CRTC2 activation in the suprachiasmatic nucleus, but not paraventricular nucleus, varies in a diurnal fashion and increases with nighttime light exposure

Abstract

Entrainment of the intrinsic suprachiasmatic nucleus (SCN) molecular clock to the light-dark cycle depends on photic-driven intracellular signal transduction responses of SCN neurons that converge on cAMP response element-binding protein (CREB)-mediated regulation of gene transcription. Characterization of the CREB coactivator proteins CREB-regulated transcriptional coactivators (CRTCs) has revealed a greater degree of differential activity-dependent modulation of CREB transactivational function than previously appreciated. In confirmation of recent reports, we found an enrichment of crtc2 mRNA and prominent CRTC2 protein expression within the SCN of adult male rats. With use of a hypothalamic organotypic culture preparation for initial CRTC2-reactive antibody characterization, we found that CRTC2 immunoreactivity in hypothalamic neurons shifted from a predominantly cytoplasmic profile under basal culture conditions to a primarily nuclear localization (CRTC2 activation) 30 min after adenylate cyclase stimulation. In adult rat SCN, we found a diurnal variation in CRTC2 activation (peak at zeitgeber time of 4 h and trough at zeitgeber time of 16–20 h) but no variation in the total number of CRTC2-immunoreactive cells. There was no diurnal variation of CRTC2 activation in the hypothalamic paraventricular nucleus, another site of enriched CRTC2 expression. Exposure of rats to light (50 lux) for 30 min during the second half of their dark (nighttime) phase produced CRTC2 activation. We observed in the SCN a parallel change in the expression of a CREB-regulated gene (FOS). In contrast, nighttime light exposure had no effect on CRTC2 activation or FOS expression in the paraventricular nucleus, nor did it affect corticosterone hormone levels. These results suggest that CRTC2 participates in CREB-dependent photic entrainment of SCN function.

optimization of all aspects of organismic performance according to the time of day is essential for normal physical and mental function (29, 42). The primary neural integration responsible for converting periodic light cues into perception of light-dark cycle parameters (period and phase) occurs within the suprachiasmatic nucleus of the hypothalamus (SCN) (37, 45). The SCN receives direct and indirect neural input from a special subset of photosensitive melanopsin-expressing retinal ganglion cells (7, 35). The SCN is also necessary for coordination of the body's diurnal function with the environmental light-dark cycle. However, the SCN does not merely transduce light-dark periodic information and pass that information onto the rest of body; the SCN independently generates an intrinsic circadian rhythm that is sustained, even in the absence of photic information (45). Consequently, the SCN serves as the body's master circadian pacemaker that normally operates in synchronization with the light-dark cycle.

Entrainment of SCN neural function by periodic photic information depends on a dynamic interaction between photic-driven intracellular signal transduction and the operational elements of the intrinsic molecular clock. The cAMP response element-binding protein (CREB) is a key intracellular activity-dependent transcription factor that couples photic neural input with altered gene expression and, thereby, mediates phase entrainment of the molecular clock (9, 22, 45). For example, Per1 core clock gene expression in the SCN is directly regulated by CREB (43). The immediate early gene c-fos is also regulated by CREB activation, and c-fos induction contributes to normal SCN entrainment (46). Several different activity-dependent changes within SCN neurons, such as increased intracellular calcium, increased cAMP, or activation of MAPK phospho-relay cascades (33, 45), have been shown to converge on CREB activation. In general, the transactivational capability of CREB is substantially enhanced by phosphorylation of CREB's Ser¹³³ residue and the subsequent binding of CREB with the coactivator CREB-binding protein (8). However, phosphorylation of Ser¹³³ is not always sufficient for CREB-dependent gene expression (6, 17, 25). CREB's action can also depend on other posttranslational modifications, as well as interaction with other coactivators (17, 28). There is some evidence that this is also the case for CREB's role in regulating light-induced phase shifts in the SCN (12).

Recently, three related mammalian CREB coactivator proteins, referred to as CREB-regulated transcriptional coactivators (CRTC1, CRTC2, and CRTC3; also known as transducers of regulated CREB activity 1, 2, and 3), have been identified (10, 14). These proteins bind to the basic leucine zipper domain of CREB and stabilize CREB's binding to a cAMP response element (27). In some cases, these coactivators are necessary for CREB-dependent function; in other cases, they can potentiate CREB activity independent of phosphorylation at Ser¹³³ (10). The ability of CRTC proteins to bind CREB is regulated by the shuttling of those proteins between the cytoplasm and the nucleus. Phosphorylation of specific CRTC protein sites induces strong association with 14-3-3 adaptor proteins primarily localized in the cytoplasm. Dephosphorylation of those sites results in translocation and retention of CRTC proteins in the nucleus, thereby allowing binding to CREB (40). The nuclear localization of CRTC proteins is positively regulated in response to increased intracellular calcium (calcineurin-mediated) and increased cAMP (PKA-mediated) and negatively regulated by AMP kinase and several other related kinases (41). Thus CRTC function is strongly activity-dependent and, consequently, may contribute an important level of intracellular signal transduction coordination to CREB-mediated gene regulation.

