is very close to 24 h (23.7 + 0.3 h (mean _ S.D.)) [17]. Phase shifts were determined by comparing two best fitted curves (pre- and post-pulse) in the same culture. The magnitude of phaseshift was obtained as the difference in phase between the curves at the onset time of the pulse. Since the difference in estimated phase among untreated individual cultures from the same cell preparation (S.D.) was 0.43 + 0.10 h (mean _ S.E.M.), phase shifts greater than one hour were considered significant. The direction of big phase shifts 30

A i-" CO

20

O. U~

h..

10

ft.

> ,<

0 0

12

0

12

0

12

Time of Day (hr)

3O

B

E

Ill

2.3. Antisense experiments 0

Antisense oligonucleotides were synthesized on an Applied Biosystem 381A D N A synthesizer and purified by an OPC column. The sequence of oligomers designed to recognize 21 nucleotides from the translation initiation codon (ATG) were as follows; Antisense AVP, 5-GTT G A G C - A T C A T G G C G A G C A T - 3 ; Control, 5-GATCCTTGTI'CATACAGCTCA-3. 2.4. Data analysis

For determination of circadian phase, the best fitted cosine curve ( [ M + A c o s ( t o t +
0 m 10

>

Time of Day (hr)

Fig. 1. Circadian oscillation of AVP release in cultures of the SCN cells. A: Static culture. Cells were cultured on a cloning plate for one week, and circadian oscillations of AVP release were measured. The midpoint of the sampling time is denoted by a symbol. Different symbols represent the results of different cultures from the same cell preparation. B: Perfusion culture. Cells were cultured on a 96-well plate for 1 week, and perfusion was started. The results are of 3 out of 4 cultures from the same cell preparation.

K. Watanabe et aL / Brain Research 684 (1995) 179-184

(advance or delay) was verified by measuring responses to lower doses of anisomycin to determine in which direction the magnitude of the phase shift changed. In this experiment, the circadian time (CT) refers to the oscillation of AVP release. Since the time-of-peak on the second day of culture correspond to about 15 h after the onset of darkness in the previous light/dark cycle [18], we designated this phase (peak of AVP release) CT 3.

181

phase was almost the same among the multiple cultures from the same cell preparation, as reported previously [17]. A similar circadian oscillation was also observed in the perfusion cultures (Fig. 1B). In some perfusion cultures, the tube was obstructed by detached cells and we had to terminate the experiment. When a 6-h anisomycin pulse was applied at the peak of AVP release on the static cultures, the amount of AVP release was reduced (Fig. 2A). Although the release was recovered after removal of anisomycin, the circadian phase was almost reversed. On the other hand, the pulse given at the trough of AVP release had little effect on the amount of AVP release and on its phase (Fig. 2B). Similar phaseshifts were induced by a single injection of anisomycin

3. Results Clear circadian oscillations in AVP release were observed in all the static cultures (Fig. 1A). The circadian

12

50

A 10 te-

4O

A

~D O.

30

A

=

O3 ¢~

m

AL

v/

20

L..

> ,~

lO

12

12

0

0 i 0

12

i 12

12

0

12

10

40

E O'1 vO.

30

\

8

O3 t~

O3

6

20 t..

t.=

13. >

n

>

<

0

D

B

m

12

12

50

U~ O~ O.

i

3

Time of Day (hr)

Time of Day (hr)

A

i

12

<

10

4 i

n

0 0

12

0

i

'

12

12

T i m e of Day (hr)

i

2

0

12

12

O

12

12

12

Time of Day (hr)

Fig. 2. Effect of anisomycin on the circadian oscillation of AVP release. A,B: Cells were cultured on a cloning plate and oscillation of AVP release was measured (open circle). After 2 days, the medium was supplemented with anisomycin (50/~M) on various 6-h intervals (arrows show the midpoint of the pulse), and the measurements were continued (closed circle). C,D: Circadian oscillation of AVP release in peffusion culture was measured. At various times (arrow), 50 /~1 of 1 mM anisomycin was injected into each well.

