Knock-down of Cdk5 in the SCN shortens period and reduces phase delays but not phase advances.

(a) Examples of double plotted wheel-running actograms of control (scr) and Cdk5 knock-down (shCdk5) male mice. Animals were kept under a 12 h light / 12 h dark cycle (white and grey areas, respectively) (LD). After 8-10 days they received a 15 min. light pulse at the indicated zeitgeber times (ZT) (yellow stars). After the light pulse animals were released into constant darkness (DD). This light pulse assessment is termed Aschoff type II. (b) Circadian period (τ) of shCdk5 mice (red) is significantly shorter compared to scr controls (blue). τ scr = 23.21 ± 0.08 h, τ shCdk5 = 22.47 ± 0.09 h. All values are mean ± SEM, unpaired t-test with Welch’s correction, n = 6, ***p < 0.001. (c) Quantification of phase shifts (φ) after a 15 min. light pulse at ZT10, ZT14 and ZT22. The phase shift at ZT14 is strongly reduced in shCdk5 animals (red) compared to scr controls (blue). scr: φ ZT10: -1.93 ± 1.43 min., φ ZT14: -105.24 ± 1.54 min., φ ZT22: 34.30 ± 2.97 min., shCdk5: φ ZT10: -2.60 ± 1.72 min., φ ZT14: -11.80 ± 2.81 min., φ ZT22: 35.88 ± 5.68 min. All values are mean ± SEM, unpaired t-test with Welch’s correction, n = 5-6, ****p < 0.0001. (d) Examples of double plotted wheel-running actograms of control (scr) and Cdk5 knock-down (shCdk5) mice. Animals were kept under DD. After 8-10 days they received a 15 min light pulse at the indicated circadian times (CT) (orange stars). This light pulse assessment is termed Aschoff type I. (e) Circadian period of shCdk5 mice (red) is significantly shorter compared to scr controls (blue). τ scr = 23.20 ± 0.05 h, τ shCdk5 = 22.48 ± 0.09 h. All values are mean ± SEM, unpaired t-test with Welch’s correction, n = 5, ***p < 0.001. (f) Quantification of phase shifts after a 15 min. light pulse at CT10, CT14 and CT22. The phase shift at CT14 is strongly reduced in shCdk5 animals (red) compared to scr controls (blue). scr: φ ZT10: 0.12 ± 3.31 min., φ ZT14: -121.52 ± 8.18 min., φ ZT22: 48.40 ± 3.43 min., shCdk5: φ ZT10: -1.68 ± 2.78 min., φ ZT14: -46.60 ± 5.84 min., φ ZT22: 54.16 ± 3.19 min. All values are mean ± SEM, unpaired t-test with Welch’s correction, n = 5, ***p < 0.001.

Cdk5 activity is modulated by light in the early night.

Immuno Western blotting (IB), immunoprecipitation (IP) and Cdk5 kinase activity assays from SCN tissue extracts harvested 30 min after light (+) and no light (-) given at ZT14. (a) Western blot depicting the amounts of phospho PKA (p-PKA), p35 co-activator, Cdk5, and tubulin (Tub, control) before and after light pulse at ZT14. (b) Quantification of p-PKA relative to tubulin. Values are the mean ± SEM. Unpaired t-test, n = 3, *p < 0.05. (c) Quantification of p35 co-activator of Cdk5. Values are the mean ± SEM. Unpaired t-test, n = 3, ***p < 0.001. (d) Cdk5 kinase activity assay. IP of SCN extracts with antibodies against Cdk5 showing the presence of Cdk5 (upper panel) and total protein with Coomassie blue staining (lower panel) as a control for the presence of H1. The middle panel depicts histone H1 phosphorylated by Cdk5, visualized as 32P-histone H1 (32P-H1). (e) Quantification of Cdk5 kinase activity relative to H1 levels. Values are the mean ± SEM. Unpaired t-test, n = 3, ***p < 0.001. (f) Co-immunoprecipitation of p35 with Cdk5 before and after a light pulse.

Cdk5 impacts on the CREB signaling pathway via calcium/calmodulin dependent kinases (CaMK).

