Circadian rhythm is essential for driving various physiological and metabolic functions in most living organisms, and coordination of their behavioral and biological processes such as sleep/wake cycles, body temperature, hormone secretion, blood pressure, and salivary secretion (Reppert and Weaver, 2002). Disturbance of the circadian rhythms can cause major systemic conditions such as cancers, central nervous system diseases, cardiovascular disorders, metabolic syndromes, and mental illness (Hastings et al., 2007). The rhythm is regulated by a master clock and other peripheral biological clocks. The master clock generates 24-h circadian rhythms in all mammals. It is located in the suprachiasmatic nucleus (SCN) of the hypothalamus and orchestrates the regulation of peripheral clocks. Several other peripheral clocks are regulated independently by endogenous self-oscillating transcription factors called the clock genes. Molecular regulation by the clock genes in the circadian oscillator is based on interconnected transcriptional/translational feedback loops (Reppert and Weaver, 2002; Suárez-Trujillo and Casey, 2016). Among them, aryl hydrocarbon receptor nuclear translocator-like protein 1 (Arntl/Bmal1), period (Per), circadian locomotor output cycles kaput (Clock), and cryptochrome circadian clock (Cry) are termed as the core clock genes. Retinoic acid receptor-related orphan receptor alpha (Rorα) and nuclear receptor subfamily 1 group D member 1 (Rev-erbα) regulate the transcription of Bmal1. The key transcription factors, CLOCK and BMAL1, form heterodimers that bind to the enhancer box (E-box) sequences and activate transcription of the Per and Cry genes. The PER and CRY proteins subsequently repress the transcription at their own promoters through a negative feedback loop, effected by acting on the CLOCK–BMAL1 complex (Shearman et al., 2000; Sato et al., 2014). This feedback loop controls the master and the peripheral clocks in most tissues.

Salivary flow rate or the secretion rate of salivary substances such as Na⁺, HCO3−, Cl⁻, K⁺, and α-amylase also follows a circadian rhythm (Dawes, 1972). Especially, it is known that unstimulated salivary flow rate is extremely low during sleep (Dawes, 1972). Recent studies have shown the circadian expression rhythm of clock gene (Per1, Per2, Bmal1, Cry1, deleted in esophageal cancer 1[Dec1], Dec2, D-site of albumin promoter binding protein [Dbp] and Rev-erbα) and amylase 1 in SGs (Furukawa et al., 2005). The localization of core clock proteins (Bmal1, Per2, and Clock) and Bmal1 and Per2 mRNAs in the mucous acini and striated ducts of salivary glands was also determined by in situ hybridization (Zheng et al., 2012). Moreover, light and food entrainment control the phase of submaxillary Per1 expression (Vujović et al., 2008). These studies suggest that not only acinar but also ductal cells play an important role in circadian oscillation of salivary secretion. Although the findings imply that clock genes influence the physiological functions in the salivary glands, detailed rhythmic and temporal expression patterns of the clock genes in salivary gland cells, that is, acinar and ductal cells, remain to be investigated. The major salivary glands, SGs as well as the parotid and sublingual glands normally contribute over 90% to the total volume of unstimulated saliva. Percentage contributions of salivary glands during unstimulated saliva are as follows: 65% from SGs, 20% from parotid, and <10% from sublingual and minor glands (Humphrey and Williamson, 2001). SGs are mainly composed of two epithelial cell types: the acinar cells, which secrete water, ions, and the salivary proteins; and the ductal cells, which modulate the ionic composition of the saliva (Humphrey and Williamson, 2001).

There are transcellular and paracellular transport pathways for the secretion of water and ions in the SGs, which are driven by changes in transmembrane osmosis and water channel gating (Turner and Sugiya, 2002). Recent studies have shown that Anoctamin 1 (ANO1) and Aquaporin 5 (AQP5) play an important role in water and ion transport in SGs (Ma et al., 1999; Yang et al., 2008). ANO1 and AQP5 are localized on the apical membrane of the SGs (Yang et al., 2008). ANO1, is a transmembrane protein that functions as a Ca²⁺-activated chloride channel (CaCC). CaCCs control the apical Cl⁻ efflux, which is essential for the vectorial transport of electrolytes and water in the bronchiolar epithelial cells, pancreatic acinar cells, proximal kidney tubule epithelium, retina, dorsal root ganglion sensory neurons, airways, and salivary glands (Caputo et al., 2008; Yang et al., 2008; Ferrera et al., 2010). AQPs are channel proteins that regulate the transmembrane movement of water in response to osmotic gradients for driving the salivary secretions. In SGs, AQP5 is one of the major aquaporins expressed on the apical membrane of the acinar and intercalated ductal cells (Delporte and Steinfeld, 2006; Matsuzaki et al., 2012). Although Ano1 and Aqp5 are key genes required for water and ion secretion in the SGs, their exact contribution to the regulation of circadian rhythm in salivary secretion remains to be validated. In the present study, we examined the interaction of temporal rhythmic expression patterns among the clock genes Ano1 and Aqp5, in rat SG acinar and ductal cells.

