Downloaded from http://rspb.royalsocietypublishing.org/ on June 29, 2015
rspb.royalsocietypublishing.org
Neural correlates of individual differences in circadian behaviour Jennifer A. Evans1,†, Tanya L. Leise2, Oscar Castanon-Cervantes1 and Alec J. Davidson1
Research Cite this article: Evans JA, Leise TL, Castanon-Cervantes O, Davidson AJ. 2015 Neural correlates of individual differences in circadian behaviour. Proc. R. Soc. B 282: 20150769. http://dx.doi.org/10.1098/rspb.2015.0769
Received: 16 April 2015 Accepted: 27 May 2015
Subject Areas: behaviour, neuroscience Keywords: circadian pacemaker, suprachiasmatic nucleus, individual differences, PERIOD2::LUCIFERASE, bioluminescence
1 2
Morehouse School of Medicine, Neuroscience Institute, 720 Westview Drive SW, Atlanta, GA 30310, USA Department of Mathematics and Statistics, Amherst College, Amherst, MA 01002, USA Daily rhythms in mammals are controlled by the circadian system, which is a collection of biological clocks regulated by a central pacemaker within the suprachiasmatic nucleus (SCN) of the anterior hypothalamus. Changes in SCN function have pronounced consequences for behaviour and physiology; however, few studies have examined whether individual differences in circadian behaviour reflect changes in SCN function. Here, PERIOD2::LUCIFERASE mice were exposed to a behavioural assay to characterize individual differences in baseline entrainment, rate of re-entrainment and free-running rhythms. SCN slices were then collected for ex vivo bioluminescence imaging to gain insight into how the properties of the SCN clock influence individual differences in behavioural rhythms. First, individual differences in the timing of locomotor activity rhythms were positively correlated with the timing of SCN rhythms. Second, slower adjustment during simulated jetlag was associated with a larger degree of phase heterogeneity among SCN neurons. Collectively, these findings highlight the role of the SCN network in determining individual differences in circadian behaviour. Furthermore, these results reveal novel ways that the network organization of the SCN influences plasticity at the behavioural level, and lend insight into potential interventions designed to modulate the rate of resynchronization during transmeridian travel and shift work.
1. Introduction Authors for correspondence: Jennifer A. Evans e-mail: [email protected] Alec J. Davidson e-mail: [email protected]
†
Present address: Department of Biomedical Sciences, Marquette University, Milwaukee, WI, USA. Electronic supplementary material is available at http://dx.doi.org/10.1098/rspb.2015.0769 or via http://rspb.royalsocietypublishing.org.
Inter-individual variability in behavioural, physiological and cellular processes affects normal function, disease susceptibility and response to medical treatments. Thus, understanding the causes and consequences of biological variability is important. Progress has been made in describing mechanisms that produce genetic variability [1–3], and analyses of genetic variation have provided insight into the regulation of complex traits and disease risk [4–6]. However, questions remain concerning how individual differences in behaviour reflect altered neural and physiological function, and to what extent non-genetic mechanisms affect individual differences in such traits. Top-down analyses that investigate neural function by identifying and exploiting natural variation at the phenotypic level have great potential to increase the understanding of individual differences in brain function during both diseased and non-disease states [7]. One of the best characterized brain–behaviour relationships is the control of daily rhythms in mammalian behaviour and physiology [8]. Daily rhythms in mammals are programmed by a central circadian pacemaker located within the suprachiasmatic nucleus (SCN) of the anterior hypothalamus [9]. The SCN is a heterogeneous population of neural oscillators that displays robust rhythms in metabolism, electrical activity and gene expression [10]. Light and other environmental signals synchronize the SCN to the 24 h day, and the SCN in turn synchronizes downstream oscillators through neural and humoral outputs [11]. At the molecular level, transcription, translation and post-translational interactions of ‘clock’ genes and their protein products generate an approximately 24 h cycle in cellular activity [12]. Thus, circadian rhythmicity is influenced by complex processes operating at multiple levels of organization, from the gene to the whole system.
& 2015 The Author(s) Published by the Royal Society. All rights reserved.
Downloaded from http://rspb.royalsocietypublishing.org/ on June 29, 2015
(a) Experimental procedures (i) Breeding and initial husbandry conditions Homozygous PERIOD2::LUCIFERASE (PER2::LUC) knock-in mice [21], backcrossed at least 12 generations to C57Bl/6, were bred and raised under a 24 h light : dark cycle (12 L : 12 D) (lights-off: 1800EST, defined as Zeitgber Time 12, ZT12). Throughout life, ambient temperature was maintained at 22 + 28C, and mice had ad libitum access to water and food.
(ii) Behavioural assay in vivo (figure 1a,b) Male PER2::LUC mice (n ¼ 28, 52 – 55 weeks old) were transferred to individual wheel-running cages. Average photophase luminance was 743 + 146 lux. Mice were maintained under 12 L : 12 D for three weeks before an eastward transmeridian trip was simulated through an abrupt 6 h advance of the scotophase, which is more disruptive compared with a westward shift [22,23]. Four weeks later, mice were transferred to constant darkness (DD) for three weeks. Lastly, mice were re-entrained to 12 L : 12 D for at least three weeks to re-establish stable entrainment before SCN were collected for ex vivo bioluminescence imaging.
(iii) Bioluminescence imaging in vitro (figure 1c–e) SCN slices were collected and imaged as in [24]. Briefly, mice were sacrificed at ZT7 –10, which does not reset the SCN [25,26]. Brains were sectioned in the coronal plane and three consecutive slices (150 mm) were retained from the rostral, middle and caudal portions of the SCN (figure 1c). Each slice was trimmed near the edges of the SCN and cultured with air-buffered medium containing 0.1 mM beetle luciferin (Gold Biotechnologies). SCN slices were imaged using a Stanford Photonics XR Mega 10Z cooled intensified charge-couple device camera.
(iv) Immunohistochemistry (figure 1f ) After imaging, SCN slices were analysed for arginine vasopressin (AVP) and vasoactive intestinal polypeptide (VIP) expression as in [24]. Briefly, slices were treated with colchicine-treated medium (25 mg ml21), fixed with 4% paraformaldehyde, and
(b) Data collection and analyses (i) Behavioural analyses (figure 1a) Wheel-running rhythms were monitored and analysed with the Clocklab system (Actimetrics, Evanston, IL, USA). Activity onset, offset, duration, midpoint and total wheel revolutions were determined (figure 1a). After the 6 h advance of the lightdark (LD) cycle (figure 1a), the number of days to resynchronize was defined for each mouse as the number of days required to shift activity midpoint by 6 h, as in [27]. Re-entrainment was also calculated using activity onset, which correlated with results using activity midpoint (r ¼ 0.79, p , 0.0001). Free-running period under DD was measured by the slope of a regression line fitted to 16 consecutive activity onsets, excluding the first 4 days after release into DD to allow for transient cycles (figure 1a). For behavioural parameters with significant variability across mice, we used the lower and upper quartile to classify mice into discrete phenotypic groups.
(ii) Criteria for slice inclusion As in [24], consistency in slice position was assessed using a metric based on the bioluminescence profile of each slice (figure 1c). For each slice, bioluminescence was summed over the first cycle in vitro, from which the width/height ratio of the bioluminescence profile was calculated (figure 1g). In addition, slice sets were screened based on AVP and VIP expression (figure 1f–g). Three slice sets were discarded owing to morphological deviation. A technical malfunction during data collection prevented inclusion of two slice sets. The remaining slice sets (n ¼ 23) were similar in bioluminescence profile and neuropeptide content. Importantly, the bioluminescence profile and neuropeptide content did not correlate with individual differences in rhythmic parameters measured in vivo or in vitro ( p . 0.2 in all cases).
(iii) MATLAB-based computational analyses As described previously [24,28], rhythmic parameters of PER2::LUC expression were calculated for each slice and for cell-like regions of interest (ROI) within each slice using MATLAB-based scripts. Individual phase maps (figure 1d ) were generated for each 12 pixel diameter ROI judged to exhibit a significant circadian rhythm. For composite phase maps, a representative slice set was selected to which other slice sets were aligned. To locate and extract data from cell-like ROIs, an iterative process was employed after background and local noise subtraction (figure 1e), as in [24,28].
(iv) Statistical analyses Statistical analyses were performed with JMP software. Pearson’s r correlation was used to test the relationship between behavioural phenotype and SCN function. Subsequent tests were performed using one-way ANOVA or full-factorial ANOVA (behaviour group, SCN region, group region interaction). For statistical analyses of SCN cellular function, the average and standard deviation for SCN neurons extracted from slices from each mouse was used to avoid spurious results produced by the large number of cell-like ROIs extracted from SCN slices (average for each mouse: 372 + 7). Cellular data for each individual mouse was normally distributed (Shapiro–Wilk W test, p , 0.0005 in all cases). Data in figures are mean + s.e.m.
2
Proc. R. Soc. B 282: 20150769
2. Material and methods
then cyprotected. Free-floating slices were rinsed in phosphate buffer solution (PBS), and then incubated in primary antibodies (anti-AVP, 1 : 1 K; anti-VIP, 1 : 500) for 48 h at 48C. Slices were rinsed in PBS before 2 h incubation in secondary antibodies (1 : 200). Confocal stacks were obtained using a Olympus FLUOVIEW confocal laser scanning system. AVP-immunoreactive (-ir) and VIP-ir cells were counted with IMAGEJ software.
rspb.royalsocietypublishing.org
Among humans and outbred animal species, there is considerable inter-individual variation in circadian behaviour and the biological basis of this variability is of great interest. For example, people display marked differences in chronotype (e.g. morning larks versus night owls) [13], and this trait is associated with clock gene polymorphisms [14,15]. Chronotype can contribute to variability in a number of other circadian behaviours, including the ability to adjust to transmeridian travel and shift work [16,17]. Understanding the basis of individual differences in circadian behaviour is of interest as this source of variability can influence physical and mental health [18 –20]. While it is clear that the SCN is vital for circadian control of behaviour, it is unclear to what degree inter-individual variability in circadian behaviour reflects individual differences in SCN function. To address this issue in the absence of genetic influences, we phenotyped inbred mice for individual differences in circadian behaviour and then used ex vivo real-time bioluminescence imaging of SCN function. Our results suggest that the network organization of the SCN influences behavioural plasticity and highlight this property as a potential target for interventions designed to modulate resynchronization during transmeridian travel and shift work.
