Per2-dluc transgenic rats (43), Per2^Luc knock-in mice (44), and Clock ^Δ19/Δ19 mice (45) were bred and genotyped as described previously (46–48). Animals were housed under a regular 12-h light/12-h dark (LD) cycle, maintained at 22 ± 2°C, with free access to food and water. All animal experiments were conducted in compliance with the Ethical Regulations of Kyoto University and performed under protocols approved by the Animal Care and Experimentation Committee of Kyoto University.

Adrenals from Per2^Luc mice (male, 8–12 weeks old) were harvested between ZT5–7 and were sliced into a section of 0.3 mm thickness (49) and separately cultured on a Millicell membrane (PICMORG50, Millipore) with serum-free minimum essential medium (MEM) containing 20 mM HEPES, 36 mM glucose, 100 units/ml penicillin, 100 μg/ml streptomycin with 1% ITS+ Premix (BD Biosciences), and 1 mM D-luciferin (Promega), in 35-mm dish. The bioluminescence from the cultured adrenal slice was measured with a highly sensitive cryogenic CCD camera (600 series, Spectral Instruments) coupled to a microscope (Axiovert 135TV, Carl Zeiss) at 35°C. Successive images were acquired every 20 min with an exposure time of 20 min. Background noise was removed from the images by applying a median filter (Metamorph Software, Molecular devices) to the whole stack of pictures, before measuring luminescence intensity from the area of interest. Luminescence intensity data were filtered by a low-pass filter with a cutoff threshold of 72 cycles/day to eliminate stochastic ultradian oscillations. Where indicated, Ang II (Peptide institute) was applied to culture medium at the final concentration of 100 nM. Phase shifts were calculated as the time-interval between the second and third peaks minus the time-interval between the first and second peaks.

The rat glomerulosa cells were obtained from adrenal glands of male rats weighing 200–250 g (7–8 weeks old) and isolated according to the method previously established by Payet et al. (50, 51). In brief, Per2-dluc rats were sacrificed at ZT5–7 and adrenal glands were immediately removed and dissected. The capsules were separated from fasciculata-reticularis by manual compression and incubated for 20 min at 37°C in oxygenized MEM (4 capsules/ml) containing 2 mg/ml collagenase and 125 units/ml deoxyribonuclease. After incubation, cells were disrupted mechanically, filtered through Cell Strainer (100 μm, BD Biosciences), and centrifuged for 10 min at 300g. The cell pellet was resuspended in OPTI-MEM medium supplemented with 2% fetal bovine serum, 100 units/ml penicillin, and 100 μg/ml streptomycin (52–54). The cells were then plated in 35-mm dishes (4 capsules/dish) and cultured with repeated medium changes at 24-h intervals for 3 days before experiments.

H295R cells (ATCC, CRL-2128) were cultured in DMEM/F12 supplemented with 2.5% Nu serum (BD Biosciences) and 1% ITS premix (BD Biosciences) as described previously (47). Cells were transfected with the following reporter plasmids using Lipofectamine LTX (Thermo Fisher Scientific): (i) pGL4.25 PER1 CRE×3-WT-Luc2CP, in which a tandem repeat of the sequence corresponding to the human PER1 CRE with flanking sequences (positions –1950 to –1921; 5′-TTC TTC CGC TTT GAC GTC ACT GCT GTC TCC-3′) was cloned into [luc2CP/minP] (Promega), (ii) pGL4.25 PRE1 CRE×3-Mut-Luc2CP, which is the same as (i) except that CRE sequences were mutated to 5′-TCACATAA-3′, (iii) pGL4.25 E4BP4 NBRE-like×3-WT-Luc2CP, in which a tandem repeat of the sequence corresponding to the human E4BP4 NBRE-like (5′-TGACCTTG-3′) with flanking sequences (positions +4452 to +4481) was cloned into pGL4.25, (iv) pGL4.25 E4BP4 NBRE×3-WT-Luc2CP, in which a tandem repeat of the human E4BP4 NBRE (5′-TGACCTTT-3′) with flanking sequences (positions +4845 to +4874) was cloned into pGL4.25, and (v) pGL4.25 E4BP4 NBRE×3-Mut-Luc2CP, which is the same as (iv) except that the human E4BP4 NBRE sequences were mutated to 5′-TGA ATTCT-3′. Bioluminescence recording was performed a day after transfection.

