Two coupled circadian oscillations regulate Bmal1-ELuc and Per2-SLR2 expression in the mouse suprachiasmatic nucleus

Abstract

Circadian rhythms in clock genes, Bmal1 and Per2 expression were monitored simultaneously in the cultured slice of mouse suprachiasmatic nucleus (SCN) by dual bioluminescent reporters. In the neonatal SCN, the phase-relation between the Bmal1 and Per2 rhythms were significantly changed during culture. Medium exchange produced phase-dependent phase shifts (PRCm) in the Bmal1 rhythms, but not in the Per2 rhythms. As a result, the two circadian rhythms were temporally dissociated after medium exchange. In the adult SCN, the phase-relation between the two rhythms was kept constant during culture at least up to 20 cycles. The amplitude of PRCm in the adult SCN was significantly attenuated in the Bmal1 rhythm, whereas a PRCm was developed in the Per2 rhythm. The circadian period was not systematically affected by medium exchange in either of rhythms, regardless of whether it was in the neonatal or the adult SCN. Tetrodotoxin, a sodium channel blocker, enhanced the phase-response in both rhythms but abolished the phase-dependency. In addition, tetrodotoxin lengthened the circadian period independent of the phase of administration. Thus, the Bmal1 and Per2 rhythms in the SCN are dissociable and likely regulated by distinct circadian oscillators. Bmal1 is the component of a Bmal1/REV-ERBa/ROR loop and Per2 a Per/Cry/BMAL1/CLOCK loop. Both loops could be molecular mechanisms of the two circadian oscillators that are coupled through the protein product of Bmal1. The coupling strength between the two oscillations depends on developmental stages.

Figure 1

Circadian rhythms in the neonatal SCN slices of Bmal1-Eluc:Per2-SLR2 mice (a) Sequential records of circadian Bmal1 (green) and Per2 (red) rhythm in a neonatal SCN slice. Original records were smoothed by a 5 point (80 minutes) moving average method and detrended by a 24 hour moving subtraction method. The abscissa indicates the day in culture and the ordinate the relative intensity of bioluminescence (right, Bmal1; left, Per2). Day 0 means the day of slice preparation. (b,c) A double plotted circadian Bmal1 (b) and Per2 (c) rhythm in a raster format. Bioluminescence data of each day were indicated as a percentile to the daily mean. (d) Successive plots of circadian peak phases (Bmal1, green; Per2, red) and regressions lines fitted to them in the same SCN slice as used in (a). (e) Circadian periods of individual Bmal1 and Per2 rhythms (n = 6) calculated from a regression line. A horizontal bar indicates the mean circadian period. Circadian periods of the rhythms from the same SCN are indicated with the same colors. (f) Chi-square periodogram demonstrates significantly different circadian periods of Bmal1 (green) and Per2 (red) rhythm. A black oblique line in the graph indicates the significance level at p < 0.05. (g) Phase difference (ψ) between the circadian Bmal1 and Per2 peaks in each cycle. The values are indicated as the mean and SD. (h) The mean and SD of ψ in the first and the last 10 cycles is indicated. **p < 0.01 (paired t-test).

Figure 2

Circadian rhythms of the adult SCN slices of Bmal1-Eluc:Per2-SLR2 mice (a) Sequential records of circadian Bmal1 (green) and Per2 (red) rhythm in an adult SCN slice. See also the legend for Fig. 1a. (b,c) A double plotted circadian Bmal1 (green) and Per2 (red) rhythm in a raster format. (d) Successive plots of circadian peak phases (Bmal1, green; Per2, red) and regressions lines fitted to them in the same SCN slice as used in (a). (e) Circadian periods of Bmal1 and Per2 rhythms in the cultured SCN (n = 8) calculated from a regression line. See also the legend for Fig. 1e. (f) Chi-square periodogram demonstrates significant but different circadian periods of Bmal1 (green) and Per2 (red) rhythm. See also the legend for Fig. 1f. (g) A phase difference (ψ) between the circadian Bmal1 and Per2 peaks in each cycle. See also the legend for Fig. 1g. (h) The mean and SD of ψ in the first and the last 10 cycles is indicated.

