Vous êtes sur la page 1sur 5
YBBRC 26901 No. of Pages 5, Model 5G 16 June 2011
YBBRC 26901
No. of Pages 5, Model 5G
16 June 2011

1

Biophysical Research Communications xxx (2011) xxx–xxx 1 Contents lists available at ScienceDirect Biochemical

Contents lists available at ScienceDirect

Biochemical and Biophysical Research Communications

journal homepage: www.else vier.com/locate/ybbrc

journal homepage: www.else vier.com/locate/ybbrc 2 Caffeine lengthens circadian rhythms in mice Hideaki Oike

2

Caffeine lengthens circadian rhythms in mice

Hideaki Oike a , ⇑ , Masuko Kobori a , Takahiro Suzuki b , c
Hideaki Oike a , ⇑ , Masuko Kobori a , Takahiro Suzuki b , c , Norio Ishida b , c
3
4
a Food Functionality Division, National Food Research Institute (NFRI), National Agriculture and Food Research Organization (NARO), 2-1-12 Kannon dai, Tsukuba,
5
Ibaraki 305-8642, Japan
6
b Ishida Group of Clock Gene, Biomedical Research Institute, National Institute of Advanced Industrial Science and Technology (AIST), Central 6, 1-1 -1 Higashi, Tsukuba,
7
Ibaraki 305-8566, Japan
8
c Graduate School of Life and Environmental Sciences, University of Tsukuba, Tsukuba, Ibaraki 305-8576, Japan
9
10
article info
abstract
2 1 4 2
13
Article history:
25
14
Received 3 June 2011
26
15
Available online xxxx
27
28
16
Keywords:
29
17
Circadian clocks
30
18
Caffeine
31
19
Coffee
32
20
Mouse
33
21
SCN
22
Although caffeine alters sleep in many animals, whether or not it affects mammalian circadian clocks
remains unknown. Here, we found that incubating cultured mammalian cell lines, human osteosarcoma
U2OS cells and mouse fibroblast NIH3T3 cells, with caffeine lengthened the period of circadian rhythms.
Adding caffeine to ex vivo cultures also lengthened the circadian period in mouse liver explants from
Per2::Luciferase reporter gene knockin mice, and caused a phase delay in brain slices containing the sup-
rachiasmatic nucleus (SCN), where the central circadian clock in mammals is located. Furthermore,
chronic caffeine consumption ad libitum for a week delayed the phase of the mouse liver clock in vivo
under 12 h light–dark conditions and lengthened the period of circadian locomotor rhythms in mice
under constant darkness. Our results showed that caffeine alters circadian clocks in mammalian cells
in vitro and in the mouse ex vivo and in vivo .
34
Liver
23
2011 Published by Elsevier Inc.
35
36
37
38
1. Introduction
obscure despite being frequently and chronically consumed by
humans.
60
61
39
Coffee, tea, soft drinks and energy drinks are popular sources of
40
caffeine worldwide. Caffeine is a psychoactive component that can
2. Materials and methods
62
41
temporarily ward off drowsiness, ease fatigue and reduce anxiety.
42
Caffeine reduces sleep duration in Drosophila as well as in mam-
2.1.
Cell-based circadian assays
63
43
mals [1,2] and together with its analogue, theophylline, lengthens
44
circadian rhythms in Neurospora , Chlamydomonas , Drosophila, and
64
45
in higher plants [3–6] . Despite these findings, little evidence indi-
65
46
cates that caffeine actually affects circadian rhythms in mammals.
66
47
The mammalian circadian clock system comprises a central
67
48
pacemaker in the suprachiasmatic nucleus (SCN) of the hypothala-
68
49
mus and various oscillators in most peripheral tissues. The molec-
69
50
ular oscillator of the circadian clocks is thought to depend on a
70
51
negative transcriptional feedback loop involving clock genes such
71
52
as Per , Cry , Clock and Bmal [7] . Light–dark cycles dominate the
72
53
entral clock as environmental input, while nutrient intake is a pre-
73
54
dominant cue for many peripheral tissue clocks [8] .
74
55
Caffeine delays the phase of circadian activity through ryano-
75
56
dine receptors in mouse brain slices containing the SCN [9] and
76
57
suppresses melatonin levels in humans when consumed at night
The U2OS- hPer2-Luc [11] cells were a gift from Dr. Hiroki R Ueda
(RIKEN, JAPAN) and NIH 3T3- mBmal1-Luc cells stably expressing the
pGL4.22 vector (Promega) harboring the mBmal1-promotor ( 611
to +154) were established from NIH3T3-3-4 cells (RCB1862; CELL
BANK, RIKEN). The cell lines were incubated at 37 C under a 5%
CO 2 atmosphere in DMEM containing high glucose (Invitrogen)
supplemented with 10% fetal calf serum (MP Biomedicals) and pen-
icillin–streptomycin (Sigma). To record luminescence, the cells were
stimulated with 50% horse serum for 2 h to synchronize rhythms
[12] , and then incubated in 35-mm dishes in medium supplemented
with 0.1 mM luciferin (Promega) and caffeine (Nacalai Tesque). Bio-
luminescence was measured and integrated for 1 min at intervals of
10 min using a dish-type luminometer supplied with 5% CO 2 at 37 C
(Kronos Dio AB-2500; ATTO).
77
58
[10] . These reports suggest that caffeine alters physiological
59
rhythms, but whether it affects circadian clocks in mammals is

