Circadian clocks of the kidney: function, mechanism, and regulation

Abstract

An intrinsic cellular circadian clock is located in nearly every cell of the body. The peripheral circadian clocks within the cells of the kidney contribute to the regulation of a variety of renal processes. In this review, we summarize what is currently known regarding the function, mechanism, and regulation of kidney clocks. Additionally, the effect of extrarenal physiological processes, such as endocrine and neuronal signals, on kidney function is also reviewed. Circadian rhythms in renal function are an integral part of kidney physiology, underscoring the importance of considering time of day as a key biological variable. The field of circadian renal physiology is of tremendous relevance, but with limited physiological and mechanistic information on the kidney clocks this is an area in need of extensive investigation.

CLINICAL HIGHLIGHTS

• Rhythms in renal function are endogenous rather than a consequence of behavior, food intake, or posture.

• A nondipping phenotype in blood pressure (BP) (<10% dip in pressure during the night) is associated with activation of the renin-angiotensin-aldosterone system, increased risk of chronic kidney disease (CKD), and adverse cardiovascular events.

• Twenty-four-hour ambulatory BP monitoring may be better at predicting target organ damage in CKD patients compared with clinic BP measures.

• Rhythms in urinary potassium excretion are lost in patients with renal insufficiencies and CKD.

• Circadian disruption of urine acidification is associated with uric acid stone formation.

1. INTRODUCTION

1.1. Mechanism of the Circadian Clock

Circadian rhythms are internal variations in behavior and physiological function coordinated by an intrinsic molecular clock. The organization of the circadian system is still referred to in hierarchical terms with the suprachiasmatic nucleus (SCN) at the top (1). The central clock resides in the SCN in the hypothalamus and is entrained by light cues that are converted to chemical signals via the retino-hypothalamic tract. This central clock in the SCN was long referred to as the “master clock,” but the prevailing understanding is that the SCN clock functions as the conductor of a circadian orchestra in which the molecular clocks in various peripheral tissues, located in nearly every cell type in the body, comprise the instruments (2) (FIGURE 1). The mechanism of the molecular clock is a transcription-translation feedback loop (see TABLE 1 for definitions of circadian terminology). In the positive arm of the molecular clock, which activates transcription, the proteins Brain and muscle ARNT-like protein-1 (BMAL1) and Circadian locomotor output cycles kaput (CLOCK) heterodimerize and bind E-box response elements in the regulatory regions of target genes. Period (Per1/2/3) and Cryptochrome (Cry1/2) genes are induced by BMAL1/CLOCK, and once translated in the cytoplasm PER and CRY proteins translocate back into the nucleus to feed back on BMAL1/CLOCK, thereby decreasing their own transcription in the negative arm. Another well-known circadian clock target gene is Dbp, which encodes the transcription factor D-site albumin-binding protein DBP. Activated by BMAL1/CLOCK, DBP acts in a secondary loop with the protein Nuclear factor, interleukin 3 regulated (NFIL3) to activate the expression of RAR-related orphan receptor alpha (Rora). BMAL1/CLOCK and DBP/NFIL3 work together through E-box and D-box response elements, respectively, to activate the expression of Nr1d1/2, which encode the nuclear receptors REV-ERBα/β. In a third regulatory loop of the clock, REV-ERBα and RORα mediate opposing actions to either inhibit or activate BMAL1 transcription via interaction with a ROR response element (RRE). These three sequence-specific regulatory elements, E-boxes, D-boxes, and RREs, are bound by circadian clock transcription factors to mediate the rhythmic expression of nearly 50% of all expressed genes in a tissue-specific manner (extensively reviewed in Refs. 3–5). It is important to note that most of the studies on peripheral clock mechanisms have been focused on the liver and other extrarenal tissues. These mechanisms remain to be validated in the kidney clocks, but what is known is discussed in this review.

FIGURE 1.

