Regulation of CRE-Dependent Transcriptional Activity in a Mouse Suprachiasmatic Nucleus Cell Line

We evaluated the signalling framework of immortalized cells from the hypothalamic suprachiasmatic nucleus (SCN) of the mouse. We selected a vasoactive intestinal peptide (VIP)-positive sub-clone of immortalized mouse SCN-cells stably expressing a cAMP-regulated-element (CRE)-luciferase construct named SCNCRE. We characterized these cells in terms of their status as neuronal cells, as well as for important components of the cAMP-dependent signal transduction pathway and compared them to SCN ex vivo. SCNCRE cells were treated with agents that modulate different intracellular signalling pathways to investigate their potency and timing for transcriptional CRE-dependent signalling. Several activating pathways modulate SCN neuronal signalling via the cAMP-regulated-element (CRE: TGACGCTA) and phosphorylation of transcription factors such as cAMP-regulated-element-binding protein (CREB). CRE-luciferase activity induced by different cAMP-signalling pathway-modulating agents displayed a variety of substance-specific dose and time-dependent profiles and interactions relevant to the regulation of SCN physiology. Moreover, the induction of the protein kinase C (PKC) pathway by phorbol ester application modulates the CRE-dependent signalling pathway as well. In conclusion, the cAMP/PKA- and the PKC-regulated pathways individually and in combination modulate the final CRE-dependent transcriptional output.


Introduction
Transcriptional activation in response to an elevation of intracellular adenosine-3 ,5 -(cyclic)monophosphate (cAMP) is a widespread phenomenon depending, among other factors, on the dissociation of PKA into regulatory and catalytic subunits and transport of the catalytic subunit of PKA from the cytosol to the nucleus [1,2]. In the suprachiasmatic nucleus (SCN) of the hypothalamus cAMP/PKA-dependent phosphorylation of cAMPregulated-element-binding protein (CREB) has been shown to be involved in the amplitude and phase shift of circadian rhythms [3]. The SCN was one of the first neuronal systems in which CREB and CRE (TGACGCTA) regulated transcriptional activity have been shown to play a physiological role, namely clock genes induction and phase shift of their circadian expression [3].
In the SCN, as well as in other cells and tissues, the second messenger cAMP is also regarded as critical for the transcriptional activation of many genes containing a CRE or similar regulatory elements [4][5][6].
In the SCN, in vivo cAMP levels vary with a diurnal (light-dark dependent) and circadian ("free running") rhythm that has a phase length of approximately 24 h. However, not all neurons inside the SCN display circadian rhythms. In the non-circadian cells of the SCN, cAMP may be regulated by neurotransmitters and hormones, and these cells may

Characterization of SCNCRE Cells
SCNCRE cells were derived from an immortalized mouse hypothalamic suprachiasmatic nucleus cell line [13]. Using an antibody against "vasoactive intestinal peptide" (VIP) in immunocytochemical (ICC) staining, most SCNCRE cells displayed VIP-positive granular material in the cytoplasm ( Figure 1A,B). These cells were strongly immunopositive for the neuropeptide VIP but only very weakly positive for arginine vasopressin (AVP) (as shown in Figure 1A,B), suggesting that SCNCRE cells represent a subpopulation of VIP-expressing parvocellular neuronal cells from the SCN. This is underscored by our finding that SCNCRE displays a strong signal for the native, unprocessed form of the vasoactive intestinal peptide, proVIP (17 kDa Figure S2). The astrocyte marker glial fibrillary acidic protein (GFAP) was, as expected, present in the SCN ex vivo extracts, but not in SCNCRE cell extracts (Supplementary Figure S2).

Elevation of cAMP in SCNCRE Cells
Application of the adenylate cyclase activator forskolin (10 µM) elevated extracellular cAMP levels more than 100-fold over basal (control) levels in two separate experiments performed on a 96-well and a 24-well plate (p ≤ 0.0001, n = 5, ANOVA with Bonferroni post-test; Figure 2) against vehicle controls. Cell culture medium and vehicle control cAMP levels were not significantly different. In an additional experiment, forskolin (1 µM) elevated intracellular cAMP levels significantly only after 30 min and 2 h of forskolin application in comparison to a time-matched control. After 4-, 6-and 12-h cAMP levels were not significantly different between control and forskolin-treated SCNCRE cell cultures (Supplementary Figure S6).