Although CRTC1 is the most abundant CRTC family member in rodent brain, CRTC2 has selectively elevated gene expression within the hippocampus and the paraventricular nucleus (PVN) and SCN subregions of the hypothalamus (44). Only a few studies have examined the in vivo role of CRTC2 in the hypothalamus (16, 23, 26, 39). Most of those studies have focused on its role in regulating corticotropin-releasing hormone (CRH) gene expression in the PVN (16, 24, 26) and energy balance coordination in the arcuate nucleus (23). As we set out to also study the role of CRTC2 in the PVN, we became intrigued by the heightened expression of CRTC2 in the SCN and the possible implications for SCN function. Thus, using an antibody that we characterized in initial hypothalamic organotypic culture studies, as well as a second antibody validated by others as selective for CRTC2 (39), we have compared CRTC2 immunoreactivity within the SCN and the PVN of the adult male rat brain. If CRTC2 plays a role in the light entrainment process of the SCN, we hypothesized that the expression level and/or activation level (nuclear localization) of CRTC2 would vary in a diurnal fashion in the SCN. Furthermore, we tested the possibility that brief exposure of rats to light during the second half of their dark period would activate CRTC2 within the SCN and that this response may be selective for CRTC2 in the SCN compared with the PVN.

MATERIALS AND METHODS

Hypothalamic Organotypic Cultures

Timed-pregnant female Sprague-Dawley rats (Harlan Laboratories, Indianapolis, IN) were housed individually at the University of Colorado vivarium and maintained on a 12:12-h light-dark cycle (lights on at 0600) with ad libitum access to rat chow and water. Dams were allowed 1 wk to acclimate to the facility. On postnatal day 10, rat pups were decapitated under sterile conditions, and brain tissue was extracted and placed into Hanks' balanced salt solution containing glucose. Each pup was processed individually to completion. The entire brain was blocked by bilateral sagittal cuts at the hypothalamic sulci, and the cortex was removed by a single transverse cut. The block was cut coronally on a McIlwain tissue chopper (Stoelting) to a thickness of 350 μm. Two sections containing the majority of the PVN were isolated, trimmed, and placed on a porous membrane (PICMORG50, Millipore). The pituitary was also isolated from each pup and placed on the membrane between the two sections containing PVN at the ventral surface of the sections. Cultures were maintained for 13 days in vitro at 35°C and 5% CO₂ in medium containing 50% Eagle's basal medium, 25% Hanks' balanced salt solution, 25% heat-inactivated horse serum, 0.45% glucose, 2 mM l-glutamine, 10 mM HEPES, and penicillin-streptomycin. Medium was replaced every other day.

Test day manipulation.

After 13 days in vitro, cultures were switched to a serum-free medium consisting of Neurobasal medium supplemented with B27 (Invitrogen), 0.45% glucose, 2 mM l-glutamine, 10 mM HEPES, 1 mM sodium pyruvate, and penicillin-streptomycin. The adenylate cyclase stimulator forskolin (Calbiochem) was made up at a stock concentration of 100× (1 mM) in 5% DMSO and diluted in fresh serum-free medium just prior to use. Cultures were treated with vehicle or 10 μM forskolin for 30 min. After treatment, cultures were washed with PBS, fixed with 4% paraformaldehyde (PFA) in PBS at 4°C for 90 min, washed with PBS, and stored in cryoprotectant (30% sucrose, 30% propylene glycol, and 1% polyvinylpyrrolidone in 0.01 M PBS, pH 7.4) at −20°C until they were further processed by immunohistochemistry.

Immunohistochemistry for CRH and CRTC2.

Sections were rinsed in PBS, incubated with 0.1% sodium borohydride in PBS for 15 min to quench background autofluorescence, washed in PBS, incubated with 10% normal goat serum + 0.1% Triton X-100 + 1% BSA + 3% nonfat milk in PBS for 2 h, incubated with primary antibody [polyclonal rabbit anti-CRTC2 (1:2,000 dilution; catalog no. ST1099, Calbiochem) or polyclonal guinea pig anti-CRH (1:10,000 dilution; catalog no. T-5007, Peninsula Laboratories)] in blocking solution at 4°C for 48 h, washed in PBS, incubated with secondary antibody [goat anti-rabbit Alexa Fluor 594-conjugated (1:1,000 dilution) or goat anti-guinea pig Alexa Fluor 488-conjugated (1:500 dilution); Invitrogen] at room temperature for 1 h, washed in 0.01 M phosphate buffer, and mounted on Superfrost Plus slides, and coverslips were applied using Prolong Gold (Invitrogen). Organotypic cultures were maintained on the Millipore membrane throughout the processing and mounting procedures.

CRTC2 immunostaining analysis.

Image capture, quantification, and analysis were performed by an individual blind to test group assignment of explant sections. PVN explant images were captured and digitized with an inverted multiphoton scanning confocal laser microscope (FluoView FV1000, Olympus). Two-channel color photomicrographs were captured from each of two bilateral PVN regions per explant for a total of four PVN images per explant with use of imaging software (FluoView, Olympus). Neurons were considered CRH-immunopositive if a clearly identifiable soma was present. In the case of CRTC2 immunofluorescence, each CRH-immunopositive cell was qualitatively scored as containing weak or strong nuclear immunofluorescence, and the percentage of CRH-immunopositive neurons containing nuclear CRTC2 was calculated. Only CRH-immunopositive neurons with strong nuclear CRTC2 immunofluorescence were categorized as containing nuclear CRTC2.