A luciferase reporter (Luc2P) driven by a mouse or human Per2 promoter sequence (positions –1670 to +53 for mouse; –1840 to +108 for human) was inserted between the inverted terminal repeat (ITR) sequences in pAAV-MCS2 plasmid (Addgene, Plasmid #46954) to obtain pAAV-mPer2-Luc2P or pAAV-hPER2-Luc2P. Recombinant AAV-DJ vectors were produced with a triple-transfection helper-free method as described (55). The day after AAV transduction, cells were split and equally plated into 35-mm dishes containing medium supplemented with 1 mM D-luciferin.

Bioluminescence recording was performed using a dish-type luminometer AB-2550 Kronos Dio (ATTO) under 5% CO₂ atmosphere at 37°C. Luminescence was measured at 20-min intervals, with an exception of 30-min intervals in pharmacological experiments (due to drug application during intervals between measurements). To evaluate dose dependency, CV11974 was administrated at a concentration of 0.001, 0.01, 0.1, and 1 μM. Where indicated, cells were treated with Ang II (100 nM), CV11974 (1 μM) or PD123319 (1 μM, Abcam). For enhancer activity reporter assay, values were normalized to the luminescence intensity that was recorded 30 min before Ang II stimulation. For single cell level recording, successive images were acquired with a highly sensitive cryogenic CCD camera (600 series, Spectral Instruments) with a 20 min exposure time for each picture (3 pictures/h). A median filter was applied to all images to eliminate cosmic-ray-induced background noise. Cell viability test and morphological inspection verify that the cells treated with CV or PD (each at 1 μM) for 24 h remained viable and morphologically unaffected ( Supplementary Figure S1 ). Trypan blue extraction assay was used for viability test (56). The half-life of CV and PD in H295R culture is not known (57).

Cells were lysed in Laemmli buffer and immunoblotting was performed according to our standard method (58) with antibodies against NGFIB (Santa Cruz, sc-5569, 1:500 dilution), E4BP4 (Santa Cruz, sc-9549, 1:1000), and β-actin (Sigma Aldrich, A5441, 1:1000). The data were normalized to the expression levels of β-actin. For qRT-PCR, cells were harvested in TRIzol reagent (Thermo Fisher). Total cell RNA was extracted using RNeasy micro kit (Qiagen) and converted to complementary DNA with SuperScript VILO cDNA synthesis kit (Thermo Fisher). Quantitative real-time PCR was achieved using the Platinum SYBR Green qPCR SuperMix-UDG with ROX (Thermo Fisher) with StepOnePlus system (Applied Biosystems). The data were normalized to a non-oscillatory housekeeping gene, RPLP0 mRNA levels. The primer sets used in this study included: PER1, Fw: 5′-GCA TCT CAG CGG AGC TCA CA-3′, Rv: 5′-GAG GCT GTA GGC AAT GGA ACT G-3′, PER2, Fw: 5′-GTG CAG CTC CAC CCT AGT GA-3′, Rv: 5′-GAT TTT CCT GCT CCA TGG GTT GAT G-3′, CRY1, Fw: 5′-AGC AAA CTC ACC TGT TGA AGC AAG G-3′, Rv: 5′-GCT GCA ACA GTA TTC CTC CTG AAT G-3′, BMAL1, Fw: 5′-AGT CTG TCT TCA AGA TCC TCA ACT AC-3′, Rv: 5′-CTG GAA GTC CAG TTT TTG CAT CTA TG-3′, DBP, Fw: 5′-GAA CCC GAC CCA GCT GAT CT-3′, Rv: 5′-CTT GGC TGC CTC GTT GTT CTT GT-3′, E4BP4, Fw: 5′-CCC GAG AGC AGG AAC ACG AT-3′, Rv: 5′-ACC CTA TCT ATG TGT GTA GGA GAA C-3′, and RPLP0, Fw: 5′-ATG CAG CAG ATC CGC ATG T-3′, Rv: 5′-TTG CGC ATC ATG GTG TTC TT-3′. For temperature entrainment, cells were subjected to 24-h temperature cycles with a 12-h warm phase (37°C) and a 12-h cold phase (33°C) for three days, and then released into constant 37°C conditions.