Characteristics of the CRE-Luciferase Response in SCNCRE Cells
As stated above, SCNCRE cells are stably expressing a CRE-luciferase reporter gene (CRE-luc) and the luminescence output produced by this CRE-luc reporter is displayed in relative luminescence units (RLU). Figure 3A shows that application of forskolin (1µM) to cultures containing 25.000 cells per well of a 96-well plate leads to an elevation of luciferase activity over untreated controls ( Figure 3A) which depends on the concentration of luciferin in the cell culture medium. A stable relation between forskolin (1 µM) and control activity is reached at luciferin concentrations above 200 µM. This forskolin-induced elevation of CRE-dependent luciferase activity also depends on the number of cells plated and reaches a stable forskolin-treated to untreated control relation above a cell number of

Characteristics of the CRE-Luciferase Response in SCNCRE Cells
As stated above, SCNCRE cells are stably expressing a CRE-luciferase reporter gene (CRE-luc) and the luminescence output produced by this CRE-luc reporter is displayed in relative luminescence units (RLU). Figure 3A shows that application of forskolin (1µM) to cultures containing 25.000 cells per well of a 96-well plate leads to an elevation of luciferase activity over untreated controls ( Figure 3A) which depends on the concentration of luciferin in the cell culture medium. A stable relation between forskolin (1 µM) and control activity is reached at luciferin concentrations above 200 µM. This forskolin-induced elevation of CRE-dependent luciferase activity also depends on the number of cells plated and reaches a stable forskolin-treated to untreated control relation above a cell number of 10.000 cells per well with a luciferin concentration fixed at 500 µM ( Figure 3B). Figure 3C shows the dose-dependent increase of CRE-dependent luciferase activity upon forskolin treatment with 25.000 cells per well and a luciferin concentration of 250 µM. Besides cell number plated, luciferin-and forskolin-concentration, the timing of the response onset, maximum and offset of activity are other important parameters to consider. At a constant cell number of SCNCRE cells per well, the maximum response to forskolin (10 µM) was approximately an hour later than the response to forskolin (1 µM), and the maximum of forskolin (0.1 µM) was even earlier ( Figure 3D). Thus, the higher the applied concentration of forskolin, the later the maximal response time.    For the following experiments, the number of cells plated out was fixed to 25.000/well and the final luciferin concentration to 250 µM if not indicated otherwise. To evaluate if the observed CRE-luciferase elevation is mediated by the suspected signal transduction proteins, we next tested the influence of the adenylyl cyclase inhibitor MDL12.330A ( Figure 4A). MDL12.330 (10 µM) showed significant inhibition of the maximum response of forskolin (1 µM) ( Figure 4A). However, MDL12.330 applied alone displayed drastically lower basal RLU values compared to the vehicle controls, which indicates a toxic effect. This presumably toxic effect is also seen in Figure 4B, where forskolin (0.1 µM) is nearly fully inhibited by MDL12.330 but only compared to the vehicle control, not to MDL12.330 (10 µM) alone. One conclusion from these data could be that MDL12.330 at 10 µM is not inhibiting specifically adenylyl cyclase but is most probably toxic to the cells. The next level of regulation is cAMP-activated protein kinase (PKA). A well-established class of PKA inhibitors are the equatorial diastereomers of 3 ,5 -cyclic adenosine phosphorothioate. In particular, the Rp-cAMPS analogue, Rp-8-Br-cAMPS, has been shown to be an efficient inhibitor of PKA. Figure 5A shows that the SCNCRE cell response to forskolin The next level of regulation is cAMP-activated protein kinase (PKA). A well-established class of PKA inhibitors are the equatorial diastereomers of 3′,5′-cyclic adenosine phosphorothioate. In particular, the Rp-cAMPS analogue, Rp-8-Br-cAMPS, has been shown to be an efficient inhibitor of PKA. Figure 5A shows that the SCNCRE cell response to forskolin (1 µM) was inhibited by approximately 50% by Rp-8-Br-cAMPS (1 mM), and the response to forskolin (0.1 µM) was nearly abolished ( Figure 5B). Another important protein class relevant to the regulation of CRE-luc activity are the 3′,5′-cyclic adenosine phosphodiesterases (PDE). Many PDEs catalyze the degradation of cAMP to 5′-AMP (or cGMP to 5′-GMP), but some also bind cAMP or cGMP in a separate site. In the presence of a PDE inhibitor in a cell expressing PDEs, the concentration of cAMP rises. A widely used PDE inhibitor exhibiting a broad inhibition spectrum across the 11 described PDE families is isobutyl-methyl-xanthine (IBMX) [19]. Here, the Another important protein class relevant to the regulation of CRE-luc activity are the 3 ,5 -cyclic adenosine phosphodiesterases (PDE). Many PDEs catalyze the degradation of cAMP to 5 -AMP (or cGMP to 5 -GMP), but some also bind cAMP or cGMP in a separate site. In the presence of a PDE inhibitor in a cell expressing PDEs, the concentration of cAMP rises. A widely used PDE inhibitor exhibiting a broad inhibition spectrum across the 11 described PDE families is isobutyl-methyl-xanthine (IBMX) [19]. Here, the application of IBMX (10 µM) showed robust induction of CRE-luc activity in SCNCRE cells with a maximum after approximately four hours ( Figure 6A). When co-incubated with forskolin (1 µM; which peaks later), the resulting activation curve in SCNCRE cells showed an intermediate maximum of CRE-luc activity. The onset was slower than under treatment with IBMX alone and the maximum time was nearly identical to the maximum time of forskolin alone. A similar result was seen with IBMX (10 µM) and forskolin (0.1 µM) combined ( Figure 6B).  