Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 Jan 2;13(1):e0190004.
doi: 10.1371/journal.pone.0190004. eCollection 2018.

Quantitative Analysis of Circadian Single Cell Oscillations in Response to Temperature

Free PMC article

Quantitative Analysis of Circadian Single Cell Oscillations in Response to Temperature

Ute Abraham et al. PLoS One. .
Free PMC article


Body temperature rhythms synchronize circadian oscillations in different tissues, depending on the degree of cellular coupling: the responsiveness to temperature is higher when single circadian oscillators are uncoupled. So far, the role of coupling in temperature responsiveness has only been studied in organotypic tissue slices of the central circadian pacemaker, because it has been assumed that peripheral target organs behave like uncoupled multicellular oscillators. Since recent studies indicate that some peripheral tissues may exhibit cellular coupling as well, we asked whether peripheral network dynamics also influence temperature responsiveness. Using a novel technique for long-term, high-resolution bioluminescence imaging of primary cultured cells, exposed to repeated temperature cycles, we were able to quantitatively measure period, phase, and amplitude of central (suprachiasmatic nuclei neuron dispersals) and peripheral (mouse ear fibroblasts) single cell oscillations in response to temperature. Employing temperature cycles of different lengths, and different cell densities, we found that some circadian characteristics appear cell-autonomous, e.g. period responses, while others seem to depend on the quality/degree of cellular communication, e.g. phase relationships, robustness of the oscillation, and amplitude. Overall, our findings indicate a strong dependence on the cell's ability for intercellular communication, which is not only true for neuronal pacemakers, but, importantly, also for cells in peripheral tissues. Hence, they stress the importance of comparative studies that evaluate the degree of coupling in a given tissue, before it may be used effectively as a target for meaningful circadian manipulation.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.


Fig 1
Fig 1. Circadian single cell bioluminescence of primary mouse fibroblasts follows external temperature cycles.
A: Representative examples of PER2::LUC bioluminescence (gray values) emitted from single mouse ear fibroblasts in confluent cell cultures exposed to six repeats of 10h of 33°C (cold): 10h 37°C (warm) (T20, left) or 12h cold: 12h warm (T24, right), respectively. The periods of the oscillations were determined by cosine fit Before (blue), During (red), and After (green) the temperature cycles. Dotted horizontal line: mean magnitude of bioluminescence. B: Double-plotted peaks (dots) of the cells shown in A) reveal that oscillations assume a semi-stable phase relationship with the cold phase. Light blue: cold phase.
Fig 2
Fig 2. Temperature cycles affect single fibroblast periods and phase relationships.
A: Scatterplot showing the distribution of periods displayed by single cells exposed to T20 (open circles) and T24 (closed circles) temperature cycles Before, During, and After the temperature paradigm. Each dot represents a period measurement of a single cell; the mean is marked by a horizontal line. Only cells that display circadian periods in all three conditions were included in the analysis. A two-way repeated measures ANOVA revealed that both treatment (T-cycle) and time (phases Before, During, and After,) and their interaction have significant effects on average periods (p = 0.013, p = 0.03, and p = 0.0008, respectively). Mean periods of cells exposed to T20 are significantly different between During and Before/After the temperature cycles (Two-Way-ANOVA for repeated measures, Bonferroni posttests, ** = p<0.001, n = 17), while periods in T24 do not change in response to temperature (Two-Way-ANOVA for repeated measures, Bonferroni posttests, p>0.05, n = 16). B: Individual period developments of single cells in constant temperature conditions (Before) and During temperature cycles in T20 (open circles) and T24 (closed circles) conditions. Each dot represents a period measurement of a single cell; repeated measures are connected by a black line. The plot shows that the broad period distribution During T24 is not a result of cellular periods being insensitive to the temperature paradigm, but rather due to very variable individual period responses. This is in contrast to T20 where the majority of cells respond with period shortening. C: Polar plots depicting single cell PER2 circadian peak phases Before (last phase before temperature cycles), During (mean of the last two phases during), and After (first phase after) the temperature cycles for T20 (open circles, n = 17) and T24 (closed circles, n = 16). Each dot represents the peak phase of a single cell measured in the respective condition. The direction of the orange arrow denotes the mean phase. The phase population clusters significantly when the arrow crosses the corresponding circle (Rayleigh's uniformity test, p<0.05). The blue shaded areas of the polar plots During denote the cold phases. Mean circadian phases are significantly different between Before and During/After T20 (Watson-Williams-F-Test, *** = p<0.0001). D: Left: The relationship of the PER2 peak phase with the onset of the cold phase (= mean of the phase relationships with the last two cold phases) is not significantly dependent on the single cell's intrinsic period (τ Before) (Linear regression, r2 = 0.08, p>0.05, n = 17). Only cells subjected to T20 temperature cycles were included in the analysis, provided that they exhibited a τ During that was close to (+/-1h) the external temperature cycle. Right: two-variable polar plot showing the frequency distribution of PER2 peak phases depending on the intrinsic period (τ Before) of the cells plotted in the left panel of D (n = 13). The radius of a wedge (concentric intervals: 2) represents the number of cells displaying a circadian peak phase within a 2h-bin (for a description of circadian phase determination see p.7, section d)), while color-coded segments denote the intrinsic periods (τ Before). There is no clear relationship between intrinsic period and phase, which supports preliminary conclusions from our linear analysis to the left.
Fig 3
Fig 3. Individual fibroblasts respond to external stimuli roughly as predicted from tissues and whole organisms.
A: Representative examples of a phase delay (negative shift, left) and a phase advance (positive shift, right) of single cell PER2::LUC bioluminescence (gray values) in response to a 10h cold phase hitting the oscillation at CT14 (left) and CT0 (right). The peak of PER2::LUC in fibroblasts was defined as CT18, in reference to [26]. Dotted horizontal line: mean magnitude of bioluminescence. B: The phase responses reveal that cells which have been hit by the cold phase between CT6 and CT18 mostly respond with a delay phase shift (mean: -3.9±1.0h SEM, n = 20). At other circadian times (CT 18 to CT6) cells respond with a phase advance (mean: 2.6±0.7h SEM, n = 7). Mean phase shifts within these two circadian ranges differ significantly (t-test, p<0.001). Dots: phase shifts of single cells; line: fourth order polynomial fit. C: The amplitude response curve (ARC) depicts the relative amplitude response (fold change) as a function of the circadian time of the onset of the 10-h temperature pulse. In contrast to the phase responses (B), there is no significant difference in the relative amplitude responses between CT6 to CT18 and CT18 to CT6 (Mann Whitney test, p>0.05, n = 20). Dots: relative amplitude changes of single cells in response to a cold stimulus; line: fourth order polynomial fit.
Fig 4
Fig 4. The degree of intercellular communication determines the impact of temperature cycles on single SCN cells.
A: PER2::LUC SCN dispersals were cultured at high (220 neurons/mm2, top left) and medium density (40 neurons/mm2, top right), and bioluminescence (gray values) from single cells monitored over the course of several days. Scale bar: 500μm. On day 3 or 5, cells were exposed to a T20 temperature cycle with deltaT = 2°C. Four representative single cell traces of a dense (left panel) and a medium dense culture (right panel) reveal that oscillations appear more robust in dense cultures. Light blue: cold phase (35°C); horizontal line: mean magnitude of bioluminescence. B: Peak bioluminescences (dots) from two representative single neurons in a dense and a medium dense SCN dispersal culture (same cells as top panels in A) depending on the time of day. Single cell SCN oscillations in the dense culture are clearly less perturbed by a temperature cycle (peak bioluminescence "free-runs" through the temperature cycle) compared to the medium dense culture (cells tend to adjust their period to the external cycle). Light blue: cold phase. Data were double plotted for a better visualization. C: Polar plots depicting single neuron PER2 circadian peak phases repeatedly measured Before (last peak before temperature cycles) and During (fourth peak during temperature cycles) in a dense (n = 25) and a medium dense (n = 8) SCN dispersal culture. Each dot represents the peak phase of a single cell measured in the respective condition. The direction of the orange arrow denotes the mean phase. The phase population clusters significantly when the arrow crosses the corresponding circle (Rayleigh's uniformity test, p<0.05). In the dense culture, peak phases cluster significantly around CT2-4, independent of whether a temperature signal is present or not. In contrast, phasing in the medium dense culture appears rather random Before, but shows a phase concentration During, with the mean peak phase being pushed to the warm phase. D: The relative amplitudes of single cell oscillations in the dense culture (n = 25) are significantly higher than those in the medium dense culture (n = 8), independent of the external signal (One-Way ANOVA with Tukey's multiple comparison posthoc tests; *** = p<0.0001). Relative amplitudes in a dense culture significantly decrease upon exposure to a temperature signal (* = p<0.01). Displayed are boxplots with min/max.