We proved that Bmal1, Per1, Per2, Clock, Cry1, Cry2, Rorα, and Rev-erbα show rhythmic circadian expressions in SGs (Figure 1). As previously shown, the phases and peaks of Bmal1, Per1, and Per2 expression exhibited opposite rhythms, which were shifted 12 h and repeated every 24 h. The Cry1 expression peaks were shifted compared to those of Per1, Per2, and Bmal1. Rorα and Rev-erbα showed rhythmic circadian expression similar to that of Per2. The phases and peaks in their expression profiles were similar to those in the SCN and other organs (Reppert and Weaver, 2002; Ueda et al., 2005). Our results were consistent with previous studies in SGs (Furukawa et al., 2005; Zheng et al., 2012). These results suggest that rat SGs have peripheral clocks and molecular clock mechanism with negative feedback loops. Although, Clock expression in whole SGs showed circadian oscillations, the peaks could not be distinguished clearly. The indistinct expression pattern may be caused by the phase differences in the Clock expression in each acinar and ductal cell individually (see below). It has been reported that temporal expression profile of Clock gene in the SCN did not show any circadian rhythm, however the peripheral clocks in organs such as liver and kidney, showed circadian rhythm in Clock expression (Ueda et al., 2005). This is in consensus with our results.

The expression of Ano1 and Aqp5 mRNA in whole SGs showed circadian rhythm, which was synchronous with Per2 expression (Figure 1). In recent studies, ANO1, CaCC, was found to express highly in the luminal membrane of the acinar cells in the SGs (Caputo et al., 2008; Yang et al., 2008). CaCC controls the apical efflux of Cl⁻ ions, which drives the transport of electrolytes and water secretion in the SGs. Transfection with mouse Ano1 siRNA significantly reduced the peak salivary flow rate induced by muscarinic-cholinergic stimulation (Yang et al., 2008). The Activation of CaCC caused an increase in intracellular Ca²⁺ concentration, which also showed circadian rhythm (Ikeda et al., 2003). In addition, Aqp5 knock-out mice showed decrease in expression of the tight junction proteins and water permeability, and a significant increase in the volume of the acinar cells compared to the wild type (Krane et al., 2001; Kawedia et al., 2007). In the Aqp5 knock-out mice, saliva production by muscarinic-cholinergic stimulation was reduced by more than 60% (Yang et al., 2008). In the present study, Aqp5 mRNA expression showed similar temporal patterns and peak times as that of Ano1. These results suggest, that the circadian rhythm in water secretion in SGs may be associated with water permeability, which is controlled by the circadian oscillation of Ano1 and Aqp5 expression. Rats displayed an increased intake of water and food during the nocturnal period (ZT12–ZT24). The expression profile of Ano1 and Aqp5 also correlated with their feeding and drinking behavior (Boulos and Terman, 1980; Mistlberger, 2009).

We showed that ANO1 protein expression displayed rhythmic circadian oscillations. There was no time-lag between the peak time of protein and mRNA expression (Figure 2). BMAL1 protein was transcribed from Bmal1 mRNA without any significant time-lag (Tamaru et al., 2000). PER2 protein expression rhythm showed 4 to 6-h phase delay compared to the mRNA expression rhythm (Field et al., 2000). The result indicating that Ano1 mRNA translated without any delay owing to factors such as phosphorylation-dependent proteolysis or shuttling by nuclear export signals.

In this study, we successfully isolated almost pure populations of acinar and ductal cells from SGs to analyze clock genes mRNA expression in each individual cell. Kallikrein, a serine protease, has been found to be localized in striated duct cells, but not in acinar cells in rat SGs (Simson et al., 1979). Immunoreactive Egf, Egfr, Hgf, Fgf2, Tgf-α, Tgf-β1, and Igf1 are localized in ductal cells (Simson et al., 1979; Motoko et al., 1992; Amano and Iseki, 1993, 2001). In contrast, Ano1 is present in acinar cells, but not in ductal cells in SGs (Yang et al., 2008). In the present study, the isolated duct-like cells showed higher kallikrein activity (Supplementary Figure 1) and Egf, Egfr, Hgf, Fgf2, Tgf-α Tgf-β1, and Igf1 expression compared with Ano1-positive acinar-like cells (Supplementary Figure 2), indicating that we could isolate almost pure populations of acinar and ductal cells. In addition, we could not find any significant differences in Aqp5 expression between acinar and ductal-like cells (data not shown). These results were consistent with the finding that Aqp5 was expressed in both the apical plasma membrane in acinar cells and cells in the striated duct (Matsuzaki et al., 2012).