Another cellular system of potential relevance for cAMP-mediated CRE-luc activity is the organic anion exchanger (OAE). Under the influence of high intracellular cAMP levels (like when cells are incubated with strong cAMP-elevating agents like forskolin or IBMX), OAE enables the release of cAMP from the cell into the extracellular space. A well-established inhibitor for OAE is probenecid, an established medical compound which serves to reduce the excretion of organic ions from the kidney. Probenecid (10 µM) had no significant additional effect on CRE-luc activity in SCNCRE cells when co-applied with forskolin (1 µM) ( Figure 7A), however at lower forskolin (0.1 µM) concentration probenecid elevated the reaction to forskolin significantly. However, compared to the impact of IBMX, this approximate 20% further elevation appears to be of rather minor importance. luciferase activity induced by 1 µM and 0.1 µM forskolin (A,B). IBMX application displayed significant activation if applied alone which is further elevated when co-applied with forskolin. IBMX displays an earlier maximal response compared to forskolin (approximately 2 h earlier). (Shown are the means values ± SD; ANOVA with Bonferroni post-test; n = 5).
Another cellular system of potential relevance for cAMP-mediated CRE-luc activity is the organic anion exchanger (OAE). Under the influence of high intracellular cAMP levels (like when cells are incubated with strong cAMP-elevating agents like forskolin or IBMX), OAE enables the release of cAMP from the cell into the extracellular space. A wellestablished inhibitor for OAE is probenecid, an established medical compound which serves to reduce the excretion of organic ions from the kidney. Probenecid (10 µM) had no significant additional effect on CRE-luc activity in SCNCRE cells when co-applied with forskolin (1 µM) ( Figure 7A), however at lower forskolin (0.1 µM) concentration probenecid elevated the reaction to forskolin significantly. However, compared to the impact of IBMX, this approximate 20% further elevation appears to be of rather minor importance.  As mentioned above, besides concentration and mechanisms of action, we observed the timing of the SCNCRE luciferase responses. To further evaluate this, we applied the PKA agonists SpcAMPS and SpcDBIMPS and the EPAC activator (EPAC-A) in addition to forskolin (Figure 8). Next, we compared the onset and maximum activity of CRE-luciferase activity for forskolin (1 µM), SpcAMPS, SpcDBIMPS and EPAC-A (all 100 µM). SpcAMPS displayed both the fastest onset and earliest significant elevation over control levels (p < 0.001 over control at 1 h). SpcAMPS is followed by SpcDBIMPS and forskolin. EPAC-A did not significantly elevate SCNCRE luciferase activity at any time point. The As mentioned above, besides concentration and mechanisms of action, we observed the timing of the SCNCRE luciferase responses. To further evaluate this, we applied the PKA agonists SpcAMPS and SpcDBIMPS and the EPAC activator (EPAC-A) in addition to forskolin (Figure 8). Next, we compared the onset and maximum activity of CREluciferase activity for forskolin (1 µM), SpcAMPS, SpcDBIMPS and EPAC-A (all 100 µM). SpcAMPS displayed both the fastest onset and earliest significant elevation over control levels (p < 0.001 over control at 1 h). SpcAMPS is followed by SpcDBIMPS and forskolin. EPAC-A did not significantly elevate SCNCRE luciferase activity at any time point. The CREluc activity increase in response to SpcAMPS application remained the highest until after 6 h of incubation. SpcAMPS, SpcDBIMPS and forskolin reached a similar maximal plateau, at which the RLU values did no longer increase and were not significantly different (Figure 8). Another class of protein kinases which was shown to be able to phosphorylate CREB and thereby activate pCREB/CRE-dependent transcriptional responses are the protein kinases C (PKC). The classical PKCs (α, β, γ) are activated by a class of substances called phorbol esters. In contrast to forskolin (10 µM), the phorbol ester PMA (1 µM) did not change cAMP levels ( Figure 9A). However, PMA did elevate pCREB levels significantly higher than Forskolin and as high as the combination of both PMA and forskolin ( Figure  9B) when compared by Western blot analysis. In the CRE reporter gene activity assay after one hour, the RLU levels under PMA treatment did not differ from vehicle control, whereas the forskolin RLU was significantly elevated and the combination of forskolin and PMA ( Figure 9C). The complete time course presented in Figure 9D shows that PMA elevated CRE-dependent RLU values with an early and low maximum around 3-4 h after onset, an acceleration of the onset in combination with forskolin and a more than additive maximum ( Figure 9D). Another class of protein kinases which was shown to be able to phosphorylate CREB and thereby activate pCREB/CRE-dependent transcriptional responses are the protein kinases C (PKC). The classical PKCs (α, β, γ) are activated by a class of substances called phorbol esters. In contrast to forskolin (10 µM), the phorbol ester PMA (1 µM) did not change cAMP levels ( Figure 9A). However, PMA did elevate pCREB levels significantly higher than Forskolin and as high as the combination of both PMA and forskolin ( Figure 9B) when compared by Western blot analysis. In the CRE reporter gene activity assay after one hour, the RLU levels under PMA treatment did not differ from vehicle control, whereas the forskolin RLU was significantly elevated and the combination of forskolin and PMA ( Figure 9C). The complete time course presented in Figure 9D shows that PMA elevated CRE-dependent RLU values with an early and low maximum around 3-4 h after onset, an acceleration of the onset in combination with forskolin and a more than additive maximum ( Figure 9D).