Circadian physiology and molecular mechanism. The circadian system is organized as a hierarchical multioscillator network, with the central clock in the suprachiasmatic nucleus (SCN) of the hypothalamus. The SCN directly receives photic cues for entrainment to the light-dark cycle and synchronizing peripheral clocks. Although the SCN is responsive to photic cues, the peripheral oscillators are more responsive to nonphotic cues such as feeding. The mechanism of the molecular clock is a transcriptional/translational feedback loop. The core loop consists of a positive (BMAL1, CLOCK) and a negative (PER, CRY) arm, which act on the E-box response element. This clock mechanism regulates tissue-specific target genes to regulate rhythmic outputs in physiology and behavior, with these rhythmic outputs influencing the circadian system. Diagram created with BioRender.com, with permission.

Table 1.

Circadian terminology

Term	Definition
Active/inactive period—rodents/humans	Nocturnal animals, such as rodents, experience their active period during the night and their inactive period during the day, whereas diurnal animals, such as humans, experience their active period during the day and their inactive period during the night.
BMAL1/CLOCK	Protein heterodimer that makes up the positive arm of the molecular circadian machinery. Binds E-box response elements upstream of clock target genes to upregulate their transcription.
Central clock	Light-entrained “master clock” in the suprachiasmatic nucleus (SCN) of the hypothalamus that synchronizes peripheral clocks in other organ systems.
Circadian	Latin translation to “about a day.” Used to describe the intrinsic 24-h cycle that most organisms follow behaviorally and physiologically.
Circadian disruption	Loss of 24-h synchronicity due to disturbance of biological timing because of changes in molecular machinery, behavioral cycles, or environment.
Desynchrony	Loss of synchrony between rhythm and its zeitgeber (external stimulus, see below).
Misalignment	Inappropriately timed.
Molecular clock	Molecular basis of controlling the circadian nature of an organism; a transcription/translation feedback loop comprised of proteins BMAL1 and CLOCK, or the positive arm, and PERIOD and CRYPTOCHROME, the negative arm.
Pacemaker	Generation of endogenous rhythmicity and imposing this rhythmicity on one or more other entities.
PER/CRY	Protein heterodimer that makes up the negative arm of the molecular circadian machinery. Inhibits action of BMAL1/CLOCK heterodimer to downregulate transcription of circadian clock genes.
Period	Time taken for a full circadian cycle, from one peak of the oscillation to the next, typically, 24 h.
Peripheral clocks	Molecular circadian machinery found in tissues outside of the brain and synchronized by the central clock.
Phase	Time period in which peak and trough of a circadian measurement occurs; typically mentioned in reference to a “phase shift” where the timing of the beginning and end of the phase changes.
Zeitgeber—external time signals	External stimulus that acts as a cue to regulate the body’s circadian rhythms. This includes but is not limited to food intake, activity, and light signals.

Many renal physiological functions, including renal blood flow, glomerular filtration rate (GFR), renal cortico-medullary osmotic gradient, and tubular water and electrolyte transport, display circadian rhythms and have been hypothesized to be driven, at least in part, by the intrinsic kidney clocks (6). The kidney is made up of many different types of cells that have unique characteristics and collectively contribute to these renal functions. The functional unit of the kidney is the nephron, which includes different sections of cell types. This includes the glomerulus, a cluster of capillaries enclosed in the Bowman’s capsule, which acts as a filtration barrier to filter large volumes of blood. This filtration barrier is composed of fenestrated endothelium, the glomerular basement membrane, and podocytes, providing support and size and charge selectivity to filtered particles. This filtered blood (ultrafiltrate) enters the tubular system, comprised of the proximal tubule (PT), loop of Henle, distal convoluted tubule (DCT), and finally the collecting duct (CD), where solutes are reabsorbed back into the bloodstream (peritubular capillaries) and waste products excreted (reviewed in Ref. 7). The tissue clocks are often referred to in the singular, such as “the” kidney clock. However, given the heterogeneity of the kidney, it is reasonable to speculate that in fact there are as many distinct “kidney clocks” as there are cell types in the kidney. In this review, we discuss recent studies that support this and highlight areas in need of investigation.