Downloaded from http://rspb.royalsocietypublishing.org/ on June 29, 2015
(b) (i)
3
rspb.royalsocietypublishing.org
days
days
(a)
(ii)
days
Proc. R. Soc. B 282: 20150769
middle
caudal
(g) bioluminescence width/height ratio
(c) rostral
(d )
1.0
a
0.8
b
0.6
c
0.4 0.2 0
8
11
14
ZT peak time (e) 152
147
142
no. AVP cells
500 400
b a
300 200
c
100 0
( f ) AVP
VIP
no. VIP cells
100 80 60
a
a
40 20 0
b rostral middle caudal
Figure 1. Methods for phenotyping individual differences in circadian rhythms in vivo and in vitro. (a) A representative double-plotted actogram illustrating the behavioural assay used to quantify individual differences in circadian rhythms. The phenotype of this mouse was slow re-entrainment, with intermediate activity offset and activity duration. The white and black bar above the actogram illustrates baseline 12 L : 12 D entrainment conditions, with changes in lighting conditions represented with internal shading. Daily times of activity onset, offset and midpoint are represented by the yellow dots, blue dots and red dots, respectively. Arrow indicates the day of re-entrainment. Lines superimposed onto the actogram during constant darkness represent linear regression analyses used to measure the free-running period. (b) Representative actograms illustrating behaviour of a mouse with fast re-entrainment and intermediate activity offset/duration (i) and a mouse with intermediate re-entrainment and long activity offset/duration (ii). (c) Representative slice set illustrating the PER2::LUC bioluminescence profile of SCN slices for mouse in (a). (d) Representative phase maps illustrating the spatio-temporal organization of SCN slices in (c). Phase is colour-coded to indicate the time of peak of PER2::LUC expression on the first cycle in vitro, as quantified in Zeitgeber Time (ZT, ZT12 ¼ projected time of lights-off). (e) Representative images of cell-like ROIs extracted from SCN slices in (c). Colour of circles indicates the ZT peak time of each cell (scale as in (d)). Number in upper left corner indicates number of cell-like ROIs extracted from each slice. ( f ) Immunohistochemical staining of AVP and VIP cell bodies and fibre processes in each slice in (c). (g) Positional analyses of bioluminescence profiles and neuropeptide content of slices (n ¼ 23 SCN slices per rostrocaudal position). Bioluminescence morphology varied systematically among rostrocaudal slices ( p , 0.0001), but not between behavioural groups ( p . 0.2 for each test). Furthermore, the number of AVP and VIP cells differed among rostrocaudal slices ( p , 0.0001 for each test), but not between behavioural groups ( p . 0.4 for each test). Letters above bars indicate the results of post hoc tests using Tukey’s HSD, with significant differences between groups that do not share the same letter, p , 0.02. n ¼ 23/slice. (Online version in colour.)
3. Results (a) Individual differences in circadian behaviour in vivo PER2::LUC mice were phenotyped for individual differences in circadian behaviour (figure 1a,b). Measures of entrainment quantified under 12 L : 12 D before and after the 6 h shift in
the LD cycle were positively correlated, as were measures of entrainment quantified under 12 L : 12 D before and after release into DD (electronic supplementary material, table S1). This indicates that individual differences in circadian behaviour were highly consistent across time even though mice were exposed to multiple environmental manipulations.
Downloaded from http://rspb.royalsocietypublishing.org/ on June 29, 2015
late
(b)
days
4 10 8
4 2
c
10 5 0
b a 11 6 6 fast intermediate slow
Figure 2. Individual differences in circadian behaviour in vivo. (a) Representative double-plotted actograms illustrating individual differences in the time of activity offset and activity duration. (b) Differences in the time of activity offset across phenotypic categories of mice. (c) Representative double-plotted actograms illustrating differences in rate of re-entrainment. (d ) Differences in rate of re-entrainment across phenotypic categories of mice. As in figure 1g, letters above bars indicate the results of Tukey’s HSD, p , 0.02. Numbers at the base of each bar indicate group sample size.
This stability in entrainment measures supports the hypothesis that individual variability in circadian behaviour reflects intrinsic differences in circadian function. We detected individual differences in several rhythmic parameters at the behavioural level (electronic supplementary material, table S2), which were evaluated using a coefficient of variation threshold of more than or equal to 5%. First, mice displayed variation in the time of activity offset and activity duration (figure 2a,b; electronic supplementary material, table S2). These two measures were inter-related because activity onset did not vary between mice. Thus, a subset of mice displayed longer activity duration owing to a later activity offset, whereas other mice displayed shorter activity duration owing to an earlier activity offset. We also detected individual differences in the rate of re-entrainment (figure 2c,d; electronic supplementary material, table S2). In particular, a subset of mice displayed a large advance immediately after the shift in the LD cycle and thereafter re-entrained quickly (less than 6 days), whereas another group of mice displayed little to no change for the first several days after the shift and reentrained relatively slowly (more than 10 days). By contrast, circadian period did not vary appreciably (electronic supplementary material, table S2). Based on these results, we focused subsequent analyses largely on individual differences in the timing of activity offset, activity duration and the rate of re-entrainment. Variability among mice did not reflect differences in environmental light levels or other circadian parameters. First, light intensity was not correlated with the timing of activity offset ( p ¼ 0.30) or the rate of re-entrainment ( p ¼ 0.89). Photophase light intensity did not differ between groups distinct in activity offset (one-way ANOVA, p ¼ 0.75) or rate of re-entrainment (one-way ANOVA, p ¼ 0.82). Further, rate of re-entrainment was not systematically related to other circadian parameters, including time of activity onset, activity offset, activity duration, free-running period or activity levels during the dark phase or light phase ( p . 0.2 in all cases).
We next assessed which measures of SCN function displayed the greatest variability. At the level of the field rhythm for the whole SCN slice, both the ZT peak time and the peak width on the first cycle in vitro varied between samples (electronic supplementary material, table S2). Similar to the low variability in circadian period at the behavioural level, the coefficient of variation for SCN period length was less than 5% (electronic supplementary material, table S2). At the level of SCN neurons, samples varied in the average ZT peak time on the first cycle in vitro, as well as the standard deviation of ZT peak time among SCN neurons on the first cycle in vitro (electronic supplementary material, table S2). By contrast, the coefficient of variation for average period and peak width of SCN neurons was less than 5% (electronic supplementary material, table S2). Based on these results, we focused subsequent analyses on individual differences in the ZT peak time and peak width for the SCN slices, as well as variability in the average and standard deviation of ZT peak time displayed by SCN neurons.
(b) Relationship between timing of behavioural rhythms and suprachiasmatic nucleus rhythms Individual differences in the timing of behavioural rhythms were correlated with rhythms of PER2::LUC expression in SCN slices and neurons (figures 3 and 4; electronic supplementary material, table S3). Behavioural activity offset and activity duration were each positively correlated with the ZT peak time of the rostral, middle and caudal SCN slices (electronic supplementary material, table S3; p 0.005 in each case), as well as the average ZT peak time of all three SCN slices (electronic supplementary material, table S3, p 0.005 in each case). Also, behavioural activity offset and activity duration were each negatively correlated with the peak width of field rhythms displayed by the rostral SCN slice (electronic supplementary material, table S3; one-way ANOVA, p 0.001 in each case). In addition, the timing of activity offset correlated with the ZT peak time
Proc. R. Soc. B 282: 20150769
days
4 13 6 early intermediate late
(d) 15 days to re-entrain
slow
days
intermediate
a
6
0 (c) fast
b
c
rspb.royalsocietypublishing.org
time of activity offset
intermediate
days
(a) early
Downloaded from http://rspb.royalsocietypublishing.org/ on June 29, 2015
early rostral
300
5
middle
300
0
0
–300 60
80
100
caudal
150 0
–300 40
20 (b)
late
–150 20
40 60 80 time in vitro (h)
100
20
40
60
80
100
14
rspb.royalsocietypublishing.org
biolum (arb. units)
(a)
12 a
ab
b
a
ab
b
b a
ab
b a
10
0 early intermediate late
early intermediate late
early intermediate late
early intermediate late
peak width (h)
14
12
a
a ab
b
ab b
10
0
early intermediate late
early intermediate late
early intermediate late
early intermediate late
all
rostral
middle
caudal
Figure 3. Relationship between the time of activity offset in vivo and SCN function in vitro. (a) Representative time series illustrating PER2::LUC rhythms of SCN slices collected from mice that displayed an early time of activity offset (early) or a late time of activity offset (late). (b) Mice that differed in the timing of activity offset differed in the ZT peak time and peak width of PER2::LUC rhythms. As in figure 1g, letters above bars indicate the results of Tukey’s HSD, p , 0.02 (for sample sizes, cf. figure 2b).
of SCN neurons in each slice (electronic supplementary material, table S3, p 0.005 in each case). We further tested these relationships by comparing PER2::LUC rhythms across phenotypic groups (figure 2b). Mice with later times of activity offset displayed later peak times of PER2::LUC rhythms in each SCN slice (figure 3b; one-way ANOVA, p , 0.05). Similarly, longer activity duration was associated with later ZT peak times of PER2::LUC rhythms (data not shown). Further, mice with later times of behavioural activity offset displayed a shorter peak width of PER2::LUC rhythms in the rostral SCN slice and across slices (figure 3b; one-way ANOVA, p , 0.05), and again results were similar for activity duration (data not shown). Lastly, ZT peak time of SCN neurons differed among phenotypic groups, with later times of activity offset associated with later ZT peak times of SCN neurons in each slice (figure 4a; one-way ANOVA, p , 0.05).
(c) Relationship between the rate of re-entrainment and suprachiasmatic nucleus rhythms Individual differences in the rate of re-entrainment were correlated with changes in the function of the SCN at the neuronal level rather than at the whole slice level (electronic
supplementary material, table S3). Rate of re-entrainment was not associated with ZT peak time on the first cycle in vitro for the field rhythm of each SCN slice (electronic supplementary material, table S3; p . 0.15 in each case). On the other hand, rate of re-entrainment was positively correlated with degree of phase heterogeneity among SCN neurons (electronic supplementary material, table S3). When all cells were analysed together, the rate of behavioural re-entrainment was positively correlated with the standard deviation in peak time on the first cycle (electronic supplementary material, table S3; p ¼ 0.03) and the second cycle in vitro ( p ¼ 0.04). When the standard deviation of cellular peak time was assessed in each SCN slice separately, rate of behavioural re-entrainment was positively correlated with phase heterogeneity within the middle SCN on both the first cycle (electronic supplementary material, table S3; p ¼ 0.001) and second cycle in vitro ( p ¼ 0.046). Like results for timing of behavioural rhythms, the relationship between rate of re-entrainment and standard deviation in the peak time of SCN neurons was also detected across distinct behavioural groups (figure 4b). Relative to mice with a fast rate of re-entrainment, mice with a slower rate of re-entrainment displayed a greater degree of phase heterogeneity among SCN neurons in the middle and caudal SCN slice, as well as across slices (figure 4b).
Proc. R. Soc. B 282: 20150769
ZT peak time (h)
b
Downloaded from http://rspb.royalsocietypublishing.org/ on June 29, 2015
(a)
6
b 12 a
b a
b
b a
early intermediate late
early intermediate late
early intermediate late
1.5
standard deviation ZT peak time (h)
b 1.0
a
b
b a
ab
b a
b
0.5
0
fast intermediate slow
fast intermediate slow
fast intermediate slow
fast intermediate slow
all
rostral
middle
caudal
Figure 4. Relationship between circadian behaviour in vivo and PER2::LUC rhythms of SCN neurons in vitro. (a) Mice that differed in the time of activity offset differed in the ZT peak time of SCN neurons. (b) Mice that differed in the rate of re-entrainment differed in the heterogeneity of ZT peak time among SCN neurons. As in figure 1g, letters above bars indicate the results of Tukey’s HSD, p , 0.02 (for sample sizes, cf. figure 2d ).