Discussion
The suprachiasmatic nucleus (SCN) of the mouse consists of a number of different cell types [7]. Among the larger neuronal populations of the SCN are cells containing the neuropeptide vasoactive intestinal peptide (VIP) [20]. In the VIP-positive population, some cells apparently do not display circadian expression of clock genes. One theory is that these cells serve as "input sensors" for external stimuli, perceive these and modulate the neighbouring rhythmic cells [13,21].
From the SCN cell line made available by the group of David Earnest, we extracted cells which did not display any detectable luciferase output. These cells were stably transfected with a CRE-luciferase construct (CRE-luc) and further characterized. The clones stably expressing CRE-luciferase and named SCN-CRE (SCNCRE) were strongly VIP-

Discussion
The suprachiasmatic nucleus (SCN) of the mouse consists of a number of different cell types [7]. Among the larger neuronal populations of the SCN are cells containing the neuropeptide vasoactive intestinal peptide (VIP) [20]. In the VIP-positive population, some cells apparently do not display circadian expression of clock genes. One theory is that these cells serve as "input sensors" for external stimuli, perceive these and modulate the neighbouring rhythmic cells [13,21].
From the SCN cell line made available by the group of David Earnest, we extracted cells which did not display any detectable luciferase output. These cells were stably transfected with a CRE-luciferase construct (CRE-luc) and further characterized. The clones stably expressing CRE-luciferase and named SCN-CRE (SCNCRE) were strongly VIP-positive. Like the original cell line, they seem to consist of two morphologically different populations, large flat cells rapidly forming a dense monolayer and small, round cells located on top of the large cells forming long protrusions and establishing connections to each other under prolonged cultivation duration [13].
We tested a series of marker proteins to evaluate the neuronal nature of the SCNCRE cells and their developmental stage. Interestingly, although SCNCRE is an immortalized cell line, they expressed in the confluent stage two makers for mature neurons, ßIII-tubulin and "microtubule-associated-protein-2B" (MAP2B), no marker for immature (Nestin) or maturing neurons (Doublecortin). Taken together, SCNCRE cells express neuronal markers suggestive of rather mature neuronal phenotype cells but do not express GFAP.
SCNCRE cells displayed an elevated CRE-luc activity after the application of forskolin, a diterpene secondary metabolite from the Indian foul tree, Coleus forskolii [22]. Forskolin is widely used to induce cAMP elevation [23]. In the data presented here, cAMP-synthesis by adenylate cyclase (AC), cAMP-degradation by 3 ,5 -phosphodiesterase (PDE) and cAMPexport via organic anion exchangers (OAE) affected CRE-dependent signalling. Besides the various signalling pathway components involved, differential timing of cAMP-production, -degradation and -export processes are important for the latency time of serine 133 pCREB phosphorylation, CRE-dependent luciferase activity onset and maximum reaction, especially when interacting with other parallel pathways like the protein kinase C (PKC) pathway.
Regarding the type of adenylate cyclase present in SCNCRE cells, data are inconclusive. In our experiments, MDL12330 became toxic around the concentrations where it should inhibit AC and SQ22536, another presumed AC inhibitor, did not inhibit forskolin-mediated CREluc activity in SCNCRE cells (Supplementary Figure S7). Thus, the type and regulation of the AC in SCNCRE cells remain to be investigated.
SCNCRE also reacted to the application of the PDE-resistant cAMP analogues, SpcAMPS and SpcDBIMPS, with SpcAMPS being the faster elevator of CRE-luc activity. This finding may be explained by SCNCRE cells predominantly expressing PKA regulatory subunit isoforms IIα and Iiβ for which SpcAMPS is a better activator than SpcDBIMPS, which displays a strong selectivity for PKA I isoforms [24]. SpcAMPS was also the agent with the fastest CRE-luc response, already elevating levels significantly over basal after 60 min, followed by SpcDBIMPS and forskolin.