The cartoons on the left of each figure depict the CaMK pathway with the red rectangle indicating the visualization of a particular the component. (a) Immunohistochemistry on the SCN of control (scr) and shCdk5 mice using an antibody recognizing phospho-serine 133 of CREB (p-CREB-S133) before and after a light pulse at ZT14. The red color shows p-CREB-S133 and the blue color represents Dapi stained nuclei of SCN cells. Scale bar: 8 µm. The right panel shows the quantification of the p-CREB-S133 signal. Values are the mean ± SEM. Unpaired t-test with Welch’s correction, n = 3, *p < 0.05. (b) Immunohistochemistry on the SCN of control (scr) and shCdk5 mice using an antibody recognizing Cam kinase II (CaMKII) before and after a light pulse at ZT14. The red color shows CaMKII and the blue color represents Dapi stained nuclei of SCN cells. Scale bar: 8 µm. The right panel shows the quantification of the CaMKII signal. Values are the mean ± SEM. Unpaired t-test with Welch’s correction, n = 3, *p < 0.05. (c) Translocation of calmodulin (CAM) in response to a light pulse at ZT14 in SCN neurons of control (scr) and Cdk5 knock-down (shCdk5) animals. CAM (yellow) accumulates around the nuclei in scr controls. In shCdk5 SCN neurons, this accumulation around the nuclei is already seen before the light pulse which is clearly different from scr controls. Scale bar: 7 µm The right panel shows the quantification of CAM rings. Values are the mean ± SEM. Unpaired t-test with Welch’s correction, n = 3, *p < 0.05, **p < 0.01. (d) Immunohistochemistry on the SCN of control (scr) and shCdk5 mice using an antibody recognizing Cam kinase IV (CaMKIV) before and after a light pulse at ZT14. The red color shows CaMKIV and the blue color represents Dapi stained nuclei of SCN cells. Scale bar: 5 µm. The right panel shows the quantification of the CaMKIV signal. Values are the mean ± SEM. Unpaired t-test with Welch’s correction, n = 3, *p < 0.05. (e) Immunohistochemistry on the SCN of control (scr) and shCdk5 mice using an antibody recognizing the calcium channel Cav3.1 before and after a light pulse at ZT14. The green color shows Cav3.1 and the blue color represents Dapi stained nuclei of SCN cells. Scale bar: 5 µm. The right panel shows the relative Cav3.1 signal. Values are the mean ± SEM. Unpaired t-test with Welch’s correction, n = 3, *p < 0.05.

Neuronal activity in response to light at ZT14 is modulated by CDK5.

(a) Illustration of the chronic optic fiber implantation in the SCN for fiber photometry recording in freely moving mice (left). The animals were previously infected either with AAV9-hSyn1-chl[1x(shNS)]-jGCaMP7b-WPRE-SV40p(A)(scr) or AAV9-hSyn1-chl[mouse(shCdk5)]-jGCaMP7b-WPRE-SV40p(A)(shCdk5). The experimental timeline of one trial is shown on the right. White and dark boxes represent the light and dark phase, respectively. Fiber photometry recordings were done in the 5 min. before the light pulse, during the light pulse, and 15 min. after. The light pulse was 15 min. long and was delivered at ZT14 (blue box; dashed lines between minutes 14-15 are not included in the analysis). (b) Mean traces ± SEM of cell activity (normalized ΔF/F0) of GCaMP7b-expressing SCN neurons (black: scr, red: shCdk5) 5 min. before, 15 min. during, and 15 min. after the light pulse (± 20 s) light pulse delivered at ZT14. N = 15 trials, n = 5 mice / red = shCdk5, N = 12 trials, n = 4 mice. (c) Bar plot showing the percentage of ΔF/F0 changes ± SEM 5 min. before the light pulse, in the first and last 5 min. during the light pulse and the first and last 5 min. after the light pulse. 9-14 minutes of light pulse: scramble (black bar) 105.5 ± 19.3 ΔF/F0 vs. shCdk5 (red bar) 74.7 ± 17.2 ΔF/F0. 1-5 minutes after light pulse: scramble (black bar) 111.2 ± 19.3 ΔF/F0 vs. shCdk5 (red bar) 81.0 ± 17.8 ΔF/F0. Black = scr, N = 15 trials, n = 5 mice / red = shCdk5, N = 12 trials, n = 4 mice. Bar values represent the mean ± SEM. ***p < 0.001; two-way ANOVA corrected with Bonferroni post-hoc test. (d) Photomicrograph of the expression of GCaMP7b (green) in the SCN in both control (scr, left) and experimental (shCdk5, right) animals. The red hatched oval indicates the placement of the optic fiber. Blue: Dapi, green: GFP (produced by jGCaMPP7). Scale bar 50 µm.

CDK5 regulates PKA phosphorylation via DARPP32 phosphorylation.