(d) Relationship between behavioural rhythms and suprachiasmatic nucleus spatio-temporal organization In agreement with the results of a previous study [24], composite maps generated using slices collected from all the mice confirm there are pronounced regional phase differences both across and within rostrocaudal SCN slices (figure 5a). As found in previous detailed analyses [24], one of the most salient organizational features is a late-peaking node located within the rostral slice. Additionally, the middle SCN contained early-peaking regions within the dorsal-most and ventral-most regions and the caudal SCN appeared to be the most homogeneous in phase. Similar to analyses of SCN slices and all SCN neurons, phase analyses indicated that the time of activity offset was related to the timing of PER2::LUC expression (figure 5b,c). For nearly all SCN regions, mice with later times of activity offset showed later ZT peak times of PER2::LUC expression for SCN neurons in regions (full-factorial ANOVA, behaviour group: p , 0.0001, p , 0.02 and p , 0.0005 for rostral, middle and caudal SCN, respectively). Regional phase differences were evident within each slice (full-factorial ANOVA, SCN region: p , 0.05), but this did not interact with behaviour group (full-factorial ANOVA, behaviour region: p . 0.50). In fact, the only SCN region that did not differ among groups was the ventral region of the middle SCN (figure 5c). Overall, similar phase maps were found for groups of mice that differed in the rate of re-entrainment, but there were subtle features that distinguished these two groups (figure 5d– f ). First, fast and slow re-entrainers appeared to differ in the late-peaking region of the rostral SCN, with this region appearing larger in mice with fast re-entrainment (figure 5d).
Furthermore, the phase of peak PER2::LUC expression in the ventral regions within the middle slice and dorsal regions of the caudal slice also appeared to differ based on rate of reentrainment, with slow re-entrainers displaying an earlier phase (figure 5d). Relative to slices from mice that entrained quickly, mice that re-entrained slowly displayed a significantly earlier average phase for SCN neurons within the ventral region of middle SCN (figure 5e). Further, mice that re-entrained slowly displayed more deviation in the phase of SCN neurons within the central region of the middle SCN (figure 5f ).
4. Discussion Circadian rhythms are the product of a dynamic system of interacting genes and tissues. Here, we have used a topdown approach that exploits individual differences in circadian behaviour to gain insight into how the SCN regulates behavioural rhythms. We find a strong relationship between the peak time of SCN PER2::LUC expression and the timing of activity offset. We also find that individual differences in activity duration correspond to rhythmic parameters at the level of the whole slice, which suggests that the basis for this variability lies in the emergent properties of the network. Lastly, we find that the degree of SCN phase heterogeneity is associated with natural variation in the rate of reentrainment during simulated jetlag. Specifically, mice that re-entrained slowly were characterized by a wider distribution of phases among SCN neurons than mice that re-entrained faster. It remains possible that individual differences in circadian behaviour are also influenced by the function of other components of the circadian system (e.g. other clock genes, downstream oscillators). Nevertheless, this study indicates
Proc. R. Soc. B 282: 20150769
early intermediate late
b
b
a
b
10
0 (b)
b
rspb.royalsocietypublishing.org
ZT peak time (h)
14
Downloaded from http://rspb.royalsocietypublishing.org/ on June 29, 2015
rostral
caudal (b)
middle
2
13
7
5
3
8
6 8
11
early
late 14
*
*
12 11
*
* *
*
*
10
ZT peak time
late
9 0 1
2
3
4 5 6 SCN region
7
8
rspb.royalsocietypublishing.org
4
1
7
(c)
early
ZT peak time (h)
(a)
fast
ZT peak time (h)
slow
(f)
slow
fast
13 12 11 10 9
*
0 1
2
3
4 5 6 SCN region
slow
fast
2 standard deviation peak time (h)
(e)
1 * 0
7
8
1
2
3
4 5 6 SCN region
7
8
Figure 5. Relationship between circadian behaviour in vivo and SCN spatio-temporal organization in vitro. (a) Average phase maps generated using slices collected from all mice illustrating regional phase differences on the first cycle in vitro. Average phase maps were generated after superimposing data from both lobes. Numbers indicate regions used for regional analyses of SCN neuronal rhythms. (b) Average phase maps for mice in the early and late activity offset group. (c) Across most SCN regions, average ZT of peak PER2::LUC expression in SCN neurons is delayed in mice that display later times of activity offset. (d ) Average phase maps for mice in the fast and slow re-entrainment group. (e,f ) Rate of re-entrainment was associated with region-specific differences in PER2::LUC peak time and standard deviation of PER2::LUC peak time among SCN neurons. Asterisks (*) denote post hoc least square means contrasts, p , 0.02 (for sample sizes, cf. figure 2b,d ). (Online version in colour.)
that behavioural plasticity is regulated by SCN network properties, which highlights this as a potential target for interventions designed to modulate circadian behaviour. We found a robust relationship between the time of activity offset in vivo and the phase of SCN PER2::LUC expression in vitro, indicating that the entrained phase of the SCN relates to the timing of locomotor activity. While this result was expected, it represents, to our knowledge, the first direct evidence that individual differences in phase angle of entrainment are encoded by the SCN. Also, we found that individual variability in behavioural activity duration was related to properties evident at the population level, which is consistent with recent work investigating seasonal encoding within the SCN [29–34]. Of note, previous research also indicates a strong link between behavioural period and SCN function [35,36], but we were unable to address this issue here because our mice did not display marked variability in behavioural period. Our study adds to a growing body of research that suggests phase heterogeneity within the SCN affects the magnitude of photic responses. First, photic phase-resetting responses are attenuated by long-day photoperiods that cause phase dispersion among SCN neurons [37–39]. A complementary body of work likewise suggests that changes in SCN spatio-temporal organization can abolish photic resetting responses. Specifically, under 22 h light : dark cycles, rat locomotor rhythms dissociate into two components that are differentially controlled by the ventral SCN and the dorsal SCN [40,41]. This protocol allows investigation of circadian responses to light when the SCN is in a dissociated versus coherent state. When the rhythms of these two SCN regions are in phase with one another, photic stimuli elicit phase shifts and physiological correlates of photic responses [40].
When these two SCN compartments are out of phase, the transmission of photic input from the ventral to dorsal SCN is blocked and no behavioural response is elicited. Collectively, changes in photic resetting after exposure to seasonal and non-seasonal lighting conditions demonstrate that the capacity to respond to light is determined by the phase coherence among SCN neurons and the efficacy of regional communication. The current finding that greater SCN phase heterogeneity is associated with slower rate of re-entrainment may represent a less extreme form of this principle where phase coherence is reduced in some mice and re-entrainment takes several more cycles of exposure to the phase-resetting stimulus. Changes in phase coherence among SCN neurons may relate to decreased rate of re-entrainment in other contexts, like that typically observed with ageing [28]. Our study highlights that the bases of individual differences in rate of re-entrainment may relate to the function of specific regions in the SCN network. We find that the relationship between phase coherence and resetting speed was the strongest for the middle SCN slice, which is of interest as this slice contains several types of light-responsive cells [42–45]. Furthermore, the ventral and central regions of the middle SCN were those that differed most between fast and slow re-entrainers, with mice that re-entrained slowly displaying a ventral region with an earlier phase and a central region with more phase heterogeneity. That our analyses reveal an effect for these specific regions is of interest because the ventral region of the middle SCN contains VIP-ir cells (figure 1f ), which are important for the initial processing of photic input before propagation to the rest of the SCN network [46–48]. Likewise, the central region of the middle SCN has been proposed to act as a gate in the flow of information between the
Proc. R. Soc. B 282: 20150769
(d)
Downloaded from http://rspb.royalsocietypublishing.org/ on June 29, 2015
Together with previous experiments investigating individual differences in circadian behaviour [50,51], this study demonstrates the use of top-down analyses of brain function. Individual differences in behavioural, physiological and cellular processes affect numerous aspects of biological function, and thus, understanding the bases of this variation remains important. Future studies investigating the underlying genetic, epigenetic and molecular mechanisms that produce inter-individual variability in SCN function may lend insight into ways to manipulate plasticity of the circadian system under a wide variety of conditions relevant to human health and well-being.
Data accessibility. The raw data (.tif image sequences and actograms for all mice) have been uploaded to Dryad: http://datadryad.org/ review?doi=doi:10.5061/dryad.4rd1g. Author contributions. J.A.E. and A.J.D. designed the research. J.A.E. performed experiments and analysed data. J.A.E., T.L.L., O.C.C. and A.J.D. developed analytical tools. J.A.E., T.L.L., O.C.C. and A.J.D. drafted the manuscript. All authors gave final approval for publication.
Funding. This work was supported by NIH grant nos. U54NS060659, F32NS071935, and S21MD000101, the Georgia Research Alliance, and the NSF Center for Behavioral Neuroscience.
Competing interests. We declare we have no competing interests. Acknowledgements. We wish to thank Stanford Photonics for their equipment and assistance. Also, we are grateful to the Morehouse School of Medicine CLAR staff for providing excellent animal care.
References 1.
2.
3.
4.
5.
6.
7.
8.
Jakobsson M et al. 2008 Genotype, haplotype and copy-number variation in worldwide human populations. Nature 451, 998– 1003. (doi:10.1038/ nature06742) Frazer KA et al. 2007 A sequence-based variation map of 8.27 million SNPs in inbred mouse strains. Nature 448, 1050–1053. (doi:10.1038/nature06067) Ayroles JF et al. 2009 Systems genetics of complex traits in Drosophila melanogaster. Nat. Genet. 41, 299–307. (doi:10.1038/ng.332) Blakey JD, Hall IP. 2011 Current progress in pharmacogenetics. Br. J. Clin. Pharmacol. 71, 824–831. (doi:10.1111/j.1365-2125.2011.03912.x) Fanciulli M, Petretto E, Aitman TJ. 2010 Gene copy number variation and common human disease. Clin. Genet. 77, 201– 213. (doi:10.1111/j.1399-0004. 2009.01342.x) Vannberg FO, Chapman SJ, Hill AV. 2011 Human genetic susceptibility to intracellular pathogens. Immunol. Rev. 240, 105– 116. (doi:10.1111/j.1600065X.2010.00996.x) Bennett AF. 1987 Interindividual variability: an underutilized resource. In New directions in ecological physiology (eds ME Feder, AF Bennett, WW Burggren, RB Huey), pp. 147–169. New York, NY: Cambridge University Press. Gachon F, Nagoshi E, Brown SA, Ripperger J, Schibler U. 2004 The mammalian circadian timing system: from gene expression to physiology. Chromosoma 113, 103– 112. (doi:10.1007/s00412004-0296-2)
9.
10.
11.
12.
13.
14.
15.
16.
Weaver DR. 1998 The suprachiasmatic nucleus: a 25year retrospective. J. Biol. Rhythms 13, 100 –112. (doi:10.1177/074873098128999952) Welsh DK, Takahashi JS, Kay SA. 2010 Suprachiasmatic nucleus: cell autonomy and network properties. Annu. Rev. Physiol. 72, 551– 577. (doi:10.1146/annurev-physiol-021909-135919) Dibner C, Schibler U, Albrecht U. 2010 The mammalian circadian timing system: organization and coordination of central and peripheral clocks. Annu. Rev. Physiol. 72, 517 –549. (doi:10.1146/ annurev-physiol-021909-135821) Ko CH, Takahashi JS. 2006 Molecular components of the mammalian circadian clock. Hum. Mol. Genet. 15, R271 –R277. (doi:10.1093/hmg/ddl207) Roenneberg T, Wirz-Justice A, Merrow M. 2003 Life between clocks: daily temporal patterns of human chronotypes. J. Biol. Rhythms 18, 80 –90. (doi:10. 1177/0748730402239679) Katzenberg D, Young T, Finn L, Lin L, King DP, Takahashi JS, Mignot E. 1998 A CLOCK polymorphism associated with human diurnal preference. Sleep 21, 569–576. Archer J, Von Schantz M. 2003 A length polymorphism in the circadian clock gene Per3 is linked to delayed sleep phase syndrome and extreme diurnal preference. Sleep 26, 413–415. Gander PH, Myhre G, Graeber RC, Andersen HT, Lauber JK. 1989 Adjustment of sleep and the circadian temperature rhythm after flights across nine time zones. Aviat. Space Environ. Med. 60, 733–743.