The importance of the circadian clock to overall human health is underscored by decades of epidemiological evidence that disruption of circadian rhythms leads to adverse health outcomes (recently reviewed in Ref. 8). Circadian disruption is defined as a misalignment between behavioral rhythms and endogenous circadian rhythms, which may occur because of shift work, chronic jet lag, or abnormal light exposure, for example. Such misalignment leads to desynchrony among the central and peripheral clocks that is thought to underlie the disease states associated with circadian disruption. Human studies have shown a clear link between circadian disruption and sleep-wake disorders, neurological and psychiatric conditions, as well as cardiovascular disease, cancer, autoimmune disorders, and disruption of energy balance and adipose tissue metabolism (9–12). Increasing evidence demonstrates a causative role for circadian disruption in diseases of the kidney, including chronic kidney disease (CKD), hypertension, and lupus nephritis (13). In sect. 1.2, we discuss clinical evidence for the importance of circadian renal function.

1.2. Clinical Evidence for Kidney Clocks

1.2.1. Historical perspective.

The earliest reports in the literature describing rhythmic renal function date back more than a century (reviewed in Refs. 13, 14). It was not until Mills conducted controlled studies in human subjects in the 1950s and 1960s that it became clear that rhythms in renal function were endogenous rather than a consequence of behavior, food intake, or posture (15–18) (FIGURE 2). The association between CKD and loss of circadian rhythms in blood pressure (BP) and renal function has not been tested until more recently. Human studies in the 2000s demonstrated that elevated nocturnal BP and nocturia increase as renal function decreases and that the nondipping pattern may be associated with renal sodium (Na⁺) handling (19–21).

FIGURE 2.

Renal rhythms in electrolyte excretion persist in the absence of time cues. Cave studies involved human subjects underground with no knowledge of clock time. Rhythms in renal sodium, potassium, and chloride excretion, indicative of rhythms in renal function, remained intact after 8 wk underground in a male subject in complete isolation from any environmental time cues. By the end of the study (12 wk), rhythms had shifted but were still apparent. Reference data from Ref. 18. Diagram created with BioRender.com, with permission.

1.2.2. Pathophysiology.

A key function of the kidney is to balance fluid and electrolytes in the body to maintain BP. BP itself has a strong circadian rhythm that appears to be endogenous (22). This circadian rhythm in BP is under multifactorial control. The available evidence suggests a role for the molecular clock in several tissues, including the kidney, in the generation and regulation of the daily circadian variation in BP (reviewed in Ref. 23). Healthy individuals experience a 10–20% decrease in BP at night. Individuals who do not exhibit this “dip” of at least a 10% change in resting BP are termed “nondippers.” Nondipping is associated with activation of the renin-angiotensin-aldosterone system (RAAS) (24), increased risk of CKD (25, 26), and adverse cardiovascular events (27, 28) and can also be used to predict future cardiovascular events (29). Alterations in BP rhythms are associated with increased risk for worsening kidney disease as well as stroke and peripheral artery disease (30). Increased nocturnal BP is associated with microalbuminuria in type 1 diabetic patients as well (31). These studies highlight the importance of using 24-h ambulatory BP monitoring (ABPM) in at-risk patients. Indeed, ABPM measures were better predictors of target organ damage in CKD patients than clinic BP measures (32).

Maintaining acid/base balance is another critical function of the kidney. Urate excretion rhythms are very similar to urine flow rhythms in humans (33). Disruption of urine acidification rhythms is associated with uric acid stone formation (34). The circadian rhythm of urinary potassium (K⁺) excretion is altered in patients with chronic renal insufficiency, and this may be related to changes in acid/base handling within the kidney (35). Indeed, the night/day pattern of urinary K⁺ excretion is lost in patients with CKD. Daytime urinary K⁺ excretion is associated with nighttime BP and BP dipping, so loss of this night/day pattern in CKD patients may have implications for BP rhythm (reviewed in Ref. 36). Together these data demonstrate that renal function is rhythmic and that loss of this rhythmicity can be both the cause and the consequence of kidney disease. In sect. 2, we provide an in-depth discussion of preclinical studies using animal models that have increased our understanding of the role of the kidney clock in regulating renal function.