Animals and Procedures

Young adult male Sprague-Dawley rats were obtained from Harlan Laboratories. Rats (260–300 g body wt at the time of experimentation) were housed in pairs in polycarbonate tubs (47 × 23 × 20 cm) with ad libitum access to water and rat chow. The experimental procedures took place within a customized animal procedure suite, with rats housed in four separate interior sound-attenuated rooms (maintained at 22 ± 2°C). Each interior room had its own air supply and exhaust and its own independent overhead fluorescent light fixture (ambient light intensity at cage location was ∼50 lux) and light timer control. The interior rooms were adjacent to an outer procedural space, where rats could be rapidly transported for euthanization and tissue collection without disturbance of other rats within their interior housing rooms. The interior housing rooms were maintained on a 12:12-h light-dark cycle, with staggered light phase onsets to provide for simultaneous tissue collection from rats at different points in their light-dark cycle phase. Rats were allowed a 2-wk acclimation period between their arrival and the experimental test day. To assess whether basal levels of CRTC2 immunoreactivity in rat SCN and PVN varied over the course of the day, rats were killed at six evenly spaced zeitgeber times (ZT) throughout the day (n = 4 rats at each of 6 time points). To determine whether short-term (30 min) ambient light (50 lux) exposure during the second half of the dark phase had an effect on CRTC2 and FOS immunoreactivity, an experiment with three groups of rats was conducted (n = 9: 6 brains from each treatment group for image analysis-assisted CRTC2-immunoreactive cell counts/nuclear localization and 3 brains from each treatment group for qualitative comparison of CRTC2 immunostaining with 2 different CRTC2-reactive antibodies). Two groups of rats that were left undisturbed in their home cage and housing room were euthanized at ZT4 or ZT20. A third group of rats was exposed to 30 min of interior housing room ambient light starting at ZT19.5 prior to euthanization at ZT20. All husbandry, handling, and use of animals were conducted in accordance with ethical guidelines in the “Guide for the Care and Use of Laboratory Animals” [DHHS Publication No. (NIH) 80-23, revised 2010, 8th edition] and were approved by the University of Colorado's Institutional Animal Care and Use Committee.

Crtc2 mRNA In Situ Hybridization

Crtc2 mRNA expression throughout adult male rat forebrain under light phase (ZT4) basal conditions was examined using in situ hybridization according to previously published procedures (13). Coronal brain cryosections (10 μm) were thaw-mounted onto electrostatically charged glass microscope slides (ColorFrost Plus) and stored at −80°C. The Crtc2 mRNA riboprobe was transcribed (T7 polymerase) from a linearized (EcoRV restriction endonuclease) Crtc2 expression plasmid (pSC-A-amp/kan, StrataClone) generated in-house from rat brain total RNA. A 358-bp nucleotide sequence that corresponds to nucleotides 203–560 of rat Crtc2 mRNA (National Center for Biotechnology Information Ref Seq NM_001033895) was cloned into the expression plasmid. The identity of the cloned Crtc2 DNA insert was verified by DNA sequencing. There was no detectable autoradiographic signal above background for test of a Crtc2 sense riboprobe on rat brain tissue.

Tissue and Blood Collection

Rats were taken directly from their interior home rooms to the adjacent procedural space and guillotine-decapitated within 30 s after removal from their home cage. Red light conditions were used in the procedural space for decapitation of rats within their dark-phase condition. Trunk blood was collected into EDTA-coated Vacutainers (Becton Dickinson, Franklin Lakes, NJ), and plasma was stored at −80°C.

Because of the rapid activation of intracellular signal transduction pathways, which can occur in the brain as a result of the influence of handling and anesthesia necessary for transcardial perfusion of rats with fixative (18, 19, 34), we instead used decapitation followed by a postfixation method for brain tissue preservation suitable for immunohistochemical analysis, as adapted from Khan et al. (18). Brains were extracted from the skull, and coronal blocks containing the rostral-to-caudal extent of the hypothalamus were placed in vials containing 4% buffered PFA (pH 7.4) for ∼10 h at 4°C on a shaker to ensure penetration of PFA. The brain sections were transferred into vials containing 30% sucrose dissolved in 0.1 M PBS and shaken for 48 h (4°C). Brain tissue was then frozen on powdered dry ice and stored at −80°C. Brains were cut via cryostat (model 1850, Leica Microsystems, Nussloch, Germany) at −24°C into 25-μm-thick coronal sections (4 series of 4 sections each) and stored in cryoprotectant (30% sucrose, 30% propylene glycol, and 1% polyvinylpyrrolidone in 0.01 M PBS buffer, pH 7.4) until subsequent immunohistofluorescence processing.

Plasma Corticosterone Measurement

Total plasma corticosterone was measured using a commercial ELISA kit (Enzo Life Sciences, Farmingdale, NY). Plasma was diluted 1:50 in assay buffer and heated for 1 h at 60°C to denature corticosteroid-binding globulin. Diluted samples were run in duplicate, and the average intra- and interassay coefficients of variation were 9.6% and 2.5%, respectively.

Semiquantification of CRTC2 and FOS Immunoreactivity in Postmortem Rat Brain Tissue

Immunohistofluorescence procedure.