182

K. Watanabe et al. / Brain Research 684 (1995) 179-184

0 J= I¢,.)

80

A

J= 0Q

j/

40

-12 e

r. O.

\

i

-24 0

30

r-

10

Fig. 4. Phase-response curve generated by 6-h pulses of anisomycin (50 /zM). Each symbol represents the result of individual static culture from seven cell preparations. The circadian time of the onset of the pulse is denoted by the symbols.

into the perfusion cultures (Fig. 2C, D). Since the second AVP peak following the pulse fell at almost the same time

/

70

10

m

50

30

as the first peak, the difference between the transient and

permanent phase shift is considered to be negligibly small. No morphological change could be observed after the

anisomycin pulse at this dose and duration. To determine whether the large phase-shifts were advances or delays, cultures were exposed to lower concentrations of anisomycin at the same CT (Fig. 3A). When the onset of the pulse fell between CT 20 and CT 3, increasing

i

10

24

Circadian Time (hr)

/ A

50

12

amounts of anisomycin gave increasing phase shifts to the

50

delay direction (Fig. 3B). The results o f the pulse beginning about CT 4 or 5 were complicated. Usually the shifts were considered to be very large phase-delays (sometimes

50 30

more than 20 h), since a lower concentration (2 to 1 0 / z M ) of the anisomycin pulse at the same CT or a shorter duration (2 or 4 h) of the pulse beginning at the same CT induced a smaller phase-delay in the parallel culture. In

10

12

0

0

12

12

one case, however, the phase-delay looked very small (less

than 0.2 h). Because we did not examine the effect of a lower concentration of anisomycin in the parallel culture, it is not certain whether this phase-shift is indeed a small

Time of Day (hr)

B

delay or whether it is either a large advance or an extraordinarily large delay of about 24 h. W e conclude that a

singular point exists around CT 4 or 5. From these results, /-

" ~

(d)

I'O

=

m c-

•

(a)

-12 o w

0 0

O.

t -24

I

I

I

I

I

10

20

30

40

5O

Anlsomycln (I~M)

(b) (c) 60

Fig. 3. Dose-dependent phase shift induced by 6-h anisomycin pulse in static culture. A: Cells were cultured on a cloning plate and circadian oscillation of AVP release was measured for 2 cycles (open circle). Just after the last sampling, the cultures were exposed to various concentration (0 to 50 tzM) of 6-h anisomycin pulse (10:00-16:00). The oscillation was measured for one cycle just after the pulse (data not shown) and for 2 cycles 3 days later (closed circle). (B) Dose-response curve for 6-h anisomycin pulse beginning at CT 22.8-23.4 (a), CT 23.5-0.5 (b), CT 1.7-3.2 (c) and CT 9.5-10.5 (d). Each symbol represents the result of individual static culture from five cell preparations.

K. Watanabe et al. /Brain Research 684 (1995) 179-184 40

e.

30

to O) Q. 20 0 m @) D.

> <

10

0 0

12

0

12

12

Time of Day (hr) Fig. 5. Effect of anti-sense AVP on circadian oscillation of AVP release. Cells were cultured on a cloning plate and the oscillation of AVP release was measured. The medium was supplemented with 20 ~ M (final conc.) of the control (open circle) or anti,sense AVP (closed circle) oligonucleotide for the whole period depicted in the figure.

we obtained the phase-response curve (PRC) for anisomycin (Fig. 4). These results suggest that the pacemaker requires some protein(s) in a phase-dependent way. One of the candidate proteins is AVP, since anisomycin induced both reduction of AVP release and phase-shift. To study whether the phase-shift is caused by the reduction of AVP synthesis, we tried to reduce the synthesis of AVP in a different way. Fig. 5 shows the effect of the oligonucleotide, antisense AVP, on the oscillation of AVP release. AVP release was apparently reduced within several hours after the administration of antisense AVP, and kept at low level during the treatment. However, the time-of-peak was almost the same as that of the control culture. In another experiment, there was no difference in phase and amplitude between the control and the antisense-treated culture after administration for 2 days (data not shown). These results suggest that anisomycin-induced phase-shift is not attributable to the reduction of AVP synthesis.