2. THE CIRCADIAN CLOCK AND RENAL FUNCTION

Circadian variation in renal function has been observed in humans as well as several different species, including nonhuman primates (37), sheep (38), dogs (39), rats (40, 41), and mice (42). In this section, we review studies done in squirrel monkeys and rodents that have increased our understanding of the connection between circadian biology and renal physiology.

2.1. Squirrel Monkey Studies

Moore-Ede and colleagues carried out extensive circadian studies using nonhuman primate squirrel monkeys (Saimiri sciureus). This animal model offers an advantage in that the squirrel monkey is diurnal, or active during the daylight, like most humans. Using male animals acclimatized to a specialized chair for collection of urine and monitoring of food and water intake, Moore-Ede and colleagues (37) were able to study physiological function in conscious animals under carefully controlled conditions. Consistent with previous studies in humans, urine flow, as well as urinary Na⁺ and K⁺ excretion, exhibited a clear circadian pattern, with peaks occurring toward the end of the active phase. The monkeys were entrained to a constant schedule or deprived of food or water intermittently to determine whether rhythms in renal function were endogenous or a result of behavioral patterns. The results clearly showed that rhythms in urinary electrolyte excretion and urine flow were independent of activity or food intake, providing strong evidence for the presence of an endogenous kidney clock.

2.2. Tau Mutant Hamster

Golden hamsters with a single point mutation in the gene encoding casein kinase 1 epsilon (CK1ɛ) are termed “tau mutants” because they have a shortened circadian period [20 h in homozygotes, 22 h in heterozygotes vs. 24 h in wild type (WT)] (43). The tau mutant hamster was one of the first mammalian species in which a single gene mutation was found to affect the circadian system. This model has provided a wealth of information about circadian physiology.

Interestingly, heterozygous tau mutants exhibit decreased longevity (44). In a landmark study, Martino et al. (45) tested the hypothesis that early death in these animals was due to cardiac and kidney disease. When tissues were collected at age 17 mo, tau mutant hamsters exhibited severe cardiac hypertrophy, myocardial fibrosis, collagen deposition, and hyperkalemia. Relative to the kidney, tau mutants exhibited proteinuria, widespread fibrosis and collagen deposition, and tubular dilatation. Importantly, this dramatic cardiorenal phenotype occurred in tau mutants that were raised in a normal light-dark (LD) cycle, which for golden hamsters is a 14:10-h LD cycle. Given that the endogenous period of these mutants is 22 h versus 24 h, Martino and colleagues tested the effect of altering the LD cycle to match the endogenous circadian rhythm of the mutants, thus synchronizing their external and internal environments. A separate cohort of tau mutants was raised in 12:10 LD. Remarkably, tau mutants raised in 12:10 LD and aged to 16 mo did not develop any signs of cardiac or renal damage. To test whether this effect could be due to desynchrony between the central and peripheral clocks, SCN ablation to remove the central clock was performed on tau mutants kept in 14:10 LD. These animals did not develop cardiac hypertrophy, providing further evidence that the cardiorenal phenotype of the tau mutant is due to the uncoupling of the central and peripheral clocks. Since the heterozygous tau mutants suffer from desynchrony and die young with cardiorenal disease but the homozygous mutants, who are arrhythmic, do not display this phenotype, it appears that having no rhythms may be more advantageous than having misaligned or desynchronized rhythms.