EPAC-A selectively activates the EPAC proteins (EPAC1 and EPAC2, also known as Rap guanine nucleotide exchange factor 3, GEF3) representing a family of intracellular sensors for cAMP, which function as nucleotide exchange factors for the Rap subfamily of RAS-like small GTPase [25]. Here, the EPAC-activator (EPAC-A) did not exert any significant change in CRE-luc activity compared to controls. EPAC-A does not activate PKA, nor does it significantly influence AC, PDE or OAE activity [25]. We conclude that EPAC plays no significant role in CRE-luciferase activity modulation.
Another finding of relevance is the shift of onset and the estimated time point of maximum when agents were combined. The onset of CREluc activity in SCNCRE cells is faster for IBMX than for forskolin. When both are applied in combination the onset of activity is not significantly slower than that of IBMX alone. However, the maximal response of the combination is at a time point similar to the maximal time response to the forskolin application. Thus, the combination response behaves similar to the faster substance (IBMX) at the onset of activity and similar to the slower substance (forskolin) at the maximum time. The effect that the higher the dose of forskolin, the later the maximal response time also applies for the combination of forskolin with IBMX in that forskolin determines at what exact time point this maximum occurs. Thus, our data support the interpretation that the extent of the temporal shift of the maximum time depends on the dose of forskolin.
Beside PKA, several other protein kinases have been shown to be able to phosphorylate CREB at serine-133 [26]. One of these "CREB-kinases" is protein kinase C (PKC) which represents a family of fifteen isozymes in humans [27]. The classical PKCs (α, β, γ) can be activated by a class of substances called phorbol esters, like the cell-permeable substance phorbol-12-myristat-13-acetat (PMA) [28]. PKCs are known to activate other protein kinases like the p42/44-MAP-kinases, which can also be present inside the cell nucleus or act via shuttle systems like the one involving pp90RSK [29]. Since PMA and forskolin combination is neither simply additive nor synergistic or antagonistic on CRE-dependent transcription activity, the kinases involved in the PMA-activated pathway at least partly compete with those involved in the forskolin-induced pathway.
In contrast to forskolin, PMA did not modulate cAMP levels in SCNCRE cells in our experiments. However, PMA did elevate pCREB levels after one-hour incubation, as shown by Western blot analysis. Under PMA treatment, pCREB levels were higher than under forskolin treatment and as high as the combination of both PMA and forskolin. However, after one-hour CRE-luc activity under PMA treatment did not differ from vehicle control but rose significantly over vehicle control levels after approximately two hours, whereas forskolin significantly elevated CRE-luc activity already after one hour. Thus, neither cAMP levels nor CREB phosphorylation at serine 133 after one hour is a reliable marker for the maximum or the timing of maximal induction of CRE-dependent transcriptional activity in SCNCRE cells.
It is widely assumed that in the SCN in vivo, different populations of parvocellular neurons located in different regions exert very different tasks with respect to rhythm generation, as well as input and output signalling [8][9][10][11][30][31][32][33][34][35]. We used SCNCRE cells, a VIP-expressing transformed cell population derived from the mouse SCN, to investigate the factors determining the cAMP/PKA/CRE-mediated transcriptional response of such cells.
Our conclusion from the experimental data presented here is that there is a doseand time-sensitive cAMP-mediated transcriptional response that is mostly depending on AC-activation and PDE-expression and -inhibition and, to a lesser degree and only at lower stimulant concentration, on the egress of cAMP through OAG. This cAMP-dependent pathway can interact with PKC-signalling but does not or only marginally depend on other cAMP-target proteins like EPAC.
For the SCN in situ, these data suggest a strong coupling and interaction of the different neuronal cell populations of which those that are non-rhythmic should not be neglected regarding their potential influence on input-and phase-regulation and output functions of other cell populations inside the SCN.