Isolated rat glomerulosa cells were seeded on poly-D-lysine-coated coverslips and cultured for 12 h. Cells were fixed with acetone for 10 min at –20°C, immediately rehydrated in 0.1 M phosphate buffered saline (PBS) for 5 min, and incubated with mouse anti-CYP11B2 antibody (Millipore, MAB6021, 1:100) in PBS containing 0.3% Triton X-100 (PBX) for 5 days at 4°C. After washing with PBX, the samples were incubated with Alexa594-conjugated donkey anti-mouse IgG (Thermo Fisher, 1:5000) for 12 h at 4°C. Nuclei were visualized by staining with 4′,6′-diamino-2-phenylindole (DAPI).

Radioisotopic in situ hybridization was performed with following gene-specific probes: for Agtr1a, nucleotides 1387–2138 (GenBank, NM_177322); for Agtr1b, nucleotides 1298–2015 (NM_175086); for Cyp11b2, nucleotides 1744−2456 (S85260). The corresponding cDNA fragment was cloned and used as a template for the generation of riboprobes. The riboprobes were radiolabeled with [³³P]UTP (PerkinElmer), using a standard protocol for the cRNA synthesis. In situ hybridization was performed as described (37). Briefly, paraformaldehyde fixed tissues were frozen and sectioned at a thickness of 30 μm. Then, the free-floating tissue sections were transferred through 4 × SSC, proteinase K (1 μg/ml, 0.1 M Tris-HCl [pH 8.0]; 50 mM EDTA) for 15 min at 37°C, 0.25% acetic anhydride in 0.1 M triethanolamine for 10 min, and 4 × SSC for 10 min. The sections were then incubated in the Denhardt’s hybridization buffer containing 55% formamide, 10% dextran sulfate, 10 mM Tris-HCl (pH 8.0), 1 mM EDTA, 0.6 M NaCl, 0.2% N-laurylsarcosine, 500 μg/ml tRNA, 0.25% SDS, 10 mM dithiothreitol (DTT) and radiolabeled riboprobes for 16 h at 60°C. Following a high-stringency posthybridization wash, the sections were treated with RNase A. Air-dried sections were exposed to X-ray films (Kodak Biomax).

Cosinor analysis and Rayleigh’s uniformity test were performed using Python 3.8. Western blot band intensities were quantified using ImageJ software. All statistical analyses were performed using GraphPad Prism 8, using the statistical tests for each figure as indicated in the legend.

To test expression of the core clock gene in mouse adrenal, we used isolated adrenal explants from Per2^Luc knock-in mice, which facilitate the analysis of circadian expression of Per2 in tissues and cells (44). Bioluminescence imaging revealed abundant PER2::LUC luminescence located mostly in the outer layer of the adrenal cortex, where ZG cells are located ( Figure 1A ); These observations are similar to those reviewed by other researchers (59). We found that ZG cells in Per2^Luc explants exhibit persistent luminescence oscillations for more than a week in culture, with a period length of approximately 24 hours (23.78 ± 0.27 h, mean ± SEM, n = 5 biologically independent slices) ( Figure 1B ). The PER2::LUC luminescence rhythms were abolished in adrenals from Per2^Luc ; Clock^Δ19/Δ19 mice ( Figure 1A ), indicating that these rhythms are generated by the endogenous circadian clock mechanism.

Next, we tested dissociated cell culture of ZG cells ( Figure 2A ). For this particular purpose, we used Per2-dluc rat (43), since isolation of primary ZG cells has been established in rats (50, 51). Immunocytochemistry confirmed that approximately 90 percent of the cells isolated were immunolabelled for the ZG cell-marker Cyp11b2, an enzyme responsible for aldosterone production, verifying enrichment of ZG cells as reported (54) ( Figure 2A ). The freshly isolated rat ZG cells were maintained with repeated medium changes at 24-h intervals for 3 consecutive days (for establishment of primary cell culture) and subsequently maintained under constant culture conditions without medium change for an additional 96 h for luminescence tracing. Under these conditions, sustained circadian bioluminescence rhythms were observed ( Figure 2A ), indicating that cultured ZG cells can act as a circadian oscillator.