2.3. Rodent Studies

2.3.1. Insights into pathophysiology of the kidney clocks.

Several recent, exciting studies have investigated the effects of renal injury or renal disease models on the kidney clock. Shibata and colleagues (46) performed a compelling study using an adenine-induced model of tubulointerstitial nephropathy to show that the kidney clock is significantly disrupted in CKD. The PER2::Luciferase (LUC) transgenic mouse model, in which the firefly Luc gene is under the control of the Per2 gene (a member of the PER family), has been a useful tool to study circadian rhythms in various tissues in mice. Using male PER2::LUC mice treated with adenine to model renal dysfunction, these investigators found that circadian rhythms of the clock protein PER2 were blunted in the kidney of CKD mice and a similar effect was observed in the liver. Strikingly, rhythms gradually became dampened in the SCN of CKD mice, suggesting that disruption of the central clock may play a role in CKD. Rhythmicity of expression of circadian-regulated genes in the kidney of CKD mice was also blunted, including Aqp2 (encoding aquaporin 2 transporter) and Avpr2 (encoding vasopressin receptor 2), which were reported to be associated with increased urination and water consumption during their rest period and likely important in the development of renal dysfunction (46).

Zha et al. treated male WT C57BL/6 mice with cisplatin, a nephrotoxin leading to renal tubular cell death and inflammation (47), at two time points, zeitgeber time (ZT)1 or ZT13, where ZT0 is the time of activity offset for rodents (lights on) and ZT12 is the time of activity onset (lights off) (48). Consistent with the known circadian toxicity of cisplatin, serum creatinine and renal injury markers KIM-1 and NGAL1 were higher in the ZT1 group of mice. BMAL1 expression was significantly increased in response to cisplatin, leading these investigators to test the hypothesis that BMAL1 is a mediator of cisplatin-induced kidney injury. Consistent with this hypothesis, short hairpin RNA (shRNA)-mediated knockdown of BMAL1 prevented the cisplatin-induced increase in blood urea nitrogen (BUN), serum creatinine, KIM-1, and NGAL1. Moreover, overexpression of BMAL1 with adenovirus worsened renal injury markers in cisplatin-treated mice. These interesting findings strongly suggest a role for BMAL1 in mediating cisplatin-induced renal damage.

The unilateral ureteral obstruction (UUO) model causes tubulointerstitial fibrosis as a result of obstructed urine flow and is believed to mimic chronic obstructive nephropathy in humans (49). Using the UUO model to study the role of BMAL1 in tubulointerstitial fibrosis, Zhang et al. (50) found that inducible, whole body knockout (KO) of BMAL1 was protective against the development of fibrosis. Consistent with previous work from the Fitzgerald lab (51), Zhang et al. also showed that global BMAL1 KO in mice resulted in decreased BP and attenuation of the night/day difference in BP. After induction of UUO, the inducible BMAL1 KO mice exhibited reduced kidney fibrosis compared with control mice. Furthermore, expression of profibrotic genes Fn1 encoding fibronectin, Acta2 encoding α-SMA, Col-I, Col-III, Mmp9, and Timp1 was reduced in the KO compared with control mice after UUO. These BMAL1-dependent effects in the UUO model appeared to be limited to fibrosis, as differences in plasma creatinine, BUN, KIM-1, and NGAL1 were not observed between KO and control mice.

Taken together, these recent studies demonstrate a deleterious role for BMAL1 in the setting of kidney injury and kidney disease, in that KO of BMAL1 in these disease models resulted in protective effects. A recent and notable exception to this concept is that inducible, whole tubule KO of BMAL1, driven by Pax8 Cre, exacerbated hyperglycemia in male C57BL/6 mice treated with streptozotocin to induce type 1 diabetes (52). This effect was associated with increased gluconeogenesis in the proximal tubule of KO mice compared with control animals. GFR remained in the normal range in the streptozotocin-treated mice and albuminuria was mild; thus it remains unknown what role BMAL1 might play in the setting of more severe diabetic nephropathy. Nevertheless, this study demonstrates a clear connection between diabetic pathology and the molecular clock. There is a lack of human studies, which should be a focus in future investigations, as these animal studies raise the intriguing concept that circadian-based interventions could be therapeutic in the treatment of kidney injury, CKD, and perhaps diabetic nephropathy.

3. THE CIRCADIAN CLOCK AND HORMONE SIGNALING

Hormones target the kidney for regulation of many physiological processes, including water and electrolyte transport and BP homeostasis (53). It is evident that the molecular clock is important for hormonal rhythms. In this section, we review the current data on the clock mechanisms behind regulation of hormonal homeostasis and its implications on renal function.