Cell Culture
Immortalized mouse suprachiasmatic nucleus cells (200.000 per well in a 6-well plate; kindly supplied by Prof. David J. Earnest PhD, Texas A&M University, Bryan, TX, USA [13] devoid of spontaneous (or basal) luminescence activity were cultivated in Dulbecco's Modified Eagle's Medium (DMEM; Sigma, Deisenhofen, Germany) supplied with 10% fetal bovine serum (FBS), penicillin/streptomycin (100 U/mL) and GlutaMax (Life Technologies, Darmstadt, Germany). This preparation is from here on called complete medium, while the preparation without FBS is called serum-free medium.
These cells were transfected using FuGene HD (Roche, Mannheim, Germany) and a commercially available CRE-luciferase plasmid (pGL4.29[luc2P/CRE/Hygro]) from Promega (Heidelberg, Germany). After maintaining the cells with the transfection reagents overnight, the medium was supplemented with Hygromycin (250 µg/mL) and left for at least three days. The cells were then washed three times with sterile Hank's Balanced Salt Solution (HBSS; Gibco, Life Technologies, Darmstadt, Germany). Those cells still attached to the cell culture plate were enzymatically detached from the cell culture substrate using Accutase-Solution (Accutase; Sigma, Deisenhofen, Germany), counted (manually or by using Scepter 2.0, Merck-Millipore, Darmstadt, Germany) and supplied with medium containing the antibiotic selection agent Hygromycin (50 µg/mL; Enzo Lifescience GmbH, Lörrach, Germany). The surviving (Hygromycin-resistant therefore CRE-reporter-positive) cells were plated out for a selection of stably transfected clones. Mixed clones showing stable induction of CRE-luc-activity induced by forskolin application (10 µM; more than four-fold elevated luminescence in the presence of forskolin over vehicle control (1% DMSO)) were expanded and frozen at a density of 1.5 million/vial at −80 • C in freezing medium (IBIDI, Planegg, Germany). The resulting cell line was named SCNCRE. SCNCRE cells thawed from these preparations were given passage number 1 and used up to passage 13 without detectable changes in control experiments comparing forskolintreated with untreated cells. Cells were passaged once a week at a density of 1 million cells per 75 cm 2 flask.
After incubation with the first antibody, the membranes were washed four times for 2 min with TBS-T and incubated with the appropriate secondary HRP-coupled antibodies against rabbit (1:50.000; Santa Cruz Biotechnology, Heidelberg, Germany) or mouse (1:50.000; DAKO, Hamburg, Germany) in Rotiblock for 1 h at room temperature. Membranes were washed with TBS-T four times for 2 min and once for 5 min. Signal detection was performed using the chemiluminescent substrate Luminata forte (Millipore, Darmstadt, Germany) or AceGlow TM Ultrasensitive (PeqLab, Munich, Germany) and a CCD cameraequipped luminescence analysis system (Quantity One, ChemiDoc XRS, Bio-Rad, Hercules, CA, USA; or a Fusion XL system from PeqLab, Munich, Germany).