We used a double-label immunofluorescence procedure, with DAPI counterstain, to examine CRTC2- or FOS-immunoreactive cells (1st label) within the hypothalamic SCN or PVN. Vasoactive intestinal peptide (VIP) antibody (1:5,000 dilution, guinea pig polyclonal antibody; catalog no. T-5030, Peninsula Laboratories) (4) and CRH antibody (1:15,000 dilution, guinea pig polyclonal antibody; catalog no. T-5007, Peninsula Laboratories) (34) were used as regional markers (2nd label) for the SCN and PVN, respectively. We used the same CRTC2 antibody that we characterized in hypothalamic organotypic cultures (1:4,000 dilution, rabbit polyclonal antibody; catalog no. ST1099, Calbiochem Laboratories) (24, 26), and we obtained the FOS antibody (1:5,000 dilution, rabbit polyclonal antibody, catalog no. sc-52) from Santa Cruz Biotechnology. PBS (0.01 M) was used as assay buffer (pH 7.4) for all incubation and wash steps except the final wash steps, for which 0.01 M phosphate buffer was used. Tissue was incubated in buffered 0.1% sodium borohydride for 30 min and then placed for 1 h in a buffered blocking solution (5% normal goat serum, 3% dry milk, 1% bovine serum albumin, and 0.3% Triton X-100). Primary antibodies were incubated simultaneously for ∼24 h with shaking at 4°C. After 24 h, tissue was incubated in both secondary antibodies simultaneously for 1.5 h at room temperature. The secondary antibody for detection of CRTC2 and FOS antibodies was a goat anti-rabbit Ig antibody that was conjugated with Alexa Fluor 594 (1:1,000 dilution; catalog no. A11012, Invitrogen Molecular Probes). The secondary antibody for detection of CRH and VIP antibodies was a goat anti-guinea pig Ig antibody that was conjugated with Alexa Fluor 488 (1:1,000 dilution; catalog no. A11073, Invitrogen Molecular Probes). Control sections incubated without each combination of primary and secondary antibodies were also included in each assay. Before the final phosphate buffer washes, tissue was incubated in dilute (1:5) DAPI nuclear staining solution for 3 min. Sections were mounted in a 1% glycerol phosphate-buffered solution onto ColorFrost Plus slides with Fluoromount-G reagent (SouthernBiotech, Birmingham, AL), and coverslips were applied.

Image analysis.

Image capture, quantification, and analysis were performed by an individual blind to test group assignment of tissue sections. PVN and SCN images were captured and digitized with a Zeiss microscope (×400 magnification; AxioImager M1, Zeiss, Oberkochen, Germany). Three-channel color epifluorescent (mercury bulb illumination) photomicrographs were captured from at least four brain sections per region of interest (ROI) and per animal with use of imaging software (AxioVision, Zeiss). For each assay, tissue from a single collection series per animal (75-μm spacing between serial sections) was used. Neurons were considered CRH-immunopositive if a clearly identifiable soma was present, whereas VIP immunofluorescence, which consisted of some prominent labeled soma, as well as a surrounding rich plexus of nerve fibers, was used as a marker for the presence of the SCN on the tissue section. For each SCN-containing tissue section, ROIs were drawn around the cell-dense region of the SCN. For each PVN-containing tissue section, ROIs were drawn around the CRH immunofluorescence portion of the PVN. With use of AxioVision software, all CRTC2- or FOS-immunopositive cells within the SCN or those colocalized with CRH immunofluorescence within the PVN were marked, counted, and expressed as number of cells per ROI area (density). In the case of CRTC2 immunofluorescence, each marked cell that had a visible nucleus (DAPI staining) was qualitatively scored as containing either weak or strong nuclear immunofluorescence. FOS immunofluorescence was exclusively localized within the cell nucleus.

Immunohistofluorescence Procedure for Qualitative Comparison of Two Different CRTC2 Antibodies

Tissue from a subset of brains taken from rats killed at ZT4 or ZT20 ± 30 min light exposure was used for a qualitative comparison of immunostaining in the SCN by two different CRTC2-reactive antibodies. For this assessment, we compared the Calbiochem CRTC2 antibody (catalog no. ST1099), used for the experiments described above, with a second commercial CRTC2 antibody (catalog no. 12497-1-AP, Proteintech Group, Chicago, IL) that has been validated in vitro to label cells that express CRTC2, but not CRTC1. Without use of an amplification process, we obtained an overall weak immunofluorescence signal with the Proteintech CRTC2 antibody. Consequently, for this direct antibody comparison, we used a tyramide signal amplification (TSA) procedure (TSA Plus Fluorescent Cyanine 3 Kit, PerkinElmer, Waltham, MA). For this comparison, tissue was counterstained with the nuclear stain DAPI, but we did not include a VIP-immunoreactive double label, so as to rule out any possible artifactual signal within the CRTC2 imaging channel due to secondary antibody-tyramide cross-reactivity or wavelength “bleed through.” The primary antibody concentrations were 1:8,000 for the Calbiochem anti-CRTC2 antibody and 1:1,000 for the Proteintech anti-CRTC2 antibody. Tissue was incubated for 30 min in 1% sodium borohydride diluted in 0.01% PBS, rinsed, and then incubated for 30 min in 2% hydrogen peroxide diluted in 0.01% PBS. The tissue was rinsed and then switched to 0.1 M Tris-buffered saline as diluent for the remainder of the assay (with 0.05% Tween 20 added for rinse steps). Tissue was incubated for 1 h in the blocking solution (0.5% in Tris-bufferred saline) provided with the TSA Plus kit and then incubated overnight (4°C) in primary antibody diluted in blocking solution. The tissue was rinsed and then incubated for 1 h at room temperature in anti-rabbit horseradish peroxidase-conjugated antibody (1:5,000 dilution; Novex, Life Technologies, Grand Island, NY) diluted with the blocking solution. The tissue was rinsed and then incubated for 30 min at room temperature in Cy3 TSA (1:100 dilution) in TSA diluent. The tissue was rinsed and then incubated in DAPI nuclear staining solution (1:5 dilution) for 3 min and mounted on gelatin-subbed slides with Fluoromount-G reagent, and coverslips were applied. Tissue was viewed and digitized with an inverted multiphoton scanning confocal laser microscope (FluoView FV1000, Olympus). Two-channel color photomicrographs were captured from SCN-containing sections from each treatment group viewed with a ×60 oil objective and using imaging software (FluoView, Olympus).