3.1. Aldosterone

The adrenal mineralocorticoid aldosterone plays an important role in tight regulation of electrolyte and water balance in the distal portion of the nephron, the functional unit of the kidney. Studies in the 1950s were the first to demonstrate a rhythm in urinary aldosterone excretion in humans (reviewed in Ref. 14). Subsequent studies from Bartter and Halberg and colleagues (54, 55) provided further evidence that urinary aldosterone levels exhibit a circadian pattern in humans, with higher levels during the day compared to the night and a clear peak in the morning. Clear night/day differences were observed in the plasma for 17-hydroxycorticosteroids (17-OHCS), product of the breakdown of cortisol, as well (54). Cugini et al. (56, 57) emphasized the importance of time of day in interpreting the results of plasma renin activity and aldosterone diagnostic tests in humans. These clinical observations were quite reproducible, but the molecular mechanisms were unknown. In a landmark study, Doi et al. (58) made the novel finding that aldosterone synthesis may be directly regulated by the molecular clock, using male KO mice lacking CRY1 and CRY2. In this compelling study, the authors found that Cry1/Cry2 double-KO mice exhibited salt-sensitive, nondipping hypertension along with dramatically higher levels of plasma aldosterone. The hypertensive phenotype of the double-KO mice was ameliorated after treatment with the mineralocorticoid receptor (MR) blocker eplerenone. Using an unbiased transcriptomic approach in adrenal glands, Doi et al. identified Hsd3b6 as an upregulated gene in the double-KO mice compared with control mice. Overexpression of HSD3B6 was limited to cells of the zona glomerulosa, where aldosterone is produced. Investigation into the molecular mechanism underlying this regulation performed with promoter luciferase assays in an immortalized human adrenal gland cell line, H295R, demonstrated that a D-box response element in the Hsd3b6 promoter contributed to the regulation of its expression. DBP mRNA levels were clamped at a high level in Cry1/Cry2 KO mice and its circadian rhythm of expression was lost, consistent with the proposed model of Hsd3b6 gene expression being subject to regulation by the circadian clock via DBP.

Work from our laboratory demonstrated a role for PER1 in the regulation of aldosterone levels. Using heterozygous Per1 mice on a 129/sv background, we found that decreased PER1 expression was associated with lower mRNA levels of Hsd3b6 in the adrenal gland, reduced plasma and urinary aldosterone, and increased urinary Na⁺ excretion in mice (59). Expression of CYP11B2, which encodes for aldosterone synthase, was not different between Per1 heterozygous mice and WT control mice. In contrast to these findings in 129/sv mice, our work in C57BL/6 mice demonstrated that global Per1 KO led to increased urinary aldosterone levels in male but not female mice on a normal diet (60). In response to a high-salt diet plus aldosterone analog (DOCP) treatment (HS/DOCP), urinary aldosterone excretion decreased in males and females in WT and Per1 KO, but male Per1 KO exhibited higher urinary aldosterone in the first 12 h after the administration of DOCP. Of note, we previously found that global male Per1 KO mice failed to downregulate renin in the kidney after HS/DOCP (61).

Using Ksp-cadherin Cre to drive KO of PER1 specifically in the distal nephron and collecting duct of the kidney, we recently found that this kidney-specific Per1 KO resulted in increased plasma aldosterone in male mice (62). We investigated gene expression in the adrenal glands of KS-Per1 KO and control mice on a normal diet or after 3 days of HS/DOCP. Hsd3b6 expression was not affected by either treatment or genotype. As expected, HS/DOCP treatment led to dramatic downregulation of Cyp11b2 in both genotypes; however, KS-Per1 KO mice exhibited higher levels of Cyp11b2 at the mRNA level under normal diet and after HS/DOCP treatment. Related to the work of Doi et al. (58), it is interesting to note that Cry1 (P = 0.06) and Cry2 (P < 0.01) expression were increased in the adrenal glands of KS-Per1 KO mice compared with control mice.