4. Discussion

As we have previously reported, the SCN cells in primary culture showed clear circadian oscillation of AVP release. Since the circadian phase of AVP release is almost the same among replicated cultures, this system seems to be suitable for measuring phase-shifts of mammalian pacemakers. The present results clearly showed that anisomycin, an inhibitor of protein synthesis on 80 S ribosomes, induced phase-shifts in the mammalian circadian pacemaker. The magnitude of the phase-shift was phaseand dose-dependent. While a 6-h anisomycin pulse induced more than 15 h of phase-delay at a peak of AVP

183

release, the pulse given at 7 h after the peak induced only few hours of phase-delay. Anisomycin induced smaller phase-delays at lower dosages. Although there are not many points in the PRC, the shape shows Type 0 resetting, and it resembles the previously reported PRCs for protein synthesis inhibitors in other organisms rather than those in rodents. There exist the following problems in CT determination. (1) The time-of-peak on the second day in culture from which CT has been determined has some variance (S.D. = 1.81 h, 15 cultures from 5 independent cell preparations) [18]. (2) The peak time 15 h after onset of darkness in L:D = 14:10 could be designated as CT 5 instead of CT 3. (3) Instead of the onset of the anisomycin pulse, the midpoint or the end of the pulse could be chosen as stimulation time. However, these problems with respect to the determination of CT are common to all the static cultures. Although the PRC might have to be shifted along the x-axis for a couple of hours as a consequence of this, it will be still be true that anisomycin induces large delays at an early subjective day. In the present study, anisomycin did not induce phase advances at any circadian phase. The magnitudes of the maximum phase-advance and the maximum phase-delay induced by the inhibitors varies among organisms. For example, in Neurospora, the maximum phase-advance and delay induced by cycloheximide are more than 12 h and less than 3 h, respectively [13], whereas in avian pineal culture, the maximum phase-advance and delay induced by anisomycin are about 6 h and 12 h, respectively [15]. Furthermore, anisomycin induces very small phase-advances in Aplysia [9]. Recently, it has been reported that the circadian pacemaker of Bulla was stopped by the inhibitors [8]. This means that the inhibitors induce phasedelays but not phase-advances at any CI'. Although there are some differences, these PRCs resemble each other in shape. The singular point, which should be the most effective phase, is the early subjective day in all of them. It is considered that the response to protein synthesis inhibitors is essentially the same among these organisms including rat SCN cells in primary culture. Protein synthesis may be involved in the generation of circadian rhythms in mammals as well as in other organisms. It has not yet been clarified whether the circadian oscillation of AVP release from the SCN cells is regulated at the level of transcription, translation or secretion. At least daily synthesis of the peptide must be necessary for the circadian oscillation of AVP release, since addition of anisomycin or antisense AVP reduced AVP release within several hours. The antisense reduced the amplitude of the oscillation without changing its circadian phase. These results suggest that the phase-shift induced by anisomycin is not due to the reduction of AVP synthesis. Therefore, the AVP release (synthesis) process in the SCN may be located on output pathway but not on 'causal loop' of the rhythm generation.

184

K. Watanabe et al. / Brain Research 684 (1995) 179-184

Acknowledgements We thank to Dr. J.H. Meijer for her critical reading of the manuscript. This work was supported in part by a Grant-in-Aid for Scientific Research (No. 04670099) from the Ministry of Education, Science and Culture of Japan.