Corresponding author. Fax: +81 29 838 8041. E-mail address: oike@affrc.go.jp (H. Oike).

0006-291X/$ - see front matter 2011 Published by Elsevier Inc. doi: 10.1016/j.bbrc.2011.06.049

2.2. Animals and handling

Animals were handled according to the guidelines of the Minis- try of Agriculture, Forestry and Fisheries for laboratory animal studies and the study was reviewed and approved by the Animal

78

79

80

81

YBBRC 26901 No. of Pages 5, Model 5G 16 June 2011
YBBRC 26901
No. of Pages 5, Model 5G
16 June 2011

2

H. Oike et al. / Biochemical and Biophysical Research Communications xxx (2011) xxx–xxx

82 Care and Use Committee of the National Food Research Institute,

with a vibratome in cold HBSS. Regions of the brain identified un-

108

83 National Agriculture and Food Research Organization (NARO),

der a dissecting microscope were isolated using scalpels as

109

84 Japan.

1.5 mm triangles of tissue and placed on Millicell membranes

110

85 Young adult male BALB/cAn mice (20–30 g) were obtained from

(Millipore) with 1.3 ml of DMEM (Nissui) containing 4.5 g/l glu-

111

86 the Institute for Animal Reproduction (Charles River Japan). Per2::-

cose, 0.35 g/l NaHCO 3 , 1 B27 supplement (Invitrogen), 10 mM

112

87 Luc knockin mice [13] established by Dr. Joseph Takahashi (North-

HEPES, penicillin–streptomycin and 0.1 mM luciferin. The explants

113

88 western University) were supplied by the Jackson Laboratory and

were cultured at 37 C without CO 2 and bioluminescence was

114

89 bred as homozygotes. All mice were housed at 24 ± 1 C and

monitored using the Kronos luminometer. 115

90 55 ± 5% humidity under a 12 h light–dark photocycle (light period

91 from 08:00 to 20:00) with free access to water and to a standard

92 diet (NMF, Oriental Yeast).

