Reversible modulation of circadian time with chronophotopharmacology

Abstract

The circadian clock controls daily rhythms of physiological processes. The presence of the clock mechanism throughout the body is hampering its local regulation by small molecules. A photoresponsive clock modulator would enable precise and reversible regulation of circadian rhythms using light as a bio-orthogonal external stimulus. Here we show, through judicious molecular design and state-of-the-art photopharmacological tools, the development of a visible light-responsive inhibitor of casein kinase I (CKI) that controls the period and phase of cellular and tissue circadian rhythms in a reversible manner. The dark isomer of photoswitchable inhibitor 9 exhibits almost identical affinity towards the CKIα and CKIδ isoforms, while upon irradiation it becomes more selective towards CKIδ, revealing the higher importance of CKIδ in the period regulation. Our studies enable long-term regulation of CKI activity in cells for multiple days and show the reversible modulation of circadian rhythms with a several hour period and phase change through chronophotopharmacology.

Fig. 1. The concept of the reversible chronophotopharmacology.

a Schematic representation of the circadian clock transcription-translation autoregulatory feedback loop, and its photo-modulation by photoswitchable inhibitor of CKI. Both isomers of the photoswitchable bioactive molecule show inhibition of CKI, but the affinity can be modulated with light, with one photoisomer (shown in red) being a stronger inhibitor than the other (blue). b Key parameters of the photoswitchable circadian modulators (2), based on longdaysin (1) structure, that must be considered for a successful application in reversible modulation of the circadian rhythm. Trans-to-cis isomerization can be performed by irradiation with one wavelength (purple light), switching the thermally adapted trans-isomer (all-trans) to the cis-enriched isomer (PSS₂). Back-isomerization can be achieved by another wavelength (white light) yielding the trans-enriched isomer distribution (PSS₁). Photochemical stability is measured by repetitive isomerization at two different wavelengths (photoisomerization cycles), and if there is no fatigue present, PSS₁ and PSS₂ will be reached each time. If left in the dark, the cis-isomer will thermally isomerize back to more stable trans-isomer (thermal stability of the cis-isomer) in a first order process. The time needed for half the cis-isomer to isomerize back to the trans is called half-life (t_1/2). Additionally, due to lengthy assays to measure circadian rhythms (5–6 days), azobenzene modulators should possess metabolic stability to maintain light-dependent circadian period modulation.

Fig. 2. Photochemical and biological evaluation of modulators 3-5.

a Structures and photoisomerization scheme of modulators 3-5. b Cis-to-trans thermal isomerization in cell culture medium followed in the dark. Photostationary states (PSSs, pie charts) were reached upon irradiation of DMSO solution (2 mM, 25 °C) with UV light (λ_max = 365 nm) and subsequent dilution in the cell culture medium to obtain the final concentration of 40 µM. The amount of the cis-isomer is shown in pie chart (purple). After 12–14 h in the dark (gray background), blue light (λ_max = 450 nm, blue rectangle) was applied for 8 min to confirm the presence of the remaining cis-isomer and provided the cis-to-trans ratio. c Photo-modulation of the CKIα inhibition, shown as the dose-response curve for the inhibitor kept in the dark (black line) and irradiated with UV light (purple line) before addition to the enzymatic reaction mixture (30 min) and during the assay (3 h). d Luminescence rhythm profiles (average of n = 6) and period-changes of U2OS reporter cells relative to DMSO control, applied with the compounds kept in the dark (black lines) or irradiated with UV light (purple line). Data for longdaysin and DMSO-treated cells are shown in Supplementary Fig. 24a, b, respectively. Luminescence is given in arbitrary units. n = 2 biologically independent samples for the enzymatic assay (c) and n = 6 biologically independent samples for the cellular assay (d). Two-way ANOVA followed by a Sidak’s multiple comparisons test was used for statistical analysis. P value is shown in the figure.

Fig. 3. Photochemical and biological evaluation of modulators 6-8.

a Chemical structures of 6-8, photochemical trans-to-cis isomerization (UV light, λ_max = 365 nm), thermal- or white- light- induced back-isomerization and corresponding percentage of cis isomer (pie chart, purple) that can be generated under irradiation at PSS and thermal half-lives of modifiers 6, 7, and 8 in cell culture medium (40 µM, 35 °C). Upon reaching PSS distribution under UV light irradiation, thermal back-isomerization was followed in dark for 12–14 h. b Kinase (CKIα) and c cellular (U2OS cells) assay data for modulators 6-8. Dose-response curves of the kinase inhibition or the circadian period lengthening under the dark condition are shown in black, upon irradiation with UV light in purple, and back-switching with white light is indicated with red lines. Effects of dark, UV light (purple bulb), and UV light followed by white light (yellow bulb) conditions on luminescence rhythm profiles (average of n = 6) are also shown in the top panels of c (8 µM of each inhibitor). Data for longdaysin- and DMSO-treated cells are shown in Supplementary Fig. 24c, d, respectively. Luminescence is given in arbitrary units. n = 2 biologically independent samples for the enzymatic assay (b), and n = 4–6 biologically independent samples for the cellular assay (c). Two-way ANOVA followed by a Sidak’s multiple comparisons test was used for statistical analysis. P value is shown in the figure.