(a) Top: Scheme of the forskolin-PKA-CREB signaling pathway. Bottom: FRET/CFP signal ratio changes in response to forskolin treatment in NIH 3T3 cells transfected with either a scr control (blue) or shCdk5 (red) expression construct. Values are the mean ± SD. Two-way ANOVA revealed a significant difference between the curves, n = 3, ****p < 0.0001. (b) Top: Scheme of the forskolin-PKA-CREB signaling pathway and calcium signaling. Bottom: FRET/CFP signal ratio changes in response to forskolin treatment in NIH 3T3 cells with the addition of Ca2+ (blue), without the addition of Ca2+ (salmon colored), and with the addition of Ca2+ and EGTA (orange). Values are the mean ± SD. Two-way ANOVA revealed a significant difference between the grey and blue/orange curves, n = 3, ****p < 0.0001. (c) Scheme of CDK5-DARPP32-PKA pathway. (d) Immunohistochemistry on the SCN of control (scr) and shCdk5 mice using an antibody recognizing phosphorylated Thr-75 of DARPP32 (p-DARPP32) before and after a light pulse at ZT14. The red color shows p-DARPP32 and the blue color represents Dapi stained nuclei of SCN cells. Scale bar: 10 µm. The right panel shows the quantification of the p-DARPP32 signal. Values are the mean ± SEM. Unpaired t-test with Welch’s correction, n = 3, **p < 0.01. (e) Immunohistochemistry on the SCN of control (scr) and shCdk5 mice using an antibody recognizing phosphorylated Thr-197 of PKA (p-PKA) before and after a light pulse at ZT14. The red color shows p-PKA and the blue color represents Dapi stained nuclei of SCN cells. Scale bar: 20 µm. The right panel shows the quantification of the p-PKA signal. Values are the mean ± SEM. Unpaired t-test with Welch’s correction, n = 3, *p < 0.05, **p < 0.01. (f) Immunohistochemistry on the SCN of control (scr) and shCdk5 mice using an antibody recognizing phosphorylated Thr-197 of PKA (p-PKA) and Cav3.1 before and after a light pulse at ZT14. The red color shows p-PKA, the green color Cav3.1, and the blue color represents Dapi-stained nuclei of SCN cells. The yellow color signifies the co-localization of PKA and Cav3.1. The stripes on the left and bottom of each micrograph show the z-stacks to confirm co-localization. Scale bar: 10 µm. The right panel shows the quantification of relative p-PKA/Cav3.1. Values are the mean ± SEM. Unpaired t-test with Welch’s correction, n = 3, *p < 0.05.

Cdk5 regulates light-induced gene expression in the SCN of some clock genes.