Statistical Analysis

For statistical tests of group differences, Student's t-test or one-way ANOVA was performed using the SPSS statistical analysis program 10.5 for Macintosh operating system (Chicago, IL). Tukey's test was used for post hoc tests of pairwise group comparisons of interest. Cosinor analysis was used for determination of 24-h periodicity of data, as well as for parameter estimates of amplitude and acrophase (Chronolab software 3.0) (31, 36). A significance level of P ≤ 0.05 was chosen for all statistical tests. Values are means ± SE.

RESULTS

In Vitro Characterization of CRTC2 Antibody Use in Rat Hypothalamic Neural Tissue

Recent reports indicate a relative abundance of crtc2 gene and protein expression within the PVN region of the rat hypothalamus (26, 44). To evaluate the subsequent use of a CRTC2-reactive antibody for immunohistochemical assessment of CRTC2 activity in rat hypothalamus, we used a hypothalamic organotypic culture model. We found that we obtained healthy hypothalamic tissue cultures with strong CRH neuronal expression when we cocultured the hypothalamic tissue with neonatal pituitaries, perhaps due to trophic factors secreted by the pituitary (Fig. 1A). There was extensive CRTC2 immunoreactivity throughout the hypothalamic tissue cultures, with some visible enrichment within the PVN that included some colocalization within CRH-immunoreactive neurons (Fig. 1, B and C). This CRH neuron colocalization confirmed neuronal expression of CRTC2. The majority of the CRTC2 immunoreactivity under our basal culture conditions was cytoplasmic. Acute (30 min) treatment of the cultures with forskolin produced predominantly nuclear CRTC2 immunoreactivity throughout the tissue, including a significant increase (P = 0.02, Student's t-test) in the percentage of CRH-immunoreactive neurons containing nuclear CRTC2 compared with vehicle-treated explants (Figs. 1C and 2). These in vitro studies confirmed the ability of this CRTC2 antibody to recognize the cytoplasmic and nuclear forms of CRTC2 in rat hypothalamic neurons and our ability to detect the redistribution of CRTC2 intracellular localization with acute cellular activation.

Fig. 1.

Corticotropin-releasing hormone (CRH) and cAMP response element-binding protein (CREB)-regulated transcriptional coactivator 2 (CRTC2) expression in paraventricular nucleus (PVN)-containing hypothalamic organotypic tissue culture. A–C: confocal images from Olympus FV1000 multiphoton scanning confocal laser microscope. A: coculture of PVN-containing tissue explants with pituitary resulted in robust CRH neuron expression. Note dense CRH immunostaining in cell bodies surrounding the dorsal portion of the third ventricle (3V) and the presence of CRH-immunoreactive processes that extend laterally and ventrally within the explant. B: CRTC2-immunoreactive cells (red immunofluorescence) were present throughout the PVN portion of the hypothalamic tissue culture, including overlap with CRH neuron localization (green immunofluorescence). C: after forskolin treatment, colocalization of CRTC2 immunoreactivity (red immunofluorescence) within the nuclei of some CRH neurons (green immunofluorescence) was readily apparent.

Fig. 2.

Shift from cytoplasmic to nuclear CRTC2 immunoreactivity with 30 min of forskolin treatment of hypothalamic organotypic culture. A: 30 min of forskolin (10 μM) treatment significantly increased the percentage of CRH neurons with nuclear CRTC2 immunoreactivity. *P = 0.02 (by Student's t-test). B and C: confocal images from Olympus FV1000 multiphoton scanning confocal laser microscope. B: vehicle-treated culture showing diffuse and predominantly cytoplasmic CRTC2 immunoreactivity overlapping with several CRH-immunoreactive neurons. C: forskolin-treated culture showing distinct CRTC2 nuclear immunoreactivity within CRH-immunoreactive neuron (higher magnification of Fig. 1C). Red, CRTC2 immunofluorescence; green, CRH immunofluorescence. Scale bar = 10 μm.