Similar articles

See all similar articles

Cited by 2 articles


    1. Reppert SM, Weaver DR. Coordination of circadian timing in mammals. Nature 2002. August 29;418(6901):935–41. doi: 10.1038/nature00965 - DOI - PubMed
    1. Welsh DK, Yoo SH, Liu AC, Takahashi JS, Kay SA. Bioluminescence imaging of individual fibroblasts reveals persistent, independently phased circadian rhythms of clock gene expression. Curr Biol 2004. December 29;14(24):2289–95. doi: 10.1016/j.cub.2004.11.057 - DOI - PMC - PubMed
    1. Nagoshi E, Saini C, Bauer C, Laroche T, Naef F, Schibler U. Circadian Gene Expression in Individual Fibroblasts; Cell-Autonomous and Self-Sustained Oscillators Pass Time to Daughter Cells. Cell 2004. November 24;119(5):693–705. doi: 10.1016/j.cell.2004.11.015 - DOI - PubMed
    1. Leise TL, Wang CW, Gitis PJ, Welsh DK. Persistent cell-autonomous circadian oscillations in fibroblasts revealed by six-week single-cell imaging of PER2::LUC bioluminescence. PLoS One 2012;7(3):e33334 doi: 10.1371/journal.pone.0033334 - DOI - PMC - PubMed
    1. Tahara Y, Kuroda H, Saito K, Nakajima Y, Kubo Y, Ohnishi N, et al. In vivo monitoring of peripheral circadian clocks in the mouse. Curr Biol 2012. June 5;22(11):1029–34. doi: 10.1016/j.cub.2012.04.009 - DOI - PubMed

Publication types

Grant support

The present study was supported by the Federal Ministry of Education and Research (BMBF) [grant number 01GQ1001C to H.H. and A.K.]. Work in A.K.'s laboratory is further supported by the German Science Foundation (DFG). U.A.'s salary was paid by the Federal Ministry of Education and Research [grant number 01GQ1001C] and by a research stipend granted by the Charité-Universitätsmedizin Berlin. J.K.S.'s salary was paid by the Federal Ministry of Education and Research [grant number 01GQ1001C]. The XR/Mega-10Z ICCD camera was funded by the German Science Foundation (DFG) [grant AB139/4-1 to U.A.]. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

LinkOut - more resources