2.4. Luminescence data analysis

93 2.3. Preparation and measurement of bioluminescence of explants 94 from Per2::Luc mice 95 Liver
93 2.3. Preparation and measurement of bioluminescence of explants
94 from Per2::Luc mice
95 Liver explants were prepared as described [14] . The Per2::Luc
96 mice were sacrificed at ZT3 to record bioluminescence rhythmicity
97 in the liver. Liver blocks were rapidly removed from the mice and
98 placed in cold Hanks’ balanced salt solution (HBSS; Sigma) with
Luminescence data were analyzed as described [14,15]. Firstly,
the original data were smoothed by an adjusting-averaging meth-
od with 100-min running means. The data set was detrended by
subtracting the 24-h running average from the raw data. The peaks
were defined as points at which the bioluminescence was higher
than that on both sides of the points and confirmed from wave-
forms. The peak phase time was determined from the peak that ap-
peared between 12 and 36 h of culture (peak 1). Period length was
calculated as elapsed time between peak 1 and the next peak (peak
99 10 mM HEPES (Sigma) and penicillin–streptomycin. The blocks
2).
100 were cut into pieces and explanted in 35-mm dishes, sealed with
101 Parafilm and incubated at 37 C under a 5% CO 2 atmosphere with
102 DMEM containing high glucose, 10 mM HEPES, penicillin–strepto-
2.5.
Behavioral analyses
103 mycin and 0.1 mM luciferin. Bioluminescence was monitored for
104 1 min at 10-min intervals using a Kronos luminometer.
105 Brain slices containing the SCN were prepared as described [15] .
106 The brains of Per2::Luc mice sacrificed at ZT6 were rapidly removed
107 and placed in cold HBSS. Coronal sections (400 l m thick) were cut
Behavior was analyzed as described [16] . BALB mice were
placed in standard mouse cages equipped with infrared sensors
(AS10D; Melquest) and locomotor activity data were collected
and analyzed using CIF-II software (Melquest). The mice were
A
B
C
U2OS-hPer2-Luc
500000
34.0
4000000
0
mM
1 mM
**
2
mM
5 mM
32.0
250000
30.0
0
2000000
**
28.0
-250000
26.0
0
-500000
24.0
12:00 0
12:00 24
12:00 48
12:00 72
12:00 96
12:00 120
12:00 0
12:00 24
12:00 48
12:00 72
12:00 96
12:00 120
0125
Time (h)
Time (h)
caffeine (mM)
D
NIH3T3-mBmal1-Luc
E
F
150000
28.0
600000
0
mM
1 mM
*
2
mM
5 mM
27.5
50000
400000
27.0
26.5
-50000
200000
26.0
0
-150000
25.5
12:00 0
12:00 24
12:00 48
12:00 72
12:00 96
12:00 0
12:00 24
12:00 48
12:00 72
12:00 96
0125
Time (h)
Time (h)
caffeine (mM)
Bioluminescence (cpm)
Bioluminescence (cpm)
Deviation from moving average
Deviation from moving average
Period length (h)
Period length (h)

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

Fig. 1. Caffeine lengthens period of circadian rhythms in mammalian cell lines Circadian luminescent rhythms of the reporter driven by the clock gene promot er monitored in human osteosarcoma U2OS/ hPer2-Luc cells (A–C) and mouse NIH3T3/ mBmal1-Luc fibroblasts (D–F) incubated for 2 h with 50% horse serum (time 0) followed by 0–5 mM caffeine added to culture medium. Representative luminescence profiles are shown for each cell line as raw (A and D) and detrended (B and E) data. Period lengths were determined by 4 independent assays in duplicate (C and F). Data are shown as means ± SE. p < 0.05, p < 0.01 (Dunnett’s test).

YBBRC 26901 No. of Pages 5, Model 5G 16 June 2011
YBBRC 26901
No. of Pages 5, Model 5G
16 June 2011

H. Oike et al. / Biochemical and Biophysical Research Communications xxx (2011) xxx–xxx

3

A B C D E F 0 24 48 0 24 48 24 48 0
A
B
C
D
E
F
0
24
48
0
24
48
24
48
0
G
H
I
14
16
18
14
16
18
14
16
18
3.2. Caffeine alters circadian rhythm in the mouse liver
C; p < 0.05, n = 11)

Fig. 2. Caffeine affects circadian rhythm in mouse liver. Representative circadian profiles of reporter luminescence in liver explants isolated from mPer2::Luc knockin mice (A, D, E, and F). Circadian oscillation peaked between 12 and 36 h in culture with or without 1 mM caffeine (B). Periods of circadian rhythms in liver explant s were determined as duration between peaks shown in (B) and next peak (C). Effects of consuming caffeine (D and G), coffee (E and H) or decaffeinated coffee (F and I) in vivo were examined in liver from knockin mice. Data are shown as means ± SE. p < 0.05, p < 0.01 ( t -test).