Relative mRNA values are represented as blue bars for scr control animals and as red bars for shCdk5 mice. The values were determined 0, 0.5, 1, and 2 hours after a light pulse (LP) given at ZT14. (a) Induction of Per1 mRNA expression by light with a maximum of 1 hour after light in scr control animals. In contrast Per1 is not induced in shCdk5 SCN. Scr: 0 h: 0.44 ± 0.01, 0.5 h: 0.65 ± 0.12, 1 h: 1.12 ± 0.19, 2 h: 0.44 ± 0.01; shCdk5: 0 h: 0.44 ± 0.01, 0.5 h: 0.44 ± 0.01, 1 h: 0.44 ± 0.01, 2 h: 0.77 ± 0.18. Values are the mean ± SEM. Unpaired t-test, n = 3, *p < 0.05. (b) Per2 mRNA expression is not induced by light neither in scr controls nor in shCdk5 animals. Scr: 0 h: 1.04 ± 0.16, 0.5 h: 0.95 ± 0.10, 1 h: 0.83 ± 0.10, 2 h: 0.96 ± 0.07; shCdk5: 0 h: 1.01 ± 0.05, 0.5 h: 1.07 ± 0.10, 1 h: 1.04 ± 0.14, 2 h: 1.04 ± 0.12. Values are the mean ± SEM. Unpaired t-test, n = 3 (c) Induction of Dec1 mRNA expression by light with a maximum at 1 hour after light in scr control animals. In contrast Dec1 is not induced in shCdk5 SCN. Scr: 0 h: 1.87 ± 0.37, 0.5 h: 3.32 ± 0.75, 1 h: 4.25 ± 0.49, 2 h: 3.13 ± 0.34; shCdk5: 0 h: 1.96 ± 0.34, 0.5 h: 2.13 ± 0.51, 1 h: 1.87 ± 0.07, 2 h: 2.32 ± 0.40. Values are the mean ± SEM. Unpaired t-test, n = 3, *p < 0.05. (d) Dec2 mRNA expression is not induced by light neither in scr controls nor in shCdk5 animals. Scr: 0 h: 1.93 ± 0.29, 0.5 h: 1.94 ± 0.25, 1 h: 1.75 ± 0.49, 2 h: 2.11 ± 0.07; shCdk5: 0 h: 1.94 ± 0.23, 0.5 h: 1.58 ± 0.07, 1 h: 1.61 ± 0.25, 2 h: 1.90 ± 0.33. Values are the mean ± SEM. Unpaired t-test, n = 3. (e) Bmal1 mRNA expression is not induced by light in the SCN of scr control and shCdk5 animals. Scr: 0 h: 0.60 ± 0.04, 0.5 h: 0.69 ± 0.06, 1 h: 0.65 ± 0.05, 2 h: 0.81 ± 0.12; shCdk5: 0 h: 0.61 ± 0.14, 0.5 h: 0.61 ± 0.12, 1 h: 0.56 ± 0.06, 2 h: 0.71 ± 0.15. Values are the mean ± SEM. Unpaired t-test, n = 3. (f) eGFP mRNA expression is detected in the SCN of scr control and shCdk5 animals demonstrating proper injection of expression constructs (scr: ssAAV-9/2-hSyn1-chl[1x(shNS)]-EGFP-WPRE-SV40p(A), shCdk5: ssAAV-9/2-hSyn1-chl[mouse(shCdk5)]-EGFP-WPRE-SV40p(A)). Scr: 0 h: 1.04 ± 0.19, 0.5 h: 1.23 ± 0.22, 1 h: 0.90 ± 0.07, 2 h: 1.34 ± 0.34; shCdk5: 0 h: 1.31 ± 0.28, 0.5 h: 1.04 ± 0.13, 1 h: 0.90 ± 0.05, 2 h: 1.15 ± 0.06. Values are the mean ± SEM. Unpaired t-test, n = 3. (g) Induction of cFos mRNA 0.5 hours after the light pulse in both scr controls and shCdk5 SCN. Scr: 0 h: 0.77 ± 0.07, 0.5 h: 2.55 ± 0.38, 1 h: 0.64 ± 0.25, 2 h: 0.61 ± 0.15; shCdk5: 0 h: 0.95 ± 0.08, 0.5 h: 2.57 ± 0.05, 1 h: 0.68 ± 0.04, 2 h: 0.74 ± 0.12. Values are the mean ± SEM. Unpaired t-test, n = 3, *p < 0.05, ***p < 0.001. (h) Induction of Egr1 mRNA 0.5 hours after the light pulse in scr control but not shCdk5 SCN. Scr: 0 h: 1.25 ± 0.38, 0.5 h: 2.71 ± 0.19, 1 h: 2.12 ± 0.41, 2 h: 1.26 ± 0.13; shCdk5: 0 h: 1.23 ± 0.24, 0.5 h: 1.77 ± 0.13, 1 h: 1.16 ± 0.08, 2 h: 1.18 ± 0.06. Values are the mean ± SEM. Unpaired t-test, n = 3, *p < 0.05. (i) Sik1 mRNA expression is induced by light in the SCN of scr control but not shCdk5 animals. Scr: 0 h: 0.29 ± 0.07, 0.5 h: 0.58 ± 0.02, 1 h: 0.27 ± 0.14, 2 h: 0.34 ± 0.07; shCdk5: 0 h: 0.22 ± 0.03, 0.5 h: 0.20 ± 0.01, 1 h: 0.19 ± 0.02, 2 h: 0.25 ± 0.02. Values are the mean ± SEM. Unpaired t-test, n = 3, *p < 0.05, **p < 0.01. (j) Gem mRNA expression is induced by light in the SCN of scr control but not shCdk5 animals. Scr: 0 h: 0.70 ± 0.07, 0.5 h: 0.86 ± 0.05, 1 h: 1.38 ± 0.11, 2 h: 0.94 ± 0.11; shCdk5: 0 h: 1.05 ± 0.24, 0.5 h: 1.08 ± 0.03, 1 h: 1.13 ± 0.07, 2 h: 0.93 ± 0.05. Values are the mean ± SEM. Unpaired t-test, n = 3, *p < 0.05.

Model of Cdk5 gated light signal.

(a) Cdk5 is active during the dark portion of the day. Active Cdk5 with its co-activator p35 phosphorylates PER2, which leads to stabilization and nuclear translocation of this protein that is abundant at ZT12. At the same time CDK5 phosphorylates DARPP32, which inhibits the PKA signaling pathway. (b) Light perceived in the dark phase at ZT14 leads to detachment of p35 from Cdk5 stopping Cdk5 activity. DARPP32 is not phosphorylated and hence can’t inhibit PKA. PKA that is activated by the light signal is phosphorylated and can mediate CREB phosphorylation via T-type calcium channels (Cav3.1) and the CaMK pathway leading to a transcriptionally active complex on the CRE element present in the promoters of many light responsive genes such as Per1, Dec1, Gem and Sik1. Overall, a light pulse at ZT14 will activate CREB phosphorylation and a protein complex will form. This complex needs phosphorylated PER2 that has accumulated in the nucleus between ZT12 and ZT14 to initiate the transcription of light-responsive genes. Both arms are necessary to build up a transcriptionally functional complex. Both arms depend on the presence and activity of CDK5, which therefore gates the light signal at ZT14. It is very likely that the amount of PER2 protein in the nucleus determines at least in part the magnitude of the phase delay, which depends on the timing of the light signal.

Antibodies used for the immunostainings.

Antibodies used for the western blots.