17. Gamble KL et al. 2011 Shift work in nurses: contribution of phenotypes and genotypes to adaptation. PLoS ONE 6, e18395. (doi:10.1371/ journal.pone.0018395) 18. Martino TA et al. 2008 Circadian rhythm disorganization produces profound cardiovascular and renal disease in hamsters. Am. J. Physiol. Regul. Integr. Comp. Physiol. 294, R1675 –R1683. (doi:10. 1152/ajpregu.00829.2007) 19. Sjoholm LK, Backlund L, Cheteh EH, Ek IR, Frisen L, Schalling M, Osby U, Lavebratt C, Nikamo P. 2010 CRY2 is associated with rapid cycling in bipolar disorder patients. PLoS ONE 5, e12632. (doi:10. 1371/journal.pone.0012632) 20. Lee HJ, Rex KM, Nievergelt CM, Kelsoe JR, Kripke DF. 2011 Delayed sleep phase syndrome is related to seasonal affective disorder. J. Affect. Disord. 133, 573–579. (doi:10.1016/j.jad.2011.04.046) 21. Yoo SH et al. 2004 PERIOD2::LUCIFERASE real-time reporting of circadian dynamics reveals persistent circadian oscillations in mouse peripheral tissues. Proc. Natl Acad. Sci. USA 101, 5339–5346. (doi:10. 1073/pnas.0308709) 22. Zee PC, Rosenberg RS, Turek FW. 1992 Effects of aging on entrainment and rate of resynchronization of circadian locomotor activity. Am. J. Physiol. 263, R1099–R1103. 23. Davidson AJ, Sellix MT, Daniel J, Yamazaki S, Menaker M, Block GD. 2006 Chronic jet-lag increases mortality in aged mice. Curr. Biol. 16, R914 –R916. (doi:10.1016/j.cub.2006.09.058)
Proc. R. Soc. B 282: 20150769
Ethics. The experiments described were approved by the MSM Institutional Animal Care and Use Committee.
8
rspb.royalsocietypublishing.org
ventral SCN and other SCN regions [49]. Thus, the relative phase and the degree of heterogeneity within the ventral and central SCN probably modulate the processing of photic input and transmission to other SCN compartments. An important question concerns the functional role of regional phase differences in the SCN. Our work indicates that phase differences are a limiting factor in the response to photic manipulations. The fact that regional phase differences are a common feature of SCN organization could then be viewed as a barrier to the rapid adjustment to transmeridian travel and shift work. Under the ecological conditions experienced during evolution, the circadian system would not experience abrupt, large environmental changes, but rather gradual changes owing to seasonal variation. Because the environmental signals that synchronize the clock vary annually, the circadian system must be flexible enough to track the seasons and yet robust against minor perturbations produced by stochastic variation in light levels that can occur from day to day. As a mechanism that limits rapid resetting, SCN phase heterogeneity may be advantageous in that it provides both stability under static conditions and flexibility during environmental change. When viewed in this manner, the subgroup of fast re-entrainers described here may in fact be displaying the aberrant response. But further insight into the neural and genetic mechanisms that underlie this rapid response may be used to facilitate rapid resynchronization for travellers and shift workers.
Downloaded from http://rspb.royalsocietypublishing.org/ on June 29, 2015
34.
35.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
J. Neurosci. 24, 68 –75. (doi:10.1523/JNEUROSCI. 1666-03.2004) Dardente H, Poirel VJ, Klosen P, Pevet P, MassonPevet M. 2002 Per and neuropeptide expression in the rat suprachiasmatic nuclei: compartmentalization and differential cellular induction by light. Brain Res. 958, 261– 271. (doi:10.1016/S0006-8993(02)03563-1) Romijn HJ, Sluiter AA, Pool CW, Wortel J, Buijs RM. 1996 Differences in colocalization between Fos and PHI, GRP, VIP and VP in neurons of the rat suprachiasmatic nucleus after a light stimulus during the phase delay versus the phase advance period of the night. J. Comp. Neurol. 372, 1–8. (doi:10.1002/(SICI)1096-9861(19960812)372:1,1:: AID-CNE1.3.0.CO;2-7) Hamada T, Antle MC, Silver R. 2004 Temporal and spatial expression patterns of canonical clock genes and clock-controlled genes in the suprachiasmatic nucleus. Eur. J. Neurosci. 19, 1741 –1748. (doi:10. 1111/j.1460-9568.2004.03275.x) Shigeyoshi Y et al. 1997 Light-induced resetting of a mammalian circadian clock is associated with rapid induction of the mPer1 transcript. Cell 91, 1043 –1053. (doi:10.1016/S00928674(00)80494-8) Yan L, Silver R. 2002 Differential induction and localization of mPer1 and mPer2 during advancing and delaying phase shifts. Eur. J. Neurosci. 16, 1531–1540. (doi:10.1046/j.1460-9568.2002.02224.x) Antle MC, Foley DK, Foley NC, Silver R. 2003 Gates and oscillators: a network model of the brain clock. J. Biol. Rhythms 18, 339–350. (doi:10.1177/ 0748730403253840) Margraf RR, Zlomanczuk P, Liskin LA, Lynch GR. 1991 Circadian differences in neuronal activity of the suprachiasmatic nucleus in brain slices prepared from photo-responsive and photonon-responsive Djungarian hamsters. Brain Res. 544, 42 – 48. (doi:10.1016/00068993(91)90883-W) Ciarleglio CM, Gamble KL, Axley JC, Strauss BR, Cohen JY, Colwell CS, McMahon DG. 2009 Population encoding by circadian clock neurons organizes circadian behavior. J. Neurosci. 29, 1670 – 1676. (doi:10.1523/JNEUROSCI.3801-08. 2009)
9
Proc. R. Soc. B 282: 20150769
36.
suprachiasmatic nuclei neurons and their response to photoperiod. J. Biol. Rhythms 24, 44 –54. (doi:10.1177/0748730408327918) Hazlerigg DG, Ebling FJ, Johnston JD. 2005 Photoperiod differentially regulates gene expression rhythms in the rostral and caudal SCN. Curr. Biol. 15, R449 –R450. (doi:10.1016/j.cub.2005.06.010) Low-Zeddies SS, Takahashi JS. 2001 Chimera analysis of the Clock mutation in mice shows that complex cellular integration determines circadian behavior. Cell 105, 25 –42. (doi:10.1016/S00928674(01)00294-X) Ralph MR, Hurd MW. 1995 Circadian pacemakers in vertebrates. Ciba Found. Symp. 183, 67 –81; discussion 81 –67. vanderLeest HT, Rohling JH, Michel S, Meijer JH. 2009 Phase shifting capacity of the circadian pacemaker determined by the SCN neuronal network organization. PLoS ONE 4, e4976. (doi:10. 1371/journal.pone.0004976) Refinetti R. 2002 Compression and expansion of circadian rhythm in mice under long and short photoperiods. Integr. Physiol. Behav. Sci. 37, 114 – 127. (doi:10.1007/BF02688824) Pittendrigh CS, Elliott JA, Takamura T. 1984 The circadian component in photoperiodic induction. Ciba Found. Symp. 104, 26– 47. (doi:10.1002/ 9780470720851.ch4) Schwartz MD, Congdon S, de la Iglesia HO. 2010 Phase misalignment between suprachiasmatic neuronal oscillators impairs photic behavioral phase shifts but not photic induction of gene expression. J. Neurosci. 30, 13 150 –13 156. (doi:10.1523/ JNEUROSCI.1853-10.2010) de la Iglesia HO, Cambras T, Schwartz WJ, DiezNoguera A. 2004 Forced desynchronization of dual circadian oscillators within the rat suprachiasmatic nucleus. Curr. Biol. 14, 796 –800. (doi:10.1016/j. cub.2004.04.034) Kuhlman SJ, Silver R, Le Sauter J, Bult-Ito A, McMahon DG. 2003 Phase resetting light pulses induce Per1 and persistent spike activity in a subpopulation of biological clock neurons. J. Neurosci 23, 1441–1450. Karatsoreos IN, Yan L, LeSauter J, Silver R. 2004 Phenotype matters: identification of light-responsive cells in the mouse suprachiasmatic nucleus.
rspb.royalsocietypublishing.org
24. Evans JA, Leise TL, Castanon-Cervantes O, Davidson AJ. 2011 Intrinsic regulation of spatiotemporal organization within the suprachiasmatic nucleus. PLoS ONE 6, e15869. (doi:10.1371/journal.pone.0015869) 25. Yoshikawa T, Yamazaki S, Menaker M. 2005 Effects of preparation time on phase of cultured tissues reveal complexity of circadian organization. J. Biol. Rhythms 20, 500–512. (doi:10.1177/074873040528 0775) 26. Davidson AJ, Castanon-Cervantes O, Leise TL, Molyneux PC, Harrington ME. 2009 Visualizing jet lag in the mouse suprachiasmatic nucleus and peripheral circadian timing system. Eur. J. Neurosci. 29, 171–180. (doi:10.1111/j.1460-9568.2008. 06534.x) 27. Evans JA, Elliott JA, Gorman MR. 2009 Dim nighttime illumination accelerates adjustment to timezone travel in an animal model. Curr. Biol. 19, R156 –R157. (doi:10.1016/j.cub.2009.01.023) 28. Sellix MT, Evans JA, Leise TL, Castanon-Cervantes O, Hill DD, DeLisser P, Block GD, Menaker M, Davidson AJ. 2012 Aging differentially affects the reentrainment response of central and peripheral circadian oscillators. J. Neurosci. 32, 16 193 – 16 202. (doi:10.1523/JNEUROSCI.3559-12.2012) 29. Meijer JH, Michel S, vanderLeest HT, Rohling JH. 2010 Daily and seasonal adaptation of the circadian clock requires plasticity of the SCN neuronal network. Eur. J. Neurosci. 32, 2143 –2151. (doi:10. 1111/j.1460-9568.2010.07522.x) 30. Inagaki N, Honma S, Ono D, Tanahashi Y, Honma K. 2007 Separate oscillating cell groups in mouse suprachiasmatic nucleus couple photoperiodically to the onset and end of daily activity. Proc. Natl Acad. Sci. USA 104, 7664 –7669. (doi:10.1073/pnas. 0607713104) 31. Naito E, Watanabe T, Tei H, Yoshimura T, Ebihara S. 2008 Reorganization of the suprachiasmatic nucleus coding for day length. J. Biol. Rhythms 23, 140–149. (doi:10.1177/0748730408314572) 32. Schaap J, Albus H, vanderLeest HT, Eilers PH, Detari L, Meijer JH. 2003 Heterogeneity of rhythmic suprachiasmatic nucleus neurons: implications for circadian waveform and photoperiodic encoding. Proc. Natl Acad. Sci. USA 100, 15 994 –15 999. (doi:10.1073/pnas.2436298100) 33. Brown TM, Piggins HD. 2009 Spatiotemporal heterogeneity in the electrical activity of
rspb.royalsocietypublishing.org
Neural correlates of individual differences in circadian behaviour Jennifer A. Evans1,†, Tanya L. Leise2, Oscar Castanon-Cervantes1 and Alec J. Davidson1
Research Cite this article: Evans JA, Leise TL, Castanon-Cervantes O, Davidson AJ. 2015 Neural correlates of individual differences in circadian behaviour. Proc. R. Soc. B 282: 20150769. http://dx.doi.org/10.1098/rspb.2015.0769
Received: 16 April 2015 Accepted: 27 May 2015
Subject Areas: behaviour, neuroscience Keywords: circadian pacemaker, suprachiasmatic nucleus, individual differences, PERIOD2::LUCIFERASE, bioluminescence
1 2
Morehouse School of Medicine, Neuroscience Institute, 720 Westview Drive SW, Atlanta, GA 30310, USA Department of Mathematics and Statistics, Amherst College, Amherst, MA 01002, USA Daily rhythms in mammals are controlled by the circadian system, which is a collection of biological clocks regulated by a central pacemaker within the suprachiasmatic nucleus (SCN) of the anterior hypothalamus. Changes in SCN function have pronounced consequences for behaviour and physiology; however, few studies have examined whether individual differences in circadian behaviour reflect changes in SCN function. Here, PERIOD2::LUCIFERASE mice were exposed to a behavioural assay to characterize individual differences in baseline entrainment, rate of re-entrainment and free-running rhythms. SCN slices were then collected for ex vivo bioluminescence imaging to gain insight into how the properties of the SCN clock influence individual differences in behavioural rhythms. First, individual differences in the timing of locomotor activity rhythms were positively correlated with the timing of SCN rhythms. Second, slower adjustment during simulated jetlag was associated with a larger degree of phase heterogeneity among SCN neurons. Collectively, these findings highlight the role of the SCN network in determining individual differences in circadian behaviour. Furthermore, these results reveal novel ways that the network organization of the SCN influences plasticity at the behavioural level, and lend insight into potential interventions designed to modulate the rate of resynchronization during transmeridian travel and shift work.