132 maintained under constant darkness for 1 week, then supplied

caffeine in vitro to identify the effects of caffeine on circadian rhythms in mammalian cells. Caffeine significantly and dose- dependently (1–5 mM) lengthened the circadian period of human osteosarcoma U2OS-hPer2-Luc cells ( Fig. 1 A–C; p < 0.01 for linear trend) and of mouse NIH3T3 -mPer2-Luc fibroblasts ( Fig. 1 D–F; p < 0.05 for linear trend), although higher concentrations attenu- ated the amplitude of reporter gene expression.

133 with caffeine (0.04% or 0.08%), coffee (1 g of instant coffee/100 ml

134 water), decaffeinated coffee (1 g/100 ml water) or water for 1 week

135 under constant darkness to measure the free-running period.

136 Cages were equipped with running wheels (SW-15S; Melquest).

137 The mice were provided with coffee or water ad libitum under a 12-

138 h light–dark cycle for 2 weeks, and under constant darkness for the

139 next 2 weeks. The free-running period was calculated using Metlab

140 software (MathWorks Inc.).

141 2.6. Statistical analysis

We examined the effects of caffeine on circadian rhythms in li- ver explants isolated from Per2::Luciferase ( Per2::Luc ) knockin mice [13] . Luminescence robustly oscillated for several days, and adding 1 mM caffeine to the culture medium lengthened the circadian rhythm of reporter gene expression in liver explants ( Fig. 2A and

142 All values are expressed as means ± SEM. Differences in the per-

143 iod length and peak time were statistically evaluated using Stu-

144 dent’s t -test for single comparisons, and one-way ANOVA with

145 Dunnett’s post hoc test for multiple comparisons. Caffeine dose-

as well as the cultured cell lines. On the other

146 dependence was evaluated by one-way ANOVA, followed by a

hand, caffeine did not influence the time of the rhythm peak in the liver explants ( Fig. 2 B). These results suggest that caffeine di- rectly lengthens the period of circadian rhythm in the mouse liver. Moreover, we administered caffeine as drinking water (0.04% solution; 1.9 mM) ad libitum to Per2::Luc mice for 1 week under a 12 h light–dark cycle to evaluate the effect on circadian rhythms in the liver in vivo. The caffeine concentration of was similar to that found in instant coffee, and in a study of chronic caffeine consump- tion using mice [18] . The phase of the luminescent rhythm statis- tically differed in the liver isolated from these mice compared with

147 post hoc test for linear trends using Prism 5 software (GraphPad

148 Software).

149

3. Results

150 3.1. Caffeine lengthens circadian rhythm in mammalian cells

151 We cultured cell lines expressing luciferase reporters driven by

152 the clock gene promoter ( hPer2-Luc [11] , mBmal1-Luc [17] ), with

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

YBBRC 26901 No. of Pages 5, Model 5G 16 June 2011
YBBRC 26901
No. of Pages 5, Model 5G
16 June 2011

4

H. Oike et al. / Biochemical and Biophysical Research Communications xxx (2011) xxx–xxx

A

B

A B

A B

C D 30 31 32 33 4. Discussion
C
D
30
31
32
33
4. Discussion

Fig. 3. Caffeine affects circadian rhythm in mouse SCN. Representative bioluminescence data show circadian profiles of reporter gene in brain slices isolat ed from mPer2::Luc knockin mice (A, raw data; B; detrended data) and incubated in culture medium with or without caffeine (1 mM). Peak time (C) and period length (D) of circ adian rhythms in slices were determined as described in the legend to Fig. 2 (means ± SE; p < 0.05, t -test).

177 control mice that consumed water ( Fig. 2 D and G; p < 0.01, n = 7).