Crtc2 mRNA Expression and CRTC2 Immunoreactivity Within SCN of Adult Rat

Consistent with a previous report (44), we found that Crtc2 mRNA expression in adult rat brain was relatively weak and diffuse, with the notable exception of enhanced expression in the hippocampus and in the PVN and SCN portions of the hypothalamus (Fig. 3, A and B). In initial tests of CRTC2 immunoreactivity in adult rat brain tissue, we found that the CRTC2 protein was also somewhat enriched within the SCN and PVN compared with the surrounding hypothalamus. CRTC2 immunoreactivity was present throughout the SCN and in the medial and lateral portions of the PVN (Fig. 3, G and H). Interestingly, for these initial tests, we used tissue taken from rats during the first half of their light-dark cycle, and we found that a greater proportion of CRTC2 immunoreactivity within the SCN was localized within cell nuclei than within the PVN (unpublished observations).

Fig. 3.

Crtc2 mRNA and CRTC2 protein expression in suprachiasmatic nucleus (SCN) and PVN hypothalamic subregions. Crtc2 mRNA (in situ hybridization) was selectively prominent within the SCN (A) and PVN (B) hypothalamic subregions (sample autoradiograms). CRTC2 immunoreactivity (red immunofluorescence) was also prominent within cells throughout the SCN [VIP immunostaining (C), CRTC2 immunostaining (D), and merged image of C and D (E)] and PVN [CRH immunostaining (F), CRTC2 immunostaining (G), and merged image of F and G (H)]. Epifluorescent images (C–H) were captured with Zeiss AxioImager M1 microscrope. oc, Optic chiasm; 3v, 3rd ventricle. Scale bar = 100 μm.

Diurnal Profile of CRTC2 Immunoreactivity in SCN and PVN

To assess whether CRTC2 activation, as assessed by nuclear localization, varies in a diurnal fashion in the SCN and PVN, we examined CRTC2 immunoreactivity across a 24-h period in rats that were left undisturbed in their home cage and maintained on a 12:12-h light-dark cycle. We found a significant diurnal rhythm of percent nuclear CRTC2 immunoreactivity within the SCN (cosinor analysis with 24-h period, P < 0.05), with peak nuclear localization around ZT4 and a trough around ZT16 and ZT20 (Fig. 4, A and B). Parameter estimates (±SE) for this 24-h rhythm yielded an amplitude of 15.0% (±5.0%) and an acrophase of ZT5.8 (±1.4). There was no significant difference across the day in the total number of CRTC2-immunoreactive neurons within the SCN (data not shown). In contrast to the SCN, we found no significant diurnal variation in the percentage of nuclear CRTC2-immunoreactive neurons in the PVN (Fig. 4C). As expected, in these rats we observed a normal diurnal variation in basal plasma corticosterone levels (Fig. 4D), with a pronounced peak at the onset of the dark phase (ZT12).

Fig. 4.

Diurnal variation in CRTC2 activation in the SCN, but not the PVN. A and C: significant 24-h rhythm (cosinor analysis) of percentage of cells with pronounced nuclear CRTC2 immunoreactivity in the SCN, but not the PVN. B: epifluorescent images from Zeiss AxioImager M1 microscrope. Representative SCN photomicrographs (ventral lateral SCN) illustrate a higher proportion of cells with distinct nuclear CRTC2 immunostaining at zeitgeber time (ZT) of 4 h (ZT4) than ZT20. Scale bar = 50 μm. D: as expected, there was a pronounced diurnal rhythm of plasma corticosterone (CORT) in these rats that peaked at the onset of the rats' active period (ZT12). Values for ZT0 were double-plotted at ZT24 to aid in visualization of the 24-h time course for each measure.

Effect of Light Exposure During the Second Half of the Dark Phase on CRTC2 and FOS Immunoreactivity in SCN and PVN

Light entrainment of the SCN molecular clock depends on activation of key intracellular signal transduction molecules that converge on CREB (9, 45). A hallmark feature of this entrainment process is the activation of these signal transduction molecular pathways when a light pulse occurs during the subjective night (33). To test whether CRTC2 activation within the SCN may be sensitive to photic input during the second half of the dark phase, we exposed some rats to 30 min of ambient housing room light beginning at ZT19.5, the approximate time of trough levels of SCN nuclear CRTC2 immunoreactivity. CRTC2 immunoreactivity of these nighttime-light-exposed rats was compared with tissue taken from rats at the peak of nuclear CRTC2 immunoreactivity (ZT4) and with other rats at ZT20 that were not exposed to nighttime light. For comparison purposes, we also examined FOS expression. FOS induction is one marker of photic activation of SCN neurons, and its photic induction is likely to be downstream from CREB activation (20, 38).

Consistent with the circadian profile of nuclear CRTC2 immunoreactivity that we observed in the previous experiment, there was a greater percentage of nuclear CRTC2 immunoreactivity in the SCN of rats at ZT4 than at ZT20. Importantly, 30 min of light exposure starting at ZT19.5 increased nuclear CRTC2 immunoreactivity to a level that was not significantly different from that at ZT4 (Figs. 5A, 6, and 7). Thus there was an overall significant effect of treatment on CRTC2 immunoreactivity in the dorsal medial (F_2,15 = 10.04, P < 0.005) and ventral lateral (F_2,15 = 10.94, P < 0.001) zones of the SCN, and the pattern of effect was similar for both zones. In contrast, there was no time-of-day or light effect on the percentage of CRTC2 nuclear immunoreactivity in the PVN (Fig. 5C).

Fig. 5.