1. Introduction Authors for correspondence: Jennifer A. Evans e-mail: [email protected] Alec J. Davidson e-mail: [email protected]
†
Present address: Department of Biomedical Sciences, Marquette University, Milwaukee, WI, USA. Electronic supplementary material is available at http://dx.doi.org/10.1098/rspb.2015.0769 or via http://rspb.royalsocietypublishing.org.
Inter-individual variability in behavioural, physiological and cellular processes affects normal function, disease susceptibility and response to medical treatments. Thus, understanding the causes and consequences of biological variability is important. Progress has been made in describing mechanisms that produce genetic variability [1–3], and analyses of genetic variation have provided insight into the regulation of complex traits and disease risk [4–6]. However, questions remain concerning how individual differences in behaviour reflect altered neural and physiological function, and to what extent non-genetic mechanisms affect individual differences in such traits. Top-down analyses that investigate neural function by identifying and exploiting natural variation at the phenotypic level have great potential to increase the understanding of individual differences in brain function during both diseased and non-disease states [7]. One of the best characterized brain–behaviour relationships is the control of daily rhythms in mammalian behaviour and physiology [8]. Daily rhythms in mammals are programmed by a central circadian pacemaker located within the suprachiasmatic nucleus (SCN) of the anterior hypothalamus [9]. The SCN is a heterogeneous population of neural oscillators that displays robust rhythms in metabolism, electrical activity and gene expression [10]. Light and other environmental signals synchronize the SCN to the 24 h day, and the SCN in turn synchronizes downstream oscillators through neural and humoral outputs [11]. At the molecular level, transcription, translation and post-translational interactions of ‘clock’ genes and their protein products generate an approximately 24 h cycle in cellular activity [12]. Thus, circadian rhythmicity is influenced by complex processes operating at multiple levels of organization, from the gene to the whole system.
& 2015 The Author(s) Published by the Royal Society. All rights reserved.
Downloaded from http://rspb.royalsocietypublishing.org/ on June 29, 2015
(a) Experimental procedures (i) Breeding and initial husbandry conditions Homozygous PERIOD2::LUCIFERASE (PER2::LUC) knock-in mice [21], backcrossed at least 12 generations to C57Bl/6, were bred and raised under a 24 h light : dark cycle (12 L : 12 D) (lights-off: 1800EST, defined as Zeitgber Time 12, ZT12). Throughout life, ambient temperature was maintained at 22 + 28C, and mice had ad libitum access to water and food.
(ii) Behavioural assay in vivo (figure 1a,b) Male PER2::LUC mice (n ¼ 28, 52 – 55 weeks old) were transferred to individual wheel-running cages. Average photophase luminance was 743 + 146 lux. Mice were maintained under 12 L : 12 D for three weeks before an eastward transmeridian trip was simulated through an abrupt 6 h advance of the scotophase, which is more disruptive compared with a westward shift [22,23]. Four weeks later, mice were transferred to constant darkness (DD) for three weeks. Lastly, mice were re-entrained to 12 L : 12 D for at least three weeks to re-establish stable entrainment before SCN were collected for ex vivo bioluminescence imaging.
(iii) Bioluminescence imaging in vitro (figure 1c–e) SCN slices were collected and imaged as in [24]. Briefly, mice were sacrificed at ZT7 –10, which does not reset the SCN [25,26]. Brains were sectioned in the coronal plane and three consecutive slices (150 mm) were retained from the rostral, middle and caudal portions of the SCN (figure 1c). Each slice was trimmed near the edges of the SCN and cultured with air-buffered medium containing 0.1 mM beetle luciferin (Gold Biotechnologies). SCN slices were imaged using a Stanford Photonics XR Mega 10Z cooled intensified charge-couple device camera.
(iv) Immunohistochemistry (figure 1f ) After imaging, SCN slices were analysed for arginine vasopressin (AVP) and vasoactive intestinal polypeptide (VIP) expression as in [24]. Briefly, slices were treated with colchicine-treated medium (25 mg ml21), fixed with 4% paraformaldehyde, and
(b) Data collection and analyses (i) Behavioural analyses (figure 1a) Wheel-running rhythms were monitored and analysed with the Clocklab system (Actimetrics, Evanston, IL, USA). Activity onset, offset, duration, midpoint and total wheel revolutions were determined (figure 1a). After the 6 h advance of the lightdark (LD) cycle (figure 1a), the number of days to resynchronize was defined for each mouse as the number of days required to shift activity midpoint by 6 h, as in [27]. Re-entrainment was also calculated using activity onset, which correlated with results using activity midpoint (r ¼ 0.79, p , 0.0001). Free-running period under DD was measured by the slope of a regression line fitted to 16 consecutive activity onsets, excluding the first 4 days after release into DD to allow for transient cycles (figure 1a). For behavioural parameters with significant variability across mice, we used the lower and upper quartile to classify mice into discrete phenotypic groups.
(ii) Criteria for slice inclusion As in [24], consistency in slice position was assessed using a metric based on the bioluminescence profile of each slice (figure 1c). For each slice, bioluminescence was summed over the first cycle in vitro, from which the width/height ratio of the bioluminescence profile was calculated (figure 1g). In addition, slice sets were screened based on AVP and VIP expression (figure 1f–g). Three slice sets were discarded owing to morphological deviation. A technical malfunction during data collection prevented inclusion of two slice sets. The remaining slice sets (n ¼ 23) were similar in bioluminescence profile and neuropeptide content. Importantly, the bioluminescence profile and neuropeptide content did not correlate with individual differences in rhythmic parameters measured in vivo or in vitro ( p . 0.2 in all cases).
(iii) MATLAB-based computational analyses As described previously [24,28], rhythmic parameters of PER2::LUC expression were calculated for each slice and for cell-like regions of interest (ROI) within each slice using MATLAB-based scripts. Individual phase maps (figure 1d ) were generated for each 12 pixel diameter ROI judged to exhibit a significant circadian rhythm. For composite phase maps, a representative slice set was selected to which other slice sets were aligned. To locate and extract data from cell-like ROIs, an iterative process was employed after background and local noise subtraction (figure 1e), as in [24,28].
(iv) Statistical analyses Statistical analyses were performed with JMP software. Pearson’s r correlation was used to test the relationship between behavioural phenotype and SCN function. Subsequent tests were performed using one-way ANOVA or full-factorial ANOVA (behaviour group, SCN region, group region interaction). For statistical analyses of SCN cellular function, the average and standard deviation for SCN neurons extracted from slices from each mouse was used to avoid spurious results produced by the large number of cell-like ROIs extracted from SCN slices (average for each mouse: 372 + 7). Cellular data for each individual mouse was normally distributed (Shapiro–Wilk W test, p , 0.0005 in all cases). Data in figures are mean + s.e.m.
2
Proc. R. Soc. B 282: 20150769
2. Material and methods
then cyprotected. Free-floating slices were rinsed in phosphate buffer solution (PBS), and then incubated in primary antibodies (anti-AVP, 1 : 1 K; anti-VIP, 1 : 500) for 48 h at 48C. Slices were rinsed in PBS before 2 h incubation in secondary antibodies (1 : 200). Confocal stacks were obtained using a Olympus FLUOVIEW confocal laser scanning system. AVP-immunoreactive (-ir) and VIP-ir cells were counted with IMAGEJ software.
rspb.royalsocietypublishing.org
Among humans and outbred animal species, there is considerable inter-individual variation in circadian behaviour and the biological basis of this variability is of great interest. For example, people display marked differences in chronotype (e.g. morning larks versus night owls) [13], and this trait is associated with clock gene polymorphisms [14,15]. Chronotype can contribute to variability in a number of other circadian behaviours, including the ability to adjust to transmeridian travel and shift work [16,17]. Understanding the basis of individual differences in circadian behaviour is of interest as this source of variability can influence physical and mental health [18 –20]. While it is clear that the SCN is vital for circadian control of behaviour, it is unclear to what degree inter-individual variability in circadian behaviour reflects individual differences in SCN function. To address this issue in the absence of genetic influences, we phenotyped inbred mice for individual differences in circadian behaviour and then used ex vivo real-time bioluminescence imaging of SCN function. Our results suggest that the network organization of the SCN influences behavioural plasticity and highlight this property as a potential target for interventions designed to modulate resynchronization during transmeridian travel and shift work.