178 Chronic caffeine intake delayed the phase of reporter gene expres-

179 sion by about 1 h. Furthermore, appropriate concentration of com-

We found that caffeine lengthens circadian rhythms in cultured cell lines as well as in mouse liver explants. These results suggest that caffeine affects mammalian clocks at the cellular level. Although caffeine did not lengthen the period in cultured brain slices containing SCN, it caused a phase delay in SCN. These results are consistent with the findings of a previous electrophysiological study [9] , and suggest that the effects of caffeine are tissue depen- dent. Because the SCN is the central clock that controls behavioral rhythm, the chronic consumption of caffeine or caffeinated drinks lengthened circadian locomotor rhythms in mice. It should be noted that amount of liquid consumption was similar level among any drink. The chronic phase-delay in the SCN caused by caffeine consumption might show the phenotype of period prolongation in locomotor rhythm. Our findings of a lengthened period in hu- man cells, mouse tissues and mouse behavior suggests that caf- feine affects an intracellular clock, rather than an intercellular clock system in the brain or tissues. Endogenous circadian clocks have been identified in various organisms from bacteria to humans. Despite common functional- ity, clock components and biochemical processes are species-spe- cific [19] . However, previous findings showing that caffeine lengthens circadian rhythms from bacteria to mammals, suggest that the target of caffeine is a universal mechanism that deter- mines 24-h clocks in all life-forms. Being structurally similar to adenosine, caffeine inhibits various adenosine-related enzymes such as various kinases, cAMP phosphodiesterase and adenosine receptors [18] . Reactions using adenosine triphosphate (ATP) might be key; indeed, the ATPase activity of KaiC protein deter- mines the basic timing for the cyanobacterial circadian clock [20]. Phosphorylation steps could also be targets for caffeine be- cause ATP is a phosphate donor. Interestingly, casein kinases and glycogen synthase kinases (GSK) modify period length in Drosoph- ila and in mammals [11,21–23]. Humans have consumed caffeine since the Stone Age, and caffe- inated beverages are popular all worldwide today. However, the effects of caffeine on circadian rhythms have not been considered

180 mercially available coffee (1 g of instant coffee in 100 ml water)

181 similarly delayed the phase in the liver ( Fig. 2 E and H; p < 0.05,

182 n = 7), whereas decaffeinated coffee did not ( Fig. 2 F and I; n = 7).

183 These results suggest that chronic consumption of caffeine or caf-

184 feinated coffee delays the liver clock in vivo .

185 3.3. Caffeine treatment delays circadian rhythms in the SCN

186 We added caffeine to cultured brain slices containing suprach-

187 iasmatic nuclei (SCN) isolated from Per2::Luc mice to determine

188 the effects of caffeine on the central clock. Reporter gene expres-

189 sion robustly oscillated in the SCN, and adding 1 mM caffeine to

190 the culture medium delayed the peak time of the rhythm

191 ( Fig. 3 A–C; p < 0.05, n = 10). In contrast, the period length of the

192 circadian rhythm was not significantly altered by caffeine ( Fig. 3 D).

193 3.4. Caffeine intake lengthens circadian behavior in mice

194 To determine the effects of chronic caffeine consumption on cir-

195 cadian behavioral rhythm, we provided mice with 0.04% caffeine

196 ad libitum instead of water under constant darkness. As the result,

197 the circadian period was longer in the mice given caffeine than in

198 control mice given access to water ad libitum ( Fig. 4 A and B;

199 p < 0.05, n = 12). The circadian period was further lengthened by

200 consuming 0.08% caffeine ( p < 0.01, n = 7). Chronic coffee con-

201 sumption also lengthened the circadian period ( p < 0.01, n = 11),

202 whereas decaffeinated coffee did not ( n = 8). It should also be noted

203 that chronic coffee consumption under a light–dark cycle for

204 2 weeks did not alter the circadian locomotor rhythm because light

205 cues dominated the rhythm. The ability of coffee to lengthen the

206 period manifested only when light cues were removed ( Fig. 4 C

207 and D; p < 0.05, n = 7). These results showed that chronic caffeine

208 consumption lengthens the circadian locomotor rhythm in mice.

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

YBBRC 26901 No. of Pages 5, Model 5G 16 June 2011
YBBRC 26901
No. of Pages 5, Model 5G
16 June 2011

H. Oike et al. / Biochemical and Biophysical Research Communications xxx (2011) xxx–xxx

5

Biophysical Research Communications xxx (2011) xxx–xxx 5 References [1] J.C. Hendricks, S.M. Finn, K.A. Panckeri, J.