Exposure to light during nighttime increased CRTC2 activation and FOS expression in the SCN, but not the PVN. A and B: exposure of rats to moderate-intensity (50 lux) light for 30 min, starting at ZT19.5, significantly increased the percentage of cells with CRTC2 nuclear immunostaining in the dorsal medial and ventral lateral portions of the SCN and increased the number of FOS-immunoreactive cells throughout the SCN. Representative epifluorescent images from Zeiss AxioImager M1 microscrope of CRTC2 and FOS immunostaining in ventral lateral SCN are shown below the respective treatment group bar graph. Scale bars = 50 μm. C–E: nighttime light exposure had no effect on CRTC2 or FOS immunostaining in the PVN or on plasma corticosterone (CORT) levels. *P < 0.05 vs. ZT4 (by Tukey's test).

Fig. 6.

Time of day- and nighttime light exposure effects on Calbiochem anti-CRTC2 antibody immunostaining pattern in SCN. Representative confocal images from Olympus FV1000 multiphoton scanning confocal laser microscope show greater SCN CRTC2 nuclear cellular localization (overlap with DAPI nuclear stain) at ZT4 and after nighttime light exposure than without light exposure at ZT20. Scale bar = 50 μm. Note diffuse CRTC2 immunostaining surrounding many DAPI-stained nuclei at ZT20 without light exposure.

Fig. 7.

Time of day- and nighttime light exposure effects on Proteintech anti-CRTC2 antibody immunostaining pattern in SCN. Representative confocal images from Olympus FV1000 multiphoton scanning confocal laser microscope show greater SCN CRTC2 nuclear cellular localization (overlap with DAPI nuclear stain) at ZT4 and after nighttime light exposure than without light exposure at ZT20. Scale bar = 50 μm. Note diffuse CRTC2 immunostaining surrounding many DAPI-stained nuclei at ZT20 without light exposure.

In confirmation of the findings of others, we also found that the nighttime light exposure led to a large increase in the number of FOS-immunopositive cells in the SCN (overall treatment effect: F_2,14 = 6.64, P < 0.01; Fig. 5B). In contrast, the light exposure did not increase FOS expression in the PVN, nor did it increase plasma corticosterone levels (Fig. 5, D and E). Thus the nighttime light exposure did not induce an activation of the hypothalamic-pituitary-adrenal axis, as might be expected if it produced an increase in arousal or a stress response (11). Interestingly, there was a trend for the nighttime light exposure to decrease PVN FOS expression levels and plasma corticosterone.

Direct Comparison of the SCN Immunostaining Profile of Two Different CRTC2-Reactive Antibodies

We examined whether a commercially available CRTC2-reactive antibody (Proteintech) that was different from the Calbiochem antibody used for the above-described experiments would produce a similar profile of differential nuclear immunostaining that varied with time of day and nighttime light exposure. The Proteintech anti-CRTC2 antibody gave overall a weaker signal-to-noise ratio with our immunostaining conditions, but it showed the same general profile as the Calbiochem anti-CRTC2 antibody (Fig. 6), with predominantly nuclear localization at ZT4, much less nuclear localization at ZT20, and an intermediate level of nuclear localization after 30 min of light exposure at ZT20 (Fig. 7). Interestingly, for both antibodies, despite the TSA amplification, the immunostaining pattern at ZT20 was very diffuse within the SCN, with indistinct demarcation of individual cell processes. Thus there appeared to be a cloud of diffuse immunostaining restricted to the region of the SCN, with very few cells showing CRTC2 immunostaining within the cell nucleus.

DISCUSSION

We found prominent Crtc2 mRNA and CRTC2 protein expression within the SCN of adult male rats, in confirmation of previous reports (39, 44). Although the number of CRTC2-immunoreactive cells did not change in a diurnal fashion within the SCN, we observed diurnal changes in the number of SCN cells that had strong nuclear localization of CRTC2 immunoreactivity (i.e., CRTC2 activation). We observed peak CRTC2 activation during the first third of the light phase and a trough of activation during the midportion of the dark phase. Exposure of rats to a moderate level (50 lux) of ambient light for 30 min during the second half of the dark phase produced an activation of CRTC2 to levels similar to those observed during the peak of CRTC2 diurnal activation. The diurnal variation and photic responsiveness of CRTC2 immunoreactivity were not observed in the PVN, another hypothalamic region that has prominent CRTC2 expression (24). This suggests that the circadian-relevant changes in CRTC2 activity in the SCN do not reflect a more widespread fluctuation of CRTC2 activity throughout the brain in response to diurnal changes in metabolic state or neural arousal. Instead, CRTC2 activity in the SCN may selectively participate in the photic entrainment of the SCN molecular clock.

