Entrainment of disrupted circadian behavior through inhibition of casein kinase 1 (CK1) enzymes

Abstract

Circadian pacemaking requires the orderly synthesis, posttranslational modification, and degradation of clock proteins. In mammals, mutations in casein kinase 1 (CK1) epsilon or delta can alter the circadian period, but the particular functions of the WT isoforms within the pacemaker remain unclear. We selectively targeted WT CK1epsilon and CK1delta using pharmacological inhibitors (PF-4800567 and PF-670462, respectively) alongside genetic knockout and knockdown to reveal that CK1 activity is essential to molecular pacemaking. Moreover, CK1delta is the principal regulator of the clock period: pharmacological inhibition of CK1delta, but not CK1epsilon, significantly lengthened circadian rhythms in locomotor activity in vivo and molecular oscillations in the suprachiasmatic nucleus (SCN) and peripheral tissue slices in vitro. Period lengthening mediated by CK1delta inhibition was accompanied by nuclear retention of PER2 protein both in vitro and in vivo. Furthermore, phase mapping of the molecular clockwork in vitro showed that PF-670462 treatment lengthened the period in a phase-specific manner, selectively extending the duration of PER2-mediated transcriptional feedback. These findings suggested that CK1delta inhibition might be effective in increasing the amplitude and synchronization of disrupted circadian oscillators. This was tested using arrhythmic SCN slices derived from Vipr2(-/-) mice, in which PF-670462 treatment transiently restored robust circadian rhythms of PER2::Luc bioluminescence. Moreover, in mice rendered behaviorally arrhythmic by the Vipr2(-/-) mutation or by constant light, daily treatment with PF-670462 elicited robust 24-h activity cycles that persisted throughout treatment. Accordingly, selective pharmacological targeting of the endogenous circadian regulator CK1delta offers an avenue for therapeutic modulation of perturbed circadian behavior.

Fig. 1.

Differential effects of selective inhibition of CK1ε and CK1δ on the circadian period of fibroblasts and SCN in vitro. (A) Group data (mean ± SEM) reveal dose-dependent period lengthening by CK1 inhibition in WT, Ck1ε^−/−, and Ck1ε^tau derived fibroblasts. Red, WT; blue, CK1ε^tau; green, Ck1ε^−/−. (B and C) Data in A plotted as fold over whole cell IC₅₀ of CK1δ (upper axis) or CK1ε (lower axis) for PF-670462 or PF-4800567. Note even when it is 10-fold over the IC₅₀ of CK1ε, PF-4800567 is ineffective in lengthening the circadian period. (D) Representative recordings of PER2::Luc bioluminescence from WT SCN treated with CK1 inhibitors (arrow,1 μM). (E) Group data reveal interaction between CK1ε genotype and CK1 inhibitor (1 μM) in slowing SCN pacemaker. Whereas PF-670462 lengthened period in all three genotypes, PF-4800567 was only effective against CK1ε τ mutant (mean ± SEM, *P < 0.05, ***P <0.001, n = 3–5). (F) Group data reveal dose-dependent prolongation of WT SCN circadian period with PF-670462.

Fig. 2.

Differential effects of selective inhibition of CK1ε and CK1δ on the circadian period of locomotor activity in vivo. (A) Representative wheel-running actograms of Ck1ε^+/+, Ck1ε^−/−, and Ck1ε^tau mice treated with vehicle (Left), PF-670462 (Center), or PF-4800567 (Right) reveal no effect of daily vehicle injection but significant period lengthening by PF-670462 in both WT, Ck1ε^−/−, and Ck1ε^tau mutant mice. In contrast, selective inhibition of CK1ε by PF-4800567 is only effective in gain-of-function Ck1ε^tau animals. (B) Actograms from Ck1ε^tau mice plotted on 21-h time base for ease of inspection. (C) Group data reveal interaction between CK1 inhibition and genotype in setting circadian period in vivo (mean ± SEM, *P < 0.05, **P <0.01 vs. vehicle, two-way ANOVA with Bonferonni's post hoc test) (n = 6–8).

Fig. 3.

CK1δ inhibitor increases nuclear retention of PER2 in SCN neurons in vitro and in vivo. (A) Representative recording of PER2::Luc signal from SCN slice treated with PF-670462 from the start of the experiment. Note the ultimate loss of circadian oscillation and sustained expression of PER2::Luc bioluminescence. (Insets) The bioluminescence images of the same slice by CCD camera. Post hoc immunostaining reveals nuclear localization of PER2 (green). Nuclei were counterstained by DAPI (blue). (B) Immunostaining for PER2 in SCN of WT mice sampled at ZT12 or injected with either vehicle or PF-670462 at ZT12 and then sampled at ZT19. Note retention of nuclear PER2 in SCN of drug-treated mouse. Black triangle marks time of sampling. *Marks time of drug/vehicle injection.

Fig. 4.

Inhibition of CK1 by PF-670462 restructures the internal phase relationships of the individual clock components. (A–C) Representative bioluminescence recordings of Per2, Bmal1, and Rev-erbα promoter activity in Rat-1 fibroblasts reveal reversible, dose-dependent period lengthening and ultimate arrhythmia induced by high-dose PF-670462. Note, black horizontal bar indicates exposure of cells to the inhibitor. (D) Group data reveal equivalent PF-670462-induced (1 μM) period lengthening in Rat-1 cells when reported by Per2, Bmal1, or Rev-erbα promoter constructs (mean ± SEM, ***<0.001 vs. vehicle). (E). Schematic plot of internal phase relationships (X and Y) of the various components of the oscillator (Upper) as reported by Per2, Bmal1, and Rev-erbα promoter sequences exposed to vehicle or 1 μM PF-670462. (F) Group data reveal selective lengthening of the X phase, but not the Y phase, by PF-670462 at 1 μM (mean ± SEM, ***<0.001 vs. vehicle).

Fig. 5.

Activity rhythms compromised by environmental or genetic perturbation of the SCN can be restored by daily CK1 inhibition in vivo. (A) Representative recordings (Left) of SCN slices derived from Vipr2^−/− mice that exhibited damped and disorganized oscillations of the PER2::Luc reporter. Treatment with PF-670462 (arrow; 1 μM) but not vehicle resulted in an immediate induction of high-amplitude and long-period bioluminescent oscillation (**P < 0.01). Group data (Right) (mean ± SEM) reveal significant increase in amplitude of oscillation with PF-670462. (B) Representative traces of Vipr2^−/− mice housed in DD and exhibiting arrhythmic or weakly rhythmic short period circadian activity cycles (n = 5). Daily vehicle treatment (Left) had no entraining action, whereas daily injections of PF-670462 (Right) (30 mg/kg) induced robust entrainment. (C) Representative actograms show behavioral activity of WT mice housed under LL (n = 8). Activity patterns of WT mice treated with vehicle (Left) deteriorated rapidly. In contrast, daily injection of PF-670462 (30 mg/kg) produced stable entrainment of locomotor activity in LL with activity onset ∼13 h after time of injection.