References

[1] J.C. Hendricks, S.M. Finn, K.A. Panckeri, J. Chavkin, J.A. Williams, A. Sehgal, A.I. Pack, Rest in Drosophila is a sleep-like state, Neuron 25 (2000) 129–138. [2] P.J. Shaw, C. Cirelli, R.J. Greenspan, G. Tononi, Correlates of sleep and waking in Drosophila melanogaster , Science 287 (2000) 1834–1837. [3] M.N. Wu, K. Ho, A. Crocker, Z. Yue, K. Koh, A. Sehgal, The effects of caffeine on sleep in Drosophila require PKA activity but not the adenosine receptor, J. Neurosci. 29 (2009) 11029–11037. [4] J.F. Feldman, Circadian periodicity a neurospora: alteration by inhibitors of cyclic AMP phosphodiesterase, Science 190 (1975) 789–790. [5] J.E. Goodenough, V.G. Bruce, The effects of caffeine and theophylline on the phototactic rhythm of Chlamydomonas-reinhardii, Biol. Bull. 159 (1980) 649–

655.

[6] I. Bollig, K. Mayer, W.E. Mayer, W. Engelmann, Effects of camp theophylline

imidazole and 4-(34-dimethoxybenzyl)-2-imidazolidone on leaf movement rhythm of trifolium-repens – test of camp-hypothesis of circadian-rhythms, Planta 141 (1978) 225–230.

418 (2002) 935–941. 728–742. 394 (1998) 381–384. is 350–356. 3477–3488. of 20746–20751. A (2002)
418 (2002) 935–941.
728–742.
394 (1998) 381–384.
is
350–356.
3477–3488.
of
20746–20751.
A
(2002) 816–820.

[7] S. Reppert, D. Weaver, Coordination of circadian timing in mammals, Nature

[8] C.B. Green, J.S. Takahashi, J. Bass, The meter of metabolism, Cell 134 (2008)

[9] J.M. Ding, G.F. Buchanan, S.A. Tischkau, D. Chen, L. Kuriashkina, L.E. Faiman,

J.M. Alster, P.S. McPherson, K.P. Campbell, M.U. Gillette, A neuronal ryanodine receptor mediates light-induced phase delays of the circadian clock, Nature

[10] K.P. Wright Jr., P. Badia, B.L. Myers, S.C. Plenzler, M. Hakel, Caffeine and light

effects on nighttime melatonin and temperature levels in sleep-deprived

humans, Brain Res. 747 (1997) 78–84. [11] Y. Isojima, M. Nakajima, H. Ukai, H. Fujishima, R.G. Yamada, K.H. Masumoto, R. Kiuchi, M. Ishida, M. Ukai-Tadenuma, Y. Minami, R. Kito, K. Nakao, W. Kishimoto, S.H. Yoo, K. Shimomura, T. Takao, A. Takano, T. Kojima, K. Nagai, Y. Sakaki, J.S. Takahashi, H.R. Ueda, CKIepsilon/delta-dependent phosphorylation

a temperature-insensitive, period-determining process in the mammalian

circadian clock, Proc. Natl. Acad. Sci. USA 106 (2009) 15744–15749. [12] A. Balsalobre, F. Damiola, U. Schibler, A serum shock induces circadian gene expression in mammalian tissue culture cells, Cell 93 (1998) 929–937.