As an extension of the above investigation, we also asked whether the human adrenocortical cell line H295R exhibits circadian oscillation ( Figure 2B ). H295R cells have been widely utilized as a model system for studying ZG (60), including aldosterone biosynthesis (61, 62). However, it is currently not known whether this human model cell line shares similar clock gene expression.

Temperature is known to entrain (or synchronize) cellular clocks of many tissues in vitro (63–65). We cultured H295R cells for 3 days with a temperature cycle (12 h at 37°C, 12 h at 33°C) and placed them in constant temperature conditions (37°C). Cells were harvested every 4 h over a 48-h period to study circadian variations. Quantitative RT-PCR analysis ( Figure 2B ) revealed that the expression levels of PER2 and BMAL1—representatives of two major feedback loops within the clockwork—showed high-amplitude circadian oscillations. PER2 exhibited its circadian peak expression 8 h earlier than BMAL1, and DBP mRNA expression fluctuated in the opposite phase of BMAL1—a phasic relationship consistent with the current molecular model of the circadian clock (7–9). Our data indicate that mouse and rat ZG cells and human H295R cells all retain a functional clockwork.

Because Ang II is a potent physiological regulator of ZG function (35), we hypothesized that Ang II may modulate circadian rhythms in ZG. In situ hybridization using radioisotope-labeled probes for the Ang II type 1 receptor (AT1R) subtypes Agtr1a and Agtr1b confirmed expression of AT1Rs in the mouse ZG ( Figure 3A ), further prompting us to test Ang II’s effect on ZG clock. When Ang II was administered to cultured adrenal slices at different phases of luminescence, it caused phase-dependent phase-shifts (delay versus advance) ( Figure 3B ): administration at the luminescence peak (12 h after the trough) significantly delayed the phase, whereas administration 6 h after the peak advanced the phase significantly (P = 0.001 for phase delay, P = 0.045 for phase advance, unpaired two-sided Student’s t test, Figure 3C ). These results demonstrate that mouse ZG contains an Ang II-regulatable circadian clock.

The human adrenocortical H295R cells were also treated with Ang II to test for the resetting effect of Ang II stimulation ( Figure 4 ). To do this, H295R cells were seeded onto 24-well plates and cultured for 2 days to confluency; then, after Ang II treatment, cells were harvested at 0, 2, and every 4 hour over a 68-h period (n = 3 replicates; total 57 wells, Figure 4A ). Quantitative RT-PCR analysis revealed rhythmical expression of PER1, PER2, BMAL1, CRY1, DBP, and E4BP4 that continued over 3 circadian cycles after Ang II treatment (cosinor analysis, available in Supplementary Table S1 ). There was an acute and transient upregulation of PER1 mRNA expression immediately after Ang II stimulation. In addition, a similarly prominent (~ 6 fold) upregulation of mRNA expression was also identified for E4BP4, while no appreciable elevation was observed following vehicle administration (see Supplementary Figure S2 ). These observations, taken together, indicate that Ang II stimulation has the ability to reset or (re)initiate circadian clock gene expression in H295R cells.

In order to facilitate real-time tracing of circadian clock gene expression, we next employed a luciferase reporter system. We generated AAV encoding a luciferase under the PER2 promoter from either human or mouse origin and transduced it to H295R cells. Then, luminescence was traced both before and after Ang II stimulation, in culture ( Figure 4B ). Luminescence rhythmicity gradually dampened before Ang II treatment. Following Ang II administration, a significant and high-amplitude surge was observed, along with sustained luminescence oscillations in subsequent cycles (P < 0.0001, for the first surge and second circadian cycle, vs vehicle control, Figure 4B ). Single-cell-level analysis of circadian rhythms was further performed by CCD video recording of H295R cells before and after Ang II treatment ( Figure 4C ). We observed that upon Ang II treatment, bioluminescence rhythms of individual cells were immediately resynchronized. This was evident as the vertical alignment of peak bioluminescence across all traced cells in the heatmap of cellular rhythms after Ang II treatment. These data illustrate the phase-resetting of individual cells [see also ref. (66)].