Downloaded from http://rspb.royalsocietypublishing.org/ on June 29, 2015
(b) (i)
3
rspb.royalsocietypublishing.org
days
days
(a)
(ii)
days
Proc. R. Soc. B 282: 20150769
middle
caudal
(g) bioluminescence width/height ratio
(c) rostral
(d )
1.0
a
0.8
b
0.6
c
0.4 0.2 0
8
11
14
ZT peak time (e) 152
147
142
no. AVP cells
500 400
b a
300 200
c
100 0
( f ) AVP
VIP
no. VIP cells
100 80 60
a
a
40 20 0
b rostral middle caudal
Figure 1. Methods for phenotyping individual differences in circadian rhythms in vivo and in vitro. (a) A representative double-plotted actogram illustrating the behavioural assay used to quantify individual differences in circadian rhythms. The phenotype of this mouse was slow re-entrainment, with intermediate activity offset and activity duration. The white and black bar above the actogram illustrates baseline 12 L : 12 D entrainment conditions, with changes in lighting conditions represented with internal shading. Daily times of activity onset, offset and midpoint are represented by the yellow dots, blue dots and red dots, respectively. Arrow indicates the day of re-entrainment. Lines superimposed onto the actogram during constant darkness represent linear regression analyses used to measure the free-running period. (b) Representative actograms illustrating behaviour of a mouse with fast re-entrainment and intermediate activity offset/duration (i) and a mouse with intermediate re-entrainment and long activity offset/duration (ii). (c) Representative slice set illustrating the PER2::LUC bioluminescence profile of SCN slices for mouse in (a). (d) Representative phase maps illustrating the spatio-temporal organization of SCN slices in (c). Phase is colour-coded to indicate the time of peak of PER2::LUC expression on the first cycle in vitro, as quantified in Zeitgeber Time (ZT, ZT12 ¼ projected time of lights-off). (e) Representative images of cell-like ROIs extracted from SCN slices in (c). Colour of circles indicates the ZT peak time of each cell (scale as in (d)). Number in upper left corner indicates number of cell-like ROIs extracted from each slice. ( f ) Immunohistochemical staining of AVP and VIP cell bodies and fibre processes in each slice in (c). (g) Positional analyses of bioluminescence profiles and neuropeptide content of slices (n ¼ 23 SCN slices per rostrocaudal position). Bioluminescence morphology varied systematically among rostrocaudal slices ( p , 0.0001), but not between behavioural groups ( p . 0.2 for each test). Furthermore, the number of AVP and VIP cells differed among rostrocaudal slices ( p , 0.0001 for each test), but not between behavioural groups ( p . 0.4 for each test). Letters above bars indicate the results of post hoc tests using Tukey’s HSD, with significant differences between groups that do not share the same letter, p , 0.02. n ¼ 23/slice. (Online version in colour.)
3. Results (a) Individual differences in circadian behaviour in vivo PER2::LUC mice were phenotyped for individual differences in circadian behaviour (figure 1a,b). Measures of entrainment quantified under 12 L : 12 D before and after the 6 h shift in
the LD cycle were positively correlated, as were measures of entrainment quantified under 12 L : 12 D before and after release into DD (electronic supplementary material, table S1). This indicates that individual differences in circadian behaviour were highly consistent across time even though mice were exposed to multiple environmental manipulations.
Downloaded from http://rspb.royalsocietypublishing.org/ on June 29, 2015
late
(b)
days
4 10 8
4 2
c
10 5 0
b a 11 6 6 fast intermediate slow
Figure 2. Individual differences in circadian behaviour in vivo. (a) Representative double-plotted actograms illustrating individual differences in the time of activity offset and activity duration. (b) Differences in the time of activity offset across phenotypic categories of mice. (c) Representative double-plotted actograms illustrating differences in rate of re-entrainment. (d ) Differences in rate of re-entrainment across phenotypic categories of mice. As in figure 1g, letters above bars indicate the results of Tukey’s HSD, p , 0.02. Numbers at the base of each bar indicate group sample size.
This stability in entrainment measures supports the hypothesis that individual variability in circadian behaviour reflects intrinsic differences in circadian function. We detected individual differences in several rhythmic parameters at the behavioural level (electronic supplementary material, table S2), which were evaluated using a coefficient of variation threshold of more than or equal to 5%. First, mice displayed variation in the time of activity offset and activity duration (figure 2a,b; electronic supplementary material, table S2). These two measures were inter-related because activity onset did not vary between mice. Thus, a subset of mice displayed longer activity duration owing to a later activity offset, whereas other mice displayed shorter activity duration owing to an earlier activity offset. We also detected individual differences in the rate of re-entrainment (figure 2c,d; electronic supplementary material, table S2). In particular, a subset of mice displayed a large advance immediately after the shift in the LD cycle and thereafter re-entrained quickly (less than 6 days), whereas another group of mice displayed little to no change for the first several days after the shift and reentrained relatively slowly (more than 10 days). By contrast, circadian period did not vary appreciably (electronic supplementary material, table S2). Based on these results, we focused subsequent analyses largely on individual differences in the timing of activity offset, activity duration and the rate of re-entrainment. Variability among mice did not reflect differences in environmental light levels or other circadian parameters. First, light intensity was not correlated with the timing of activity offset ( p ¼ 0.30) or the rate of re-entrainment ( p ¼ 0.89). Photophase light intensity did not differ between groups distinct in activity offset (one-way ANOVA, p ¼ 0.75) or rate of re-entrainment (one-way ANOVA, p ¼ 0.82). Further, rate of re-entrainment was not systematically related to other circadian parameters, including time of activity onset, activity offset, activity duration, free-running period or activity levels during the dark phase or light phase ( p . 0.2 in all cases).
We next assessed which measures of SCN function displayed the greatest variability. At the level of the field rhythm for the whole SCN slice, both the ZT peak time and the peak width on the first cycle in vitro varied between samples (electronic supplementary material, table S2). Similar to the low variability in circadian period at the behavioural level, the coefficient of variation for SCN period length was less than 5% (electronic supplementary material, table S2). At the level of SCN neurons, samples varied in the average ZT peak time on the first cycle in vitro, as well as the standard deviation of ZT peak time among SCN neurons on the first cycle in vitro (electronic supplementary material, table S2). By contrast, the coefficient of variation for average period and peak width of SCN neurons was less than 5% (electronic supplementary material, table S2). Based on these results, we focused subsequent analyses on individual differences in the ZT peak time and peak width for the SCN slices, as well as variability in the average and standard deviation of ZT peak time displayed by SCN neurons.
(b) Relationship between timing of behavioural rhythms and suprachiasmatic nucleus rhythms Individual differences in the timing of behavioural rhythms were correlated with rhythms of PER2::LUC expression in SCN slices and neurons (figures 3 and 4; electronic supplementary material, table S3). Behavioural activity offset and activity duration were each positively correlated with the ZT peak time of the rostral, middle and caudal SCN slices (electronic supplementary material, table S3; p 0.005 in each case), as well as the average ZT peak time of all three SCN slices (electronic supplementary material, table S3, p 0.005 in each case). Also, behavioural activity offset and activity duration were each negatively correlated with the peak width of field rhythms displayed by the rostral SCN slice (electronic supplementary material, table S3; one-way ANOVA, p 0.001 in each case). In addition, the timing of activity offset correlated with the ZT peak time
Proc. R. Soc. B 282: 20150769
days
4 13 6 early intermediate late
(d) 15 days to re-entrain
slow
days
intermediate
a
6
0 (c) fast
b
c
rspb.royalsocietypublishing.org
time of activity offset
intermediate
days
(a) early
Downloaded from http://rspb.royalsocietypublishing.org/ on June 29, 2015
early rostral
300
5
middle
300
0
0
–300 60
80
100
caudal
150 0
–300 40
20 (b)
late
–150 20
40 60 80 time in vitro (h)
100
20
40
60
80
100
14
rspb.royalsocietypublishing.org
biolum (arb. units)
(a)
12 a
ab
b
a
ab
b
b a
ab
b a
10
0 early intermediate late
early intermediate late
early intermediate late
early intermediate late
peak width (h)
14
12
a
a ab
b
ab b
10
0
early intermediate late
early intermediate late
early intermediate late
early intermediate late
all
rostral
middle
caudal
Figure 3. Relationship between the time of activity offset in vivo and SCN function in vitro. (a) Representative time series illustrating PER2::LUC rhythms of SCN slices collected from mice that displayed an early time of activity offset (early) or a late time of activity offset (late). (b) Mice that differed in the timing of activity offset differed in the ZT peak time and peak width of PER2::LUC rhythms. As in figure 1g, letters above bars indicate the results of Tukey’s HSD, p , 0.02 (for sample sizes, cf. figure 2b).
of SCN neurons in each slice (electronic supplementary material, table S3, p 0.005 in each case). We further tested these relationships by comparing PER2::LUC rhythms across phenotypic groups (figure 2b). Mice with later times of activity offset displayed later peak times of PER2::LUC rhythms in each SCN slice (figure 3b; one-way ANOVA, p , 0.05). Similarly, longer activity duration was associated with later ZT peak times of PER2::LUC rhythms (data not shown). Further, mice with later times of behavioural activity offset displayed a shorter peak width of PER2::LUC rhythms in the rostral SCN slice and across slices (figure 3b; one-way ANOVA, p , 0.05), and again results were similar for activity duration (data not shown). Lastly, ZT peak time of SCN neurons differed among phenotypic groups, with later times of activity offset associated with later ZT peak times of SCN neurons in each slice (figure 4a; one-way ANOVA, p , 0.05).
(c) Relationship between the rate of re-entrainment and suprachiasmatic nucleus rhythms Individual differences in the rate of re-entrainment were correlated with changes in the function of the SCN at the neuronal level rather than at the whole slice level (electronic
supplementary material, table S3). Rate of re-entrainment was not associated with ZT peak time on the first cycle in vitro for the field rhythm of each SCN slice (electronic supplementary material, table S3; p . 0.15 in each case). On the other hand, rate of re-entrainment was positively correlated with degree of phase heterogeneity among SCN neurons (electronic supplementary material, table S3). When all cells were analysed together, the rate of behavioural re-entrainment was positively correlated with the standard deviation in peak time on the first cycle (electronic supplementary material, table S3; p ¼ 0.03) and the second cycle in vitro ( p ¼ 0.04). When the standard deviation of cellular peak time was assessed in each SCN slice separately, rate of behavioural re-entrainment was positively correlated with phase heterogeneity within the middle SCN on both the first cycle (electronic supplementary material, table S3; p ¼ 0.001) and second cycle in vitro ( p ¼ 0.046). Like results for timing of behavioural rhythms, the relationship between rate of re-entrainment and standard deviation in the peak time of SCN neurons was also detected across distinct behavioural groups (figure 4b). Relative to mice with a fast rate of re-entrainment, mice with a slower rate of re-entrainment displayed a greater degree of phase heterogeneity among SCN neurons in the middle and caudal SCN slice, as well as across slices (figure 4b).
Proc. R. Soc. B 282: 20150769
ZT peak time (h)
b
Downloaded from http://rspb.royalsocietypublishing.org/ on June 29, 2015
(a)
6
b 12 a
b a
b
b a
early intermediate late
early intermediate late
early intermediate late
1.5
standard deviation ZT peak time (h)
b 1.0
a
b
b a
ab
b a
b
0.5
0
fast intermediate slow
fast intermediate slow
fast intermediate slow
fast intermediate slow
all
rostral
middle
caudal
Figure 4. Relationship between circadian behaviour in vivo and PER2::LUC rhythms of SCN neurons in vitro. (a) Mice that differed in the time of activity offset differed in the ZT peak time of SCN neurons. (b) Mice that differed in the rate of re-entrainment differed in the heterogeneity of ZT peak time among SCN neurons. As in figure 1g, letters above bars indicate the results of Tukey’s HSD, p , 0.02 (for sample sizes, cf. figure 2d ).