[13] S. Yoo, S. Yamazaki, P. Lowrey, K. Shimomura, C. Ko, E. Buhr, S. Siepka, H. Hong, W. Oh, O. Yoo, M. Menaker, J. Takahashi, Period2::luciferase real-time reporting of circadian dynamics reveals persistent circadian oscillations in mouse peripheral tissues, Proc. Natl. Acad. Sci. USA 101 (2004) 5339–5346. [14] A. Hirao, Y. Tahara, I. Kimura, S. Shibata, A balanced diet is necessary for proper entrainment signals of the mouse liver clock, PLoS One 4 (2009) e6909. [15] M. Abe, E.D. Herzog, S. Yamazaki, M. Straume, H. Tei, Y. Sakaki, M. Menaker, G.D. Block, Circadian rhythms in isolated brain regions, J. Neurosci. 22 (2002)

Fig. 4. Caffeine lengthens period of circadian locomotor rhythm in mice. Repre- sentative locomotor activity (A) from mice after consuming water, coffee or caffeine for 7 days ad libitum (longitudinal lines) under constant darkness (DD). Horizontal gray and black bars represent 12-h light–dark cycle before exposure to constant darkness. Periods for mice after consuming water, decaffeinated coffee, coffee, 0.04% or 0.08% caffeine are shown (B; means ± SE, p < 0.05, p < 0.01, Dunnett’s test). Panels (C) show representative actograms from mice after consuming water or coffee for 4 weeks under 12-h light–dark (LD) cycles for 2 weeks and then constant darkness for 2 weeks. Period (D) is for mice shown in (C); (means ± SE, p < 0.05, t -test).

[16] H. Oike, K. Nagai, T. Fukushima, N. Ishida, M. Kobori, High-salt diet advances molecular circadian rhythms in mouse peripheral tissues, Biochem. Biophys. Res. Commun. 402 (2010) 7–13. [17] Y. Onishi, S. Hanai, T. Ohno, Y. Hara, N. Ishida, Rhythmic SAF-A binding underlies circadian transcription of the Bmal1 gene, Mol. Cell Biol. 28 (2008)

[18] B.B. Fredholm, K. Battig, J. Holmen, A. Nehlig, E.E. Zvartau, Actions of caffeine in the brain with special reference to factors that contribute to its widespread use, Pharmacol. Rev. 51 (1999) 83–133. [19] D. Bell-Pedersen, V.M. Cassone, D.J. Earnest, S.S. Golden, P.E. Hardin, T.L. Thomas, M.J. Zoran, Circadian rhythms from multiple oscillators: lessons from diverse organisms, Nat. Rev. Genet. 6 (2005) 544–556. [20] K. Terauchi, Y. Kitayama, T. Nishiwaki, K. Miwa, Y. Murayama, T. Oyama, T. Kondo, ATPase activity of KaiC determines the basic timing for circadian clock

246 in detail. Our findings offer new insight into the action of caffeine

247 upon the circadian clocks at the cell and organism levels.

248 Acknowledgments

249 This work was supported by KAKENHI [21700778; Grant-in-Aid

cyanobacteria, Proc. Natl. Acad. Sci. USA 104 (2007) 16377–16381.

[21] T. Hirota, W.G. Lewis, A.C. Liu, J.W. Lee, P.G. Schultz, S.A. Kay, A chemical biology approach reveals period shortening of the mammalian circadian clock by specific inhibition of GSK-3beta, Proc. Natl. Acad. Sci. USA 105 (2008)

250 for Young Scientists (B)] to HO and the grant from the Nestlé Nutri-

251 tion Council Japan to HO and MK, and the grant from the Mishima

[22] Y. Tsuchiya, M. Akashi, M. Matsuda, K. Goto, Y. Miyata, K. Node, E. Nishida,

Involvement of the protein kinase CK2 in the regulation of mammalian circadian rhythms, Sci. Signal 2 (2009) ra26.

252 Kaiun Memorial Foundation to HO. All authors report no biomedi-

253 cal financial interests or potential conflicts of interest.

[23] J.M. Lin, V.L. Kilman, K. Keegan, B. Paddock, M. Emery-Le, M. Rosbash, R. Allada,

254 We thank Dr Hiroki R Ueda for donating the U2OS- hPer2-Luc

role for casein kinase 2alpha in the Drosophila circadian clock, Nature 420

255 cell line. We thank Professor Shigenobu Shibata and Ms Akiko

256 Hirao (Waseda University), and Dr Hajime Tei, Dr Katsuhiko

257 Sakamoto and Ms Megumi Kato (Mitsubishi Kagaku Institute of

258 Life Sciences) for technical advice.

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329