Figure 3

Phase shifts of circadian Bmal1 and Per2 rhythm in the neonatal SCN slices in response to medium exchange (a) Representative records of phase shifts in response to medium exchange at 4 different phases. The circadian rhythms before and after medium exchange were illustrated for Bmal1 (green) and Per2 (red). See also the legend for Fig. 1a. The time of medium exchange was indicated with a vertical arrow. (b) Successive circadian peaks with regression lines fitted to them before and after medium exchange in the same SCN as used for (a). The abscissa indicates times of day as ZT and the ordinate days in culture. A square column in each panel indicates the time of medium exchange. A horizontal arrow indicates the direction and magnitude of phase-shifts calculated from two regression lines (see text).

Figure 4

Phase-response curve and period-response curve for medium exchange of circadian Bmal1 and Per2 rhythms in the neonatal SCN (a) Phase-responses in individual slices (n = 22) of circadian Bmal1 (green circle) and Per2 (red circle) rhythms in the neonatal SCN (upper). The abscissa indicates the circadian phase in CT, where CT1 is the peak phase of Bmal1 rhythm and the ordinate indicates phase-shifts in hours with a positive sign for phase-advance shifts and with a negative for phase-delay. Mean phase-shifts with SD at CT0–CT6 (n = 5), CT6–CT12 (n = 9), CT12–CT18 (n = 4) and CT18–CT20 (n = 5) (lower). A significant phase-shift is indicated with asterisk (*) and a significant difference between the phase-shifts of two circadian rhythms is shown with dagger (^†). **p < 0.01 (one-way repeated measure ANOVA), ^‡p < 0.01 (two-way repeated measure ANOVA with post hoc Tukey-Kramer test). (b) Period-responses obtained from the same experiment as (a). Period-responses in individual slices of circadian Bmal1 (green circle) and Per2 (red circle) (upper), and the mean responses with SD at the 4 circadian phases (lower).

Figure 5

Phase shifts of circadian Bmal1 and Per2 rhythm in the adult SCN slices in response to medium exchange and effects of TTX pretreatment (a) Representative records of phase shifts in response to medium exchange in the adult SCN pretreated with distilled water (upper, upper most) and with TTX (lower, lower most). See also the legend for Fig. 3a. (b) Successive circadian peaks with regression lines fitted to them before and after medium exchange in the same SCN as used for (a). See also the legend for Fig. 3b.

Figure 6

Phase-response curve and period-response curve for medium exchange of circadian Bmal1 and Per2 rhythms in the adult SCN (a) Phase-responses in individual slices (n = 20) of circadian Bmal1 (green circle) and Per2 (red circle) rhythms in the adult SCN (upper). The SCN was pretreated with distilled water (see text). Mean phase-shifts with SD at CT0–CT6 (n = 5), CT6–CT12 (n = 4), CT12–CT18 (n = 5) and CT18–CT20 (n = 6) (lower). See also the legend for Fig. 4a. A significant phase-shift is indicated with asterisk (*). *p < 0.05, **p < 0.01, and a significant difference in the response between the two circadian rhythms is shown with dagger (^†). ^†p < 0.05. (b) Period-responses obtained from the same experiment as (a). Period-responses in individual slices of circadian Bmal1 (green circle) and Per2 (red circle) (upper), and the mean responses with SD at the 4 circadian phases (lower).

Figure 7

Phase-response curve and period-response curve for medium exchange in the adult SCN pretreated with TTX (a). Phase-responses (n = 19) of circadian Bmal1 (green circle) and Per2 (red circle) rhythms in the adult SCN (upper panel). The SCN was pretreated with TTX (see text). Mean phase-shifts with SD at CT0–CT6 (n = 5), CT6–CT12 (n = 4), CT12–CT18 (n = 5) and CT18–CT20 (n = 5) (lower panel). See also the legend for Fig. 4a. (b) Period-responses obtained from the same experiment as (a). Period-responses in individual slices of circadian Bmal1 (green circle) and Per2 (red circle) (upper), and the mean responses with SD at the 4 circadian phases (lower). A significant period-response is indicated with asterisk (*) and a significant difference in the response between the two circadian rhythms is shown with dagger (^†). *p < 0.05, ^†p < 0.05.