(d) Relationship between behavioural rhythms and suprachiasmatic nucleus spatio-temporal organization In agreement with the results of a previous study [24], composite maps generated using slices collected from all the mice confirm there are pronounced regional phase differences both across and within rostrocaudal SCN slices (figure 5a). As found in previous detailed analyses [24], one of the most salient organizational features is a late-peaking node located within the rostral slice. Additionally, the middle SCN contained early-peaking regions within the dorsal-most and ventral-most regions and the caudal SCN appeared to be the most homogeneous in phase. Similar to analyses of SCN slices and all SCN neurons, phase analyses indicated that the time of activity offset was related to the timing of PER2::LUC expression (figure 5b,c). For nearly all SCN regions, mice with later times of activity offset showed later ZT peak times of PER2::LUC expression for SCN neurons in regions (full-factorial ANOVA, behaviour group: p , 0.0001, p , 0.02 and p , 0.0005 for rostral, middle and caudal SCN, respectively). Regional phase differences were evident within each slice (full-factorial ANOVA, SCN region: p , 0.05), but this did not interact with behaviour group (full-factorial ANOVA, behaviour region: p . 0.50). In fact, the only SCN region that did not differ among groups was the ventral region of the middle SCN (figure 5c). Overall, similar phase maps were found for groups of mice that differed in the rate of re-entrainment, but there were subtle features that distinguished these two groups (figure 5d– f ). First, fast and slow re-entrainers appeared to differ in the late-peaking region of the rostral SCN, with this region appearing larger in mice with fast re-entrainment (figure 5d).
Furthermore, the phase of peak PER2::LUC expression in the ventral regions within the middle slice and dorsal regions of the caudal slice also appeared to differ based on rate of reentrainment, with slow re-entrainers displaying an earlier phase (figure 5d). Relative to slices from mice that entrained quickly, mice that re-entrained slowly displayed a significantly earlier average phase for SCN neurons within the ventral region of middle SCN (figure 5e). Further, mice that re-entrained slowly displayed more deviation in the phase of SCN neurons within the central region of the middle SCN (figure 5f ).
4. Discussion Circadian rhythms are the product of a dynamic system of interacting genes and tissues. Here, we have used a topdown approach that exploits individual differences in circadian behaviour to gain insight into how the SCN regulates behavioural rhythms. We find a strong relationship between the peak time of SCN PER2::LUC expression and the timing of activity offset. We also find that individual differences in activity duration correspond to rhythmic parameters at the level of the whole slice, which suggests that the basis for this variability lies in the emergent properties of the network. Lastly, we find that the degree of SCN phase heterogeneity is associated with natural variation in the rate of reentrainment during simulated jetlag. Specifically, mice that re-entrained slowly were characterized by a wider distribution of phases among SCN neurons than mice that re-entrained faster. It remains possible that individual differences in circadian behaviour are also influenced by the function of other components of the circadian system (e.g. other clock genes, downstream oscillators). Nevertheless, this study indicates
Proc. R. Soc. B 282: 20150769
early intermediate late
b
b
a
b
10
0 (b)
b
rspb.royalsocietypublishing.org
ZT peak time (h)
14
Downloaded from http://rspb.royalsocietypublishing.org/ on June 29, 2015
rostral
caudal (b)
middle
2
13
7
5
3
8
6 8
11
early
late 14
*
*
12 11
*
* *
*
*
10
ZT peak time
late
9 0 1
2
3
4 5 6 SCN region
7
8
rspb.royalsocietypublishing.org
4
1
7
(c)
early
ZT peak time (h)
(a)
fast
ZT peak time (h)
slow
(f)
slow
fast
13 12 11 10 9
*
0 1
2
3
4 5 6 SCN region
slow
fast
2 standard deviation peak time (h)
(e)
1 * 0
7
8
1
2
3
4 5 6 SCN region
7
8
Figure 5. Relationship between circadian behaviour in vivo and SCN spatio-temporal organization in vitro. (a) Average phase maps generated using slices collected from all mice illustrating regional phase differences on the first cycle in vitro. Average phase maps were generated after superimposing data from both lobes. Numbers indicate regions used for regional analyses of SCN neuronal rhythms. (b) Average phase maps for mice in the early and late activity offset group. (c) Across most SCN regions, average ZT of peak PER2::LUC expression in SCN neurons is delayed in mice that display later times of activity offset. (d ) Average phase maps for mice in the fast and slow re-entrainment group. (e,f ) Rate of re-entrainment was associated with region-specific differences in PER2::LUC peak time and standard deviation of PER2::LUC peak time among SCN neurons. Asterisks (*) denote post hoc least square means contrasts, p , 0.02 (for sample sizes, cf. figure 2b,d ). (Online version in colour.)
that behavioural plasticity is regulated by SCN network properties, which highlights this as a potential target for interventions designed to modulate circadian behaviour. We found a robust relationship between the time of activity offset in vivo and the phase of SCN PER2::LUC expression in vitro, indicating that the entrained phase of the SCN relates to the timing of locomotor activity. While this result was expected, it represents, to our knowledge, the first direct evidence that individual differences in phase angle of entrainment are encoded by the SCN. Also, we found that individual variability in behavioural activity duration was related to properties evident at the population level, which is consistent with recent work investigating seasonal encoding within the SCN [29–34]. Of note, previous research also indicates a strong link between behavioural period and SCN function [35,36], but we were unable to address this issue here because our mice did not display marked variability in behavioural period. Our study adds to a growing body of research that suggests phase heterogeneity within the SCN affects the magnitude of photic responses. First, photic phase-resetting responses are attenuated by long-day photoperiods that cause phase dispersion among SCN neurons [37–39]. A complementary body of work likewise suggests that changes in SCN spatio-temporal organization can abolish photic resetting responses. Specifically, under 22 h light : dark cycles, rat locomotor rhythms dissociate into two components that are differentially controlled by the ventral SCN and the dorsal SCN [40,41]. This protocol allows investigation of circadian responses to light when the SCN is in a dissociated versus coherent state. When the rhythms of these two SCN regions are in phase with one another, photic stimuli elicit phase shifts and physiological correlates of photic responses [40].
When these two SCN compartments are out of phase, the transmission of photic input from the ventral to dorsal SCN is blocked and no behavioural response is elicited. Collectively, changes in photic resetting after exposure to seasonal and non-seasonal lighting conditions demonstrate that the capacity to respond to light is determined by the phase coherence among SCN neurons and the efficacy of regional communication. The current finding that greater SCN phase heterogeneity is associated with slower rate of re-entrainment may represent a less extreme form of this principle where phase coherence is reduced in some mice and re-entrainment takes several more cycles of exposure to the phase-resetting stimulus. Changes in phase coherence among SCN neurons may relate to decreased rate of re-entrainment in other contexts, like that typically observed with ageing [28]. Our study highlights that the bases of individual differences in rate of re-entrainment may relate to the function of specific regions in the SCN network. We find that the relationship between phase coherence and resetting speed was the strongest for the middle SCN slice, which is of interest as this slice contains several types of light-responsive cells [42–45]. Furthermore, the ventral and central regions of the middle SCN were those that differed most between fast and slow re-entrainers, with mice that re-entrained slowly displaying a ventral region with an earlier phase and a central region with more phase heterogeneity. That our analyses reveal an effect for these specific regions is of interest because the ventral region of the middle SCN contains VIP-ir cells (figure 1f ), which are important for the initial processing of photic input before propagation to the rest of the SCN network [46–48]. Likewise, the central region of the middle SCN has been proposed to act as a gate in the flow of information between the
Proc. R. Soc. B 282: 20150769
(d)
Downloaded from http://rspb.royalsocietypublishing.org/ on June 29, 2015
Together with previous experiments investigating individual differences in circadian behaviour [50,51], this study demonstrates the use of top-down analyses of brain function. Individual differences in behavioural, physiological and cellular processes affect numerous aspects of biological function, and thus, understanding the bases of this variation remains important. Future studies investigating the underlying genetic, epigenetic and molecular mechanisms that produce inter-individual variability in SCN function may lend insight into ways to manipulate plasticity of the circadian system under a wide variety of conditions relevant to human health and well-being.
Data accessibility. The raw data (.tif image sequences and actograms for all mice) have been uploaded to Dryad: http://datadryad.org/ review?doi=doi:10.5061/dryad.4rd1g. Author contributions. J.A.E. and A.J.D. designed the research. J.A.E. performed experiments and analysed data. J.A.E., T.L.L., O.C.C. and A.J.D. developed analytical tools. J.A.E., T.L.L., O.C.C. and A.J.D. drafted the manuscript. All authors gave final approval for publication.
Funding. This work was supported by NIH grant nos. U54NS060659, F32NS071935, and S21MD000101, the Georgia Research Alliance, and the NSF Center for Behavioral Neuroscience.
Competing interests. We declare we have no competing interests. Acknowledgements. We wish to thank Stanford Photonics for their equipment and assistance. Also, we are grateful to the Morehouse School of Medicine CLAR staff for providing excellent animal care.
References 1.
2.
3.
4.
5.
6.
7.
8.
Jakobsson M et al. 2008 Genotype, haplotype and copy-number variation in worldwide human populations. Nature 451, 998– 1003. (doi:10.1038/ nature06742) Frazer KA et al. 2007 A sequence-based variation map of 8.27 million SNPs in inbred mouse strains. Nature 448, 1050–1053. (doi:10.1038/nature06067) Ayroles JF et al. 2009 Systems genetics of complex traits in Drosophila melanogaster. Nat. Genet. 41, 299–307. (doi:10.1038/ng.332) Blakey JD, Hall IP. 2011 Current progress in pharmacogenetics. Br. J. Clin. Pharmacol. 71, 824–831. (doi:10.1111/j.1365-2125.2011.03912.x) Fanciulli M, Petretto E, Aitman TJ. 2010 Gene copy number variation and common human disease. Clin. Genet. 77, 201– 213. (doi:10.1111/j.1399-0004. 2009.01342.x) Vannberg FO, Chapman SJ, Hill AV. 2011 Human genetic susceptibility to intracellular pathogens. Immunol. Rev. 240, 105– 116. (doi:10.1111/j.1600065X.2010.00996.x) Bennett AF. 1987 Interindividual variability: an underutilized resource. In New directions in ecological physiology (eds ME Feder, AF Bennett, WW Burggren, RB Huey), pp. 147–169. New York, NY: Cambridge University Press. Gachon F, Nagoshi E, Brown SA, Ripperger J, Schibler U. 2004 The mammalian circadian timing system: from gene expression to physiology. Chromosoma 113, 103– 112. (doi:10.1007/s00412004-0296-2)
9.
10.
11.
12.
13.
14.
15.
16.
Weaver DR. 1998 The suprachiasmatic nucleus: a 25year retrospective. J. Biol. Rhythms 13, 100 –112. (doi:10.1177/074873098128999952) Welsh DK, Takahashi JS, Kay SA. 2010 Suprachiasmatic nucleus: cell autonomy and network properties. Annu. Rev. Physiol. 72, 551– 577. (doi:10.1146/annurev-physiol-021909-135919) Dibner C, Schibler U, Albrecht U. 2010 The mammalian circadian timing system: organization and coordination of central and peripheral clocks. Annu. Rev. Physiol. 72, 517 –549. (doi:10.1146/ annurev-physiol-021909-135821) Ko CH, Takahashi JS. 2006 Molecular components of the mammalian circadian clock. Hum. Mol. Genet. 15, R271 –R277. (doi:10.1093/hmg/ddl207) Roenneberg T, Wirz-Justice A, Merrow M. 2003 Life between clocks: daily temporal patterns of human chronotypes. J. Biol. Rhythms 18, 80 –90. (doi:10. 1177/0748730402239679) Katzenberg D, Young T, Finn L, Lin L, King DP, Takahashi JS, Mignot E. 1998 A CLOCK polymorphism associated with human diurnal preference. Sleep 21, 569–576. Archer J, Von Schantz M. 2003 A length polymorphism in the circadian clock gene Per3 is linked to delayed sleep phase syndrome and extreme diurnal preference. Sleep 26, 413–415. Gander PH, Myhre G, Graeber RC, Andersen HT, Lauber JK. 1989 Adjustment of sleep and the circadian temperature rhythm after flights across nine time zones. Aviat. Space Environ. Med. 60, 733–743.