Fig. 4. Photochemical and biological evaluation of modulator 9.

a Photoisomerization scheme using green light (λ_max = 530 nm) for trans-to-cis and violet light (λ_max = 400 nm) for back-isomerization. b Thermal half-live of modifier 9 in cell culture medium (t_1/2 > 50 h, 35 °C, 40 µM). Upon reaching PSS distribution with green light, thermal back-isomerization was followed in the dark for 14 h (gray background). After this period, violet light was applied (violet rectangle) for 8 min to confirm the presence of the remaining cis-isomer and estimate cis-to-trans ratio. Green light allowed for switching to the cis-isomer once again (green rectangle), showing that in situ isomerization is possible. c Reversible photochromism in cell culture medium followed at 312 nm absorbance. Almost no fatigue was observed after nine repetitions (cell culture medium, 40 µM, 35 °C) of irradiation with violet and green light. d Light-modulation of the CKIα and CKIδ inhibition. The trans-isomer (dark, shown in black) shows higher potency than the cis-enriched mixture (green light, shown in green). e Circadian period modulation in U2OS cells with light. The thermally adapted sample is shown in black (dark), irradiated with green light in vitro (pre-incubation, shown in green), with violet light in cellulo (post-incubation, shown in gray), and with green light in vitro followed by violet light in cellulo in purple. The control experiment with longdaysin is also shown. Luminescence rhythm profiles and data for DMSO-treated cells are shown in Supplementary Figs. 25 and 26, respectively. Luminescence is given in arbitrary units. n = 3 biologically independent samples for the CKIα enzymatic assay (d), n = 2 biologically independent samples for the CKIδ enzymatic assay (d), and n = 3 biologically independent samples for the cellular assay (e). Two-way ANOVA followed by a Sidak’s multiple comparisons test (e, top panels) or one-way ANOVA followed by a Tukey’s multiple comparisons test (e, bottom panels) was used for statistical analysis. P value is shown in the figure.

Fig. 5. Biological evaluation of compound 9 for a long-term photo-modulation of the circadian period.

a The circadian period change over 6 days of the U2OS cellular assay starting with trans-9. Thermally adapted 9 was applied to the cells, and the period change was monitored in the dark for 3 days (gray background, black and green lines). Half of the plate was kept in dark for additional three days (black line) while the other half was irradiated shortly with green light (λ_max = 530 nm), followed by 3 days recording in dark (green background, green line). b The circadian period change over six days of the cellular assay starting with cis-enriched or trans-9. Three-day period change was monitored in the dark upon the application of thermally adapted 9 (black and gray lines) or following short green light irradiation in cellulo (green background, green and violet lines). On the 3rd day, half of the cells with cis-enriched 9 were irradiated shortly with violet light (λ_max = 400 nm, violet background, violet line) and half of the cells with trans-9 were irradiated shortly with violet light (gray line). The other half of the cells with cis-enriched and trans-9 were kept in dark and recorded additional 3 days (green and black lines, respectively). Luminescence is given in arbitrary units. n = 6 biologically independent samples. Two-way ANOVA followed by a Sidak’s multiple comparisons test (a, b, top panels) or one-way ANOVA followed by a Tukey’s multiple comparisons test (b, bottom panels) was used for statistical analysis. P value is shown in the figure. Data for longdaysin and DMSO-treated cells are shown in Supplementary Fig. 27.

Fig. 6. Ex vivo photo-modulation of the circadian period in mouse tissue explants.

a–c The spleen explants of the Per2::Luc knock-in reporter mice were applied with 12 µM compound 9 and irradiated with green light (λ_max = 530 nm) for 0 min (dark, black) or 30 min (green). Luminescence rhythms were then monitored and shown in a (mean of n = 5 for dark and n = 6 for green light). Baseline-subtracted data are shown in b. Period changes compared to a DMSO control are plotted in c. Data for DMSO-treated tissues are shown in Supplementary Fig. 28. Luminescence is given in arbitrary units. n = 5 biologically independent samples for dark and n = 6 for green light. Two-sided Student’s t test was used for statistical analysis. P value is shown in the figure. d–f The SCN explants of the Per2::Luc knock-in reporter mice were applied with DMSO or 18 µM compound 9 on day 4 (pink) and irradiated with green light on day 8 (green). Luminescence rhythms of a representative sample for each condition are shown in d. Baseline-subtracted and normalized data of d are shown in e. Peaks are indicated by orange dots. Period changes are plotted in f. Data are mean with SD (n = 7 biologically independent samples for DMSO and n = 12 for compound 9). One-sided Friedman test with post hoc Steel-Dwass test was used for statistical analysis. P value is shown in the figure.

Fig. 7. Circadian phase control using chronophotopharmacology.

Green light (λ_max = 530 nm)-irradiated compound 9 was applied to U2OS cells at 0 h (A). On the next day (B), the cells were irradiated with violet light (λ_max = 400 nm) for 0 min (black and green lines), 10 min (light violet line), or 20 min (dark violet line). After 48 h or 72 h (C), the cells were irradiated with green light (λ_max = 530 nm) for 0 min (black line) or 30 min (green, light violet, and dark violet lines). Luminescence rhythm profiles (average of n = 6) are shown in a. Phase and period changes relative to DMSO controls after time point C are shown in b and c, respectively. Data for DMSO-treated cells are shown in Supplementary Fig. 29. Luminescence is given in arbitrary units. n = 6 biologically independent samples. One-way ANOVA followed by a Tukey’s multiple comparisons test was used for statistical analysis. P value is shown in the figure.