CV11974 (CV) is a receptor antagonist specific to AT1R. We found that CV treatment dose-dependently suppressed the rhythm resetting of H295R cells ( Figure 5A ). In this experiment, a Per2-Luc reporter was also virally introduced into H295R cells. As the CV dose increased (0, 1, 10, 100, and 1000 nM), the Per2-Luc expression induced by Ang II decreased accordingly. This resulted in a similar dose-dependent attenuation of the Per2-Luc rhythm amplitude in the following circadian cycles ( Figure 5A ). Moreover, CV shortened the duration of the initial surge of the Per2-Luc expression, which in turn influenced the phase of the Per2-Luc rhythms in the subsequent circadian cycles ( Figure 5B ; compare phases across different CV doses). We observed that the Ang II type 2 receptor inhibitor PD did not affect either the amplitude or phase of the Per2-Luc rhythm in the same H295R cells ( Supplementary Figure S3 ). Without Ang II, CV treatment had no effect on the Per2-Luc expression ( Supplementary Figure S4 ), consistent with a conjecture that AT1R is not basally activated in the culture conditions that we used for H295R cells (67–69).

Furthermore, we observed that the magnitude of the initial induction of endogenous PER1 and E4BP4 mRNA expression in Ang II-treated H295R cells was also decreased by CV treatment in a dose-dependent manner ( Figures 5C, D ). Considering that CV is one of the most commonly used clinical drugs for hypertension treatment, its dose-dependent effects on clock gene expression imply a previously unstudied potential action of the drug [for known actions of CV, see (70)].

It has been reported that PER1 is acutely responsive to various stimuli through the cAMP-response element (CRE) located in its promoter and that Ang II causes upregulation of CRE- mediated gene transcription in H295R cells (71–76). Consistently, the isolated human PER1 CRE sequence (5′-TGACGTCA-3′) significantly increased reporter transcription upon Ang II stimulation, and CV treatment dose-dependently reduced this activation (see Figures 5E−G for PER1). However, CRE has not previously been implicated in the regulation of E4BP4 (77).

Previous studies investigating Ang II-downstream signaling in H295R cells (61, 78–80) revealed two independent, DNA-binding transcription factor families — ATF/CREB family and NGFI-B family — that mediate transient transcription of a different set of genes upon Ang II stimulation. Because CRE sequence (ATF/CREB-binding motif) is not present in the human E4BP4 locus, we rather propose that the NGFI-B-responsive element (NBRE, 5′-TGA CCTTT-3′) may play a role in the Ang II/CV-dependent mRNA induction of E4BP4 ( Figures 5H−K ). E4BP4 consists of two exons. Sequence conservation analysis revealed that the intron 1 contains a highly conserved segment (+4283 to +5181, relative to the transcription start site) harboring two putative NBRE sequences at +4459 to +4466 (NBRE-like) and +4852 to +4859 (NBRE-consensus), but no CRE sites (see Figure 5H ). Western blots ( Figure 5I ) confirmed acute accumulation of NGFI-B, noticeable as early as 2 h after Ang II treatment, consistent with previous reports (61, 68). E4BP4 protein levels peaked around 8 h after Ang II treatment ( Figure 5I ). Subsequently, we assessed the potential cis-element function using reporter vectors containing tandem repeats of the putative cis-element with flanking regions (NBRE-consensus×3-luc and NBRE-like×3-luc) ( Figure 5J ). In H295R cells, Ang II significantly increased the reporter activity of NBRE-consensus, and CV effectively blocked this effect. Contrastingly, neither Ang II nor CV had any effect on the NBRE-like sequence, indicating the sequence specificity of Ang II-induced transactivation in H295R cells. We confirmed that mutation of the NBRE-consensus sequence abolished its Ang II-dependent enhancer activity ( Figure 5J , Mut). Finally, we applied CV at the same various doses as used in Figure 5A . We found that the upregulation of NBRE-consensus reporter activity was dose-dependently reduced by CV, in a manner similar to the changes observed in endogenous E4BP4 mRNA expression in H295R cells ( Figures 5D, K ). We thus propose that the NBRE sequence in intron 1 may be involved, at least partly, in the Ang II-regulated transcription of E4BP4 in H295R cells.