Two recent reports also support a role of CRTC1 in light entrainment of SCN function (15, 39). Sakamoto et al. (39) recently reported that CRTC1 immunoreactivity fluctuates in the SCN of mice in a circadian fashion. In contrast to CRTC2 immunoreactivity, they found that CRTC1-immunoreactive cell counts were higher during the subjective day than subjective night and that a 5-min light pulse (40 lux) during the subjective night rapidly increased CRTC1-immunoreactive levels and the extent of nuclear CRTC1 immunoreactivity. Similar to our study, they did not see diurnal or light pulse changes in the number of CRTC2-immunoreactive cells; however, they also did not observe qualitative differences in the nuclear immunostaining pattern of CRTC2. The failure of Sakamoto et al. to see photic effects on CRTC2 nuclear immunoreactivity in the SCN may be due to a difference in the species studied or differences in the tissue collection method and immunohistochemistry procedure. In contrast to the study of Sakamoto et al., we did not anesthetize and perfuse our animals at the time of euthanization. Instead, rats were killed by guillotine decapitation, and brain tissue was prepared for immunohistochemical analysis following postfixation treatment. We and others have found that, in some cases, the anesthetization-and-perfusion process is sufficient to activate intracellular signaling molecules, most notably phosphorylated ERK1/2 (19, 34), but this may also be the case for CRTC2 (26). For our hypothalamic organotypic culture and semiquantitative analysis of CRCT2 immunoreactivity in rat brain, we used a commercially available CRTC2-reactive antibody (Calbiochem) different from that used by Sakamoto et al. (Proteintech). Both antibodies are directed against the same carboxy-terminal region of the CRTC2 protein, the portion of the protein that has limited sequence similarity with CRTC1 and CRTC3. When directly comparing the SCN immunostaining profiles of the Calbiochem and Proteintech CRTC2-reactive antibodies on experimental tissue, we found qualitatively a very similar profile of differential immunostaining within cell nuclei that varied with time of day and presence of nighttime light exposure. Thus it is likely that the combined tissue-processing and immunohistochemistry conditions of our study contributed to greater sensitivity in detecting changes in intracellular localization of CRTC2. Our hypothalamic organotypic culture study further validated our ability to detect with the Calbiochem antibody a redistribution of CRTC2 from the cytoplasm to the nucleus after acute adenylate cyclase stimulation. Thus our results indicate that CRTC2, in conjunction with CRTC1 (15, 39), is likely to contribute to CREB-dependent photic entrainment of SCN function. Although we did not examine in this study whether CRTC2 activation within the SCN would vary in a circadian manner in the absence of a light-dark cycle, Sakamoto et al. found that such was the case for CRTC1. Whether CRTC1 and CRTC2 proteins are colocalized within the same SCN cells and provide redundant or differential function remains to be determined.

The phase relationship that we observed for CRTC2 diurnal activation and photic responsiveness parallels the well-documented intracellular responses of SCN neurons that converge on CREB activation (e.g., increases in cAMP, intracellular calcium, and phosphorylated ERK) and the subsequent induction of several immediate early gene proteins, including FOS (9, 20, 33, 38, 45). We also saw parallel diurnal and light exposure changes in FOS expression that were selective for the SCN. Whether CRTC proteins are necessary CREB coactivators for photic entrainment of the SCN molecular clock has not been assessed. A recent study suggests that regulation of CRTC activity in the SCN by activity-dependent induction of one of CRTC's AMP kinase-related negative regulators (salt-inducible kinase) serves as a mechanism to constrain the duration of photic responses in the SCN (15). Another prospect is that the CRTC proteins contribute to the multifactorial aspect of SCN entrainment to photic and nonphotic cues. Whereas the light-dark cycle provides the dominant entraining influence on SCN function, a number of other factors either moderate the photic response or contribute independent influences on molecular clock function (5, 21). Examples of nonphotic influences on SCN operation include food availability, social interactions, wheel-running activity, methamphetamine treatment, and metabolic state (2, 5). Because of the complex activity dependence of CRTC protein function, which includes multiple positive and negative regulators (41), CRTC protein activation is likely to contribute to the molecular integration of photic and nonphotic information that impacts on SCN molecular clock operation. CRTC proteins may have a key role in the influence of abnormal metabolic state on SCN function. For example, a high-fat diet has been shown to interfere with SCN light entrainment (30). CRTC1 and CRTC2 protein activation varies in the arcuate nucleus of mice in response to glucose administration or alterations in the fasting/fed state of the animal (1, 23). Glucocorticoid hormone levels have also been found to modulate CRTC2 activation in the PVN (16), but this may not be a factor in regulating SCN CRTC2 activity because of the apparent low or absent glucocorticoid receptor expression within the SCN (3, 32).

In conclusion, the CREB coactivator CRTC2 has somewhat enriched expression within the SCN compared with many other brain regions, and its activation profile also parallels CREB-dependent photic entrainment of the SCN molecular clock. Because of the multifactor nature of CRTC1 and CRTC2 activation, these CREB coactivators may also importantly contribute to the modulatory influence of nonphotic cues on this entrainment process.

GRANTS

This work was supported by National Institute of Mental Health Grants MH-75968 and MH-065977 and by the University of Colorado Biological Sciences Initiative.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.A.H., M.J.W., L.R.H., and R.L.S. are responsible for conception and design of the research; J.A.H., M.J.W., L.R.H., and R.L.S. performed the experiments; J.A.H., M.J.W., L.R.H., and R.L.S. analyzed the data; J.A.H., M.J.W., L.R.H., and R.L.S. interpreted the results of the experiments; J.A.H., M.J.W., L.R.H., and R.L.S. prepared the figures; J.A.H., M.J.W., and R.L.S. drafted the manuscript; J.A.H., M.J.W., L.R.H., and R.L.S. edited and revised the manuscript; J.A.H., M.J.W., L.R.H., and R.L.S. approved the final version of the manuscript.