References [1] Dunlap, J.C. and Feldman, J.F., On the role of protein synthesis in the circadian clock of Neurospora crassa, Proc. Natl. Acad. Sci. USA, 85 (1988) 1096-1100. [2] Edmunds Jr., L.N., Cellular and Molecular Basis of Biological Clocks, Springer, New York, 1988. [3] Eskin, A. and Corrent, G., Effects of divalent cations and metabolic poisons on the circadian rhythm from the Aplysia eye, J. Comp. Physiol., 117 (1977) 1-20. [4] Feldman, J.F., Lengthening the period of a biological clock in Euglena by cycloheximide, an inhibitor of protein synthesis, Proc. Natl. Acad. Sci. USA, 57 (1967) 1080. [5] Inouye, S.-I.T., Takahashi, J.S., Wollnik, F. and Turek, F., Inhibition of protein synthesis phase shifts a circadian pacemaker in mammalian SCN, Am. J. Physiol., 255 (1988) R1055-R1058. [6] Jacklet, J.W., Neuronal circadian rhythm: phase shifting by a protein synthesis inhibitor, Science, 198 (1977) 69-71. [7] Johnson, C.H., An Atlas of Phase Response Curves for Circadian and Circatidal Rhythms, Dept. of Biol., Vanderbilt Univ., 1990. [8] Khalsa, S.B.S., Whitmore, D. and Block, G.D., Stopping the circadian pacemaker with inhibitors of protein synthesis, Proc. Natl. Acad. Sci. USA, 89 (1992) 10862-10866.

[9] Lotshaw, D.P. and Jacklet, J.W., Involvement of protein synthesis in circadian clock of Aplysia eye, Am. J. Physiol., 250 (1986) R5-R7. [10] Meijer, J.H. and Rietveld, W.J., Neurophysiology of the suprachiasmatic circadian pacemaker in rodents, Physiol. Rev., 69 (1989) 671-707. [11] Mrosovsky, N. and Ralph, M.R., Phase response curve to anisomycin in tau mutant hamsters, Experientia, 48 (1992) 875-877. [12] Murakami, N., Takamure, M., Takahashi, K., Utunomiya, K., Kuroda, H. and Etoh, T., Long-term cultured neurons from rat suprachiasmatic nucleus retain the capacity for circadian oscillation of vasopressin release, Brain Res., 545 (1991) 347-350. [13] Nakashima, H., Perlman, J. and Feldman, J.F., Cycloheximide-induced phase shifting of circadian clock of Neurospora, Am. J. Physiol., 241 (1981) R31-R35. [14] Raju, U., Koumenis, C., Nunez-regueiro, M. and Eskin, A., Alteration of the phase and period of a circadian oscillator by a reversible transcription inhibitor, Science, 253 (1991) 673-675. [15] Takahashi, J.S., Murakami, N., Nikaido, S.S., Pratt, B.L. and Robertson, L.M., The avian pineal, a vertebrate model system of the circadian oscillator: cellular regulation of circadian rhythms by light, second messengers, and macromolecular synthesis, Rec. Prog. Horm. Res., 45 (1989) 279-352. [16] Takahashi, J.S. and Turek, F.W., Anisomycin, an inhibitor of protein synthesis, perturbs the phase of a mammalian circadian pacemaker, Brain Res., 405 (1987) 199-203. [17] Watanabe, K., Koibuchi, N., Ohtake, H. and Yamaoka, S., Circadian rhythm of vasopressin release in primary cultures of rat suprachiasmatic nucleus, Brain Res., 624 (1993) 115-120. [18] Watanabe, K. and Yamaoka, S., Cell cultures of rat suprachiasmatic nucleus: Circadian oscillation of vasopressin release, BioL Rhythm Res., in press. [19] Wollnik, F., Turek, F.W., Majewski, P. and Takahashi, J.S., Phase shifting the circadian clock with cycloheximide: response of hamsters with an intact or split rhythm of locomotor activity, Brain Res., 496 (1989) 82-88.