17. Gamble KL et al. 2011 Shift work in nurses: contribution of phenotypes and genotypes to adaptation. PLoS ONE 6, e18395. (doi:10.1371/ journal.pone.0018395) 18. Martino TA et al. 2008 Circadian rhythm disorganization produces profound cardiovascular and renal disease in hamsters. Am. J. Physiol. Regul. Integr. Comp. Physiol. 294, R1675 –R1683. (doi:10. 1152/ajpregu.00829.2007) 19. Sjoholm LK, Backlund L, Cheteh EH, Ek IR, Frisen L, Schalling M, Osby U, Lavebratt C, Nikamo P. 2010 CRY2 is associated with rapid cycling in bipolar disorder patients. PLoS ONE 5, e12632. (doi:10. 1371/journal.pone.0012632) 20. Lee HJ, Rex KM, Nievergelt CM, Kelsoe JR, Kripke DF. 2011 Delayed sleep phase syndrome is related to seasonal affective disorder. J. Affect. Disord. 133, 573–579. (doi:10.1016/j.jad.2011.04.046) 21. Yoo SH et al. 2004 PERIOD2::LUCIFERASE real-time reporting of circadian dynamics reveals persistent circadian oscillations in mouse peripheral tissues. Proc. Natl Acad. Sci. USA 101, 5339–5346. (doi:10. 1073/pnas.0308709) 22. Zee PC, Rosenberg RS, Turek FW. 1992 Effects of aging on entrainment and rate of resynchronization of circadian locomotor activity. Am. J. Physiol. 263, R1099–R1103. 23. Davidson AJ, Sellix MT, Daniel J, Yamazaki S, Menaker M, Block GD. 2006 Chronic jet-lag increases mortality in aged mice. Curr. Biol. 16, R914 –R916. (doi:10.1016/j.cub.2006.09.058)
Proc. R. Soc. B 282: 20150769
Ethics. The experiments described were approved by the MSM Institutional Animal Care and Use Committee.
8
rspb.royalsocietypublishing.org
ventral SCN and other SCN regions [49]. Thus, the relative phase and the degree of heterogeneity within the ventral and central SCN probably modulate the processing of photic input and transmission to other SCN compartments. An important question concerns the functional role of regional phase differences in the SCN. Our work indicates that phase differences are a limiting factor in the response to photic manipulations. The fact that regional phase differences are a common feature of SCN organization could then be viewed as a barrier to the rapid adjustment to transmeridian travel and shift work. Under the ecological conditions experienced during evolution, the circadian system would not experience abrupt, large environmental changes, but rather gradual changes owing to seasonal variation. Because the environmental signals that synchronize the clock vary annually, the circadian system must be flexible enough to track the seasons and yet robust against minor perturbations produced by stochastic variation in light levels that can occur from day to day. As a mechanism that limits rapid resetting, SCN phase heterogeneity may be advantageous in that it provides both stability under static conditions and flexibility during environmental change. When viewed in this manner, the subgroup of fast re-entrainers described here may in fact be displaying the aberrant response. But further insight into the neural and genetic mechanisms that underlie this rapid response may be used to facilitate rapid resynchronization for travellers and shift workers.
Downloaded from http://rspb.royalsocietypublishing.org/ on June 29, 2015
34.
35.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
J. Neurosci. 24, 68 –75. (doi:10.1523/JNEUROSCI. 1666-03.2004) Dardente H, Poirel VJ, Klosen P, Pevet P, MassonPevet M. 2002 Per and neuropeptide expression in the rat suprachiasmatic nuclei: compartmentalization and differential cellular induction by light. Brain Res. 958, 261– 271. (doi:10.1016/S0006-8993(02)03563-1) Romijn HJ, Sluiter AA, Pool CW, Wortel J, Buijs RM. 1996 Differences in colocalization between Fos and PHI, GRP, VIP and VP in neurons of the rat suprachiasmatic nucleus after a light stimulus during the phase delay versus the phase advance period of the night. J. Comp. Neurol. 372, 1–8. (doi:10.1002/(SICI)1096-9861(19960812)372:1,1:: AID-CNE1.3.0.CO;2-7) Hamada T, Antle MC, Silver R. 2004 Temporal and spatial expression patterns of canonical clock genes and clock-controlled genes in the suprachiasmatic nucleus. Eur. J. Neurosci. 19, 1741 –1748. (doi:10. 1111/j.1460-9568.2004.03275.x) Shigeyoshi Y et al. 1997 Light-induced resetting of a mammalian circadian clock is associated with rapid induction of the mPer1 transcript. Cell 91, 1043 –1053. (doi:10.1016/S00928674(00)80494-8) Yan L, Silver R. 2002 Differential induction and localization of mPer1 and mPer2 during advancing and delaying phase shifts. Eur. J. Neurosci. 16, 1531–1540. (doi:10.1046/j.1460-9568.2002.02224.x) Antle MC, Foley DK, Foley NC, Silver R. 2003 Gates and oscillators: a network model of the brain clock. J. Biol. Rhythms 18, 339–350. (doi:10.1177/ 0748730403253840) Margraf RR, Zlomanczuk P, Liskin LA, Lynch GR. 1991 Circadian differences in neuronal activity of the suprachiasmatic nucleus in brain slices prepared from photo-responsive and photonon-responsive Djungarian hamsters. Brain Res. 544, 42 – 48. (doi:10.1016/00068993(91)90883-W) Ciarleglio CM, Gamble KL, Axley JC, Strauss BR, Cohen JY, Colwell CS, McMahon DG. 2009 Population encoding by circadian clock neurons organizes circadian behavior. J. Neurosci. 29, 1670 – 1676. (doi:10.1523/JNEUROSCI.3801-08. 2009)
9
Proc. R. Soc. B 282: 20150769
36.
suprachiasmatic nuclei neurons and their response to photoperiod. J. Biol. Rhythms 24, 44 –54. (doi:10.1177/0748730408327918) Hazlerigg DG, Ebling FJ, Johnston JD. 2005 Photoperiod differentially regulates gene expression rhythms in the rostral and caudal SCN. Curr. Biol. 15, R449 –R450. (doi:10.1016/j.cub.2005.06.010) Low-Zeddies SS, Takahashi JS. 2001 Chimera analysis of the Clock mutation in mice shows that complex cellular integration determines circadian behavior. Cell 105, 25 –42. (doi:10.1016/S00928674(01)00294-X) Ralph MR, Hurd MW. 1995 Circadian pacemakers in vertebrates. Ciba Found. Symp. 183, 67 –81; discussion 81 –67. vanderLeest HT, Rohling JH, Michel S, Meijer JH. 2009 Phase shifting capacity of the circadian pacemaker determined by the SCN neuronal network organization. PLoS ONE 4, e4976. (doi:10. 1371/journal.pone.0004976) Refinetti R. 2002 Compression and expansion of circadian rhythm in mice under long and short photoperiods. Integr. Physiol. Behav. Sci. 37, 114 – 127. (doi:10.1007/BF02688824) Pittendrigh CS, Elliott JA, Takamura T. 1984 The circadian component in photoperiodic induction. Ciba Found. Symp. 104, 26– 47. (doi:10.1002/ 9780470720851.ch4) Schwartz MD, Congdon S, de la Iglesia HO. 2010 Phase misalignment between suprachiasmatic neuronal oscillators impairs photic behavioral phase shifts but not photic induction of gene expression. J. Neurosci. 30, 13 150 –13 156. (doi:10.1523/ JNEUROSCI.1853-10.2010) de la Iglesia HO, Cambras T, Schwartz WJ, DiezNoguera A. 2004 Forced desynchronization of dual circadian oscillators within the rat suprachiasmatic nucleus. Curr. Biol. 14, 796 –800. (doi:10.1016/j. cub.2004.04.034) Kuhlman SJ, Silver R, Le Sauter J, Bult-Ito A, McMahon DG. 2003 Phase resetting light pulses induce Per1 and persistent spike activity in a subpopulation of biological clock neurons. J. Neurosci 23, 1441–1450. Karatsoreos IN, Yan L, LeSauter J, Silver R. 2004 Phenotype matters: identification of light-responsive cells in the mouse suprachiasmatic nucleus.
rspb.royalsocietypublishing.org
24. Evans JA, Leise TL, Castanon-Cervantes O, Davidson AJ. 2011 Intrinsic regulation of spatiotemporal organization within the suprachiasmatic nucleus. PLoS ONE 6, e15869. (doi:10.1371/journal.pone.0015869) 25. Yoshikawa T, Yamazaki S, Menaker M. 2005 Effects of preparation time on phase of cultured tissues reveal complexity of circadian organization. J. Biol. Rhythms 20, 500–512. (doi:10.1177/074873040528 0775) 26. Davidson AJ, Castanon-Cervantes O, Leise TL, Molyneux PC, Harrington ME. 2009 Visualizing jet lag in the mouse suprachiasmatic nucleus and peripheral circadian timing system. Eur. J. Neurosci. 29, 171–180. (doi:10.1111/j.1460-9568.2008. 06534.x) 27. Evans JA, Elliott JA, Gorman MR. 2009 Dim nighttime illumination accelerates adjustment to timezone travel in an animal model. Curr. Biol. 19, R156 –R157. (doi:10.1016/j.cub.2009.01.023) 28. Sellix MT, Evans JA, Leise TL, Castanon-Cervantes O, Hill DD, DeLisser P, Block GD, Menaker M, Davidson AJ. 2012 Aging differentially affects the reentrainment response of central and peripheral circadian oscillators. J. Neurosci. 32, 16 193 – 16 202. (doi:10.1523/JNEUROSCI.3559-12.2012) 29. Meijer JH, Michel S, vanderLeest HT, Rohling JH. 2010 Daily and seasonal adaptation of the circadian clock requires plasticity of the SCN neuronal network. Eur. J. Neurosci. 32, 2143 –2151. (doi:10. 1111/j.1460-9568.2010.07522.x) 30. Inagaki N, Honma S, Ono D, Tanahashi Y, Honma K. 2007 Separate oscillating cell groups in mouse suprachiasmatic nucleus couple photoperiodically to the onset and end of daily activity. Proc. Natl Acad. Sci. USA 104, 7664 –7669. (doi:10.1073/pnas. 0607713104) 31. Naito E, Watanabe T, Tei H, Yoshimura T, Ebihara S. 2008 Reorganization of the suprachiasmatic nucleus coding for day length. J. Biol. Rhythms 23, 140–149. (doi:10.1177/0748730408314572) 32. Schaap J, Albus H, vanderLeest HT, Eilers PH, Detari L, Meijer JH. 2003 Heterogeneity of rhythmic suprachiasmatic nucleus neurons: implications for circadian waveform and photoperiodic encoding. Proc. Natl Acad. Sci. USA 100, 15 994 –15 999. (doi:10.1073/pnas.2436298100) 33. Brown TM, Piggins HD. 2009 Spatiotemporal heterogeneity in the electrical activity of