Immunocytochemistry
5.000 or 25.000 SCNCRE cells were plated on poly-L-lysine-coated glass coverslips (10 mm diameter) and treated with 10 µM forskolin or left untreated (control/1% DMSO) for one hour. For fixation, the medium was removed at the end of each experiment and replaced with phosphate buffer (0.1 M) containing 4% (w/v) paraformaldehyde (Morphisto, Frankfurt, Germany) for at least 15 min at room temperature. Samples were washed with 0.1 M phosphate-buffered saline (PBS) for 15 min and blocked for 45 min at room temperature using 5% (v/v) normal goat serum (both from Sigma-Aldrich, Darmstadt, Germany) PBS containing 0.3% (v/v) Triton-X-100 (blocking buffer). The cells were incubated overnight with primary antibodies in a blocking buffer (Benz et al., 2010). The primary antibodies used were anti-β-actin (1:10.000; Sigma-Aldrich, Deisenhofen, Germany) and anti-phospho-CREB (1:500) (Cell Signaling, Bad Nauheim, Germany). Diaminobenzidin (DAB) staining was performed according to a standard protocol as described [37,38]. For fluorescence staining, samples were washed three times for 10 min in PBS and incubated in fluorochromecoupled secondary antibodies against rabbit (Alexa-568; 1:2.500, Invitrogen) and mouse (Alexa-488; 1:500; Active Motif, Rixensart, Belgium) in blocking buffer. Coverslips were washed three times for 10 min in PBS, mounted on slides using VectaShield HardSet Antifade mounting medium with DAPI (Vector Laboratories, Newark, CA, USA) and stored at 4 • C. Intracellular distribution and fluorescence intensity of immunolabeled proteins was determined by fluorescence microscopy (Axio, Zeiss, Göttingen, Germany).

cAMP ELISA
Extracellular cAMP was determined in SCNCRE cells using serum-free, undiluted cell culture supernatant and a commercial ELISA (Cayman, Ann Arbor, MI, USA; Item no. 581001) according to the supplier's protocol as described before [36]. In brief, SCNCRE cells were plated either at 25.000 cells per well on a 96-well plate or 400.000 cells per well on a 24-well plate, left overnight in serum-free medium and exposed with the indicated substances for one hour. Subsequently, the medium was harvested and either frozen at −20 • C for later processing or directly subjected to the sample plate provided in the cAMP ELISA kit. The cAMP ELISA procedure was performed according to the supplier's manual, as previously described [36].

Determination of Luminescence Activity
If not indicated otherwise, 25.000 cells per well of a 96-well plate were plated in 100 µL volume and left in the incubator overnight in order to promote adhesion. Experiments were performed in 200 µL medium/well containing 0.25 mM Luciferin (unless indicated otherwise) and measured in a luminometer (BMG Lumistar; BMG, Ortenberg, Germany) or Berthold Centro LB960 (Berthold Technologies, Bad Wildbad, Germany) at 35 • Celsius for 0.1 s per well. Luminescence data are displayed as relative luminescence units (RLU). Cells were measured after different times of incubation (e.g., every 15 min) with and without chemical agents potentially influencing CRE-luciferase signalling unless indicated otherwise.

Image Analysis and Statistics
Images of protein bands were digitized either using a Bio-Rad Universal Hood equipped with a CCD camera, Quantity One and ChemiDoc XRS software (BioRad, München, Germany) or a Fusion system (Vilber/PeqLab; PeqLab/VWR, Erlangen, Germany). Signal intensities of the digitized images were analyzed using online open access ImageJ software (https://imagej.net, accessed on 27 July 2022) [39]. Each area/density value for a specific protein band was normalized against the corresponding β-actin signal of each extract. Statistical significance between groups was analyzed using one-way ANOVA, followed by the Tukey post-hoc test. The criterion of significance was p ≤ 0.05, with analysis performed using GraphPad Prism 8.4 (GraphPad, San Diego, CA, USA).

SCN Ex Vivo Sample Treatment
Mice were held under a schedule of 12 h of darkness and 12 h of light with unlimited food and water supply according to the veterinary laws of Germany. Animals were sacrificed at Zeitgeber time (ZT)10 (ten hours after lights on) and ZT14 (2 h after lights off. The brains were excised, and the SCNs were punched out with a hollow needle. The punches were immediately mixed with two times concentrated sample buffer homogenized by sonification and treated as described above for the SCNCRE cell extracts for Western blot analysis.

Informed Consent Statement: Not applicable.
Data Availability Statement: All raw data and calculations are available from the senior author.