Acinar cells showed rhythmic expression of Bmal1, Per1, Per2, Cry1, Cry2, Rorα, and Rev-erbα, but not Clock (Figure 3). In ductal cells, mRNA expression of Bmal1, Per1, Per2, Clock, Cry1, Cry2, and Rev-erbα displayed a circadian rhythm, whereas Rorα did not (Figure 4). The temporal expression pattern of Bmal1, Per1, Cry1, Cry2, and Rev-erbα in both the acinar and ductal cells, isolated from the SGs, had very similar circadian rhythms. The rhythmic expression pattern of Clock was indistinct in whole SGs and acinar cells. However, ductal cells showed rhythmic Clock mRNA expression, similar in pattern to Bmal1. The obscurity in Clock rhythmic expression in whole SGs might be due to overlapping cycles of mRNA expression in individual acinar and ductal cells. Phase difference in mRNA expression for Per2 and Rorα was also observed between acinar and ductal cells. The peak mRNA expression time for Per2 in whole SGs, corresponded with that in the isolated ductal cells. The peak time of expression for Rorα in whole SGs did not coincide with that in acinar and ductal cells. Variable oscillatory expressions of Per1, a paralog of Per2, has been observed in individual cell in the SCN (Yamaguchi, 2003). These results suggest that temporal expression patterns of clock gene mRNAs in whole SGs are established by the combination of their expression profiles in both the acinar and ductal cells. Per1, Per2, Clock, Cry1, Cry2, and Rorα showed high expression in ductal cells (Figure 5). The mRNA expression levels of Bmal1 and Rev-erbα were not significantly different. These results are also consistent with previous studies demonstrating the localization of Per2 mRNA expression in mice SGs by in situ hybridization (Zheng et al., 2012). Our results also confirmed the localization of mRNAs encoding clock genes in acinar and ductal cells. In addition, we revealed that there were broad differences in clock gene expression levels between acinar and duct cells.

To confirm the relationship between clock genes and Ano1 and Aqp5 expression, we studied their DNA sequences. The promoters of rat Ano1 and Aqp5 have an enhancer box (E-box) binding sequence. The E-box is a DNA response element that acts as a protein-binding site and is necessary for the binding of BMAL1-CLOCK heterodimer to the promoter region. E-box plays a major role in regulating the transcription of Per2 and Cry1 (Bozek et al., 2009). A recent study has identified the E-box protein-binding sequence on the Aqp5 promoter in humans and mice (Zheng et al., 2012). In rats, the E-box element has been identified from −160 to +69 on the Aqp5 promoter (Nomura et al., 2007). The existence of E-box in the promoter region is a characteristic of clock-controlled genes (CCGs), which show rhythmic expression and contain clock gene binding motifs, in the promoter (Bozek et al., 2009). Thus, Ano1 and Aqp5 may be putative CCGs and key targets of the BMAL1-CLOCK heterodimer in SGs.

In conclusion, we showed circadian rhythmic expression of Bmal1, Per1, Per2, Clock, Cry1, Cry2, Rorα, and Rev-erbα mRNAs in not only whole SGs, but also isolated acinar and ductal cells. From our personal communications, Bmal1 mRNA also showed rhythmic expression in a salivary ductal cell line (He et al., 1998). Although we could not exclude the possibility of contributions of a master clock in the suprachiasmatic nucleus to circadian rhythmic salivation, cells in the salivary gland may have independent peripheral clocks. Further studies are needed to clarify the involvement of the suprachiasmatic nucleus in the salivation rhythm by using in vivo experiments, such as monitoring of salivary secretion following section of nervous input to the SGs. Clock genes may also regulate the circadian rhythmic oscillatory expression of Ano1, and Aqp5 mRNA and ANO1 protein, and may play an important role in the physiological nocturnal salivary secretion including water and electrolyte secretion. In addition, predominant rhythmic expression of clock genes in ductal cells may play an important role in modifying the ionic composition of the saliva, by reabsorbing and/or secreting ions during salivation.

RS, YS, and NS conceived and designed the experiments. RS performed all the experiments. RS and YS conceived the ideas, and wrote the paper. RS, YS, MS, MK, MT, YI, and NS contributed to the analysis of results and preparation of the figures and tables. All authors read and approved the results and the final manuscript.