The aforementioned studies provide strong evidence linking the clock proteins PER1 and CRY1/CRY2 to the regulation of aldosterone levels. The fact that PER1 appears to mediate different effects on aldosterone that are mouse strain dependent is interesting and may be explained by differences in the RAAS that are known between C57BL/6 mice and 129/sv mice. For example, 129/sv mice have an extra copy of the renin gene and are relatively more salt sensitive than C57BL/6 mice (63, 64). It is interesting to note that aldosterone levels do not appear to differ between Clock KO mice and control mice, although Clock KO mice did exhibit an altered circadian rhythm in plasma aldosterone (42). Plasma aldosterone may be lower in global Bmal1 KO mice compared with control mice, but this was not directly tested (65). Studies in a triple-KO mouse model lacking three PAR bZIP transcription factors related to the clock mechanism, DBP, HLF, and TEF, showed a role for these proteins in regulating BP and plasma aldosterone (66). The Dbp/Hlf/Tef KO mice exhibited lower BP compared with control mice (but normal rhythms) as well as reduced plasma aldosterone. One caveat to this study is that the mice were on a mixed background, 129/sv and C57BL/6, and they possessed an extra copy of the renin gene, as noted by the authors.

Despite strain-specific differences in the aldosterone phenotypes related to the clock genes discussed above, these data clearly show a strong connection between the RAAS and the negative arms of the clock mechanism. With distal nephron PER1 having a role in regulation of aldosterone homeostasis, future studies need to focus on the role of the different kidney clocks on aldosterone. Although the outcomes may appear confusing, using diverse animal models in physiology studies is an important way to approximate the diversity of physiology and pathophysiology in the human population.

3.2. Cortisol

Glucocorticoids (cortisol in humans; corticosterone in rodents), produced by the adrenal gland cortex, exhibit a 24-h circadian rhythm, with peak concentrations at the start of the active period. Circadian control of glucocorticoid production and secretion involves the central clock and the adrenal clock (67). Glucocorticoids play important roles in various physiological functions, including immune function, electrolyte balance, metabolism, development, and cognition, as well as response to stressors (reviewed in Refs. 68, 69). Glucocorticoids play a role in integrated renal function by increasing renal blood flow and GFR, with variable effects on renal vascular resistance and water and electrolyte metabolism (reviewed in Refs. 70, 71).

Clock genes were found to be expressed in a circadian manner in the mouse adrenal gland (72), where glucocorticoids are synthesized, and these rhythms are abolished by lesioning of the SCN, demonstrating a role for the central clock in the regulation of this expression (73). A microarray demonstrated that >1,500 genes were rhythmically expressed in the adrenal gland of 4-mo-old male C57BL/6J mice. Biological processes shown to be enriched for rhythmically expressed genes included steroid biosynthesis (74). Global Per1 KO and Per2 KO mice on the 129/sv background had markedly elevated concentrations in plasma corticosterone, with rhythm only disrupted in Per1 KO mice (75). Rhythmicity of corticosterone was also lost in Per2/Cry1 double-KO mice. With glucocorticoid synthesis residing in the adrenal cortex, questions arose as to the role of the adrenal clock in regulating circadian rhythms of glucocorticoid synthesis. Oster et al. (74) showed that the adrenal clock contributes to the circadian rhythm of adrenal adrenocorticotropic hormone (ACTH) sensitivity, controlling production of glucocorticoids (discussed in more detail in sect. 5.2). The authors highlighted that Per2 and Cry1 genes together were critical for this rhythm. Furthermore, ACTH has been shown to reset the rhythm of the adrenal circadian clock in vitro (76). BMAL1 in the adrenal gland has been shown to be required for proper circadian rhythmicity of glucocorticoid production from the adrenal cortex of mice (77, 78), with the role of other adrenal clock components in glucocorticoid production still unknown.

The rhythmic secretion of glucocorticoids from the adrenal cortex is important for the synchronization of peripheral circadian clocks, particularly the renal clocks, although its exact role is not fully understood (79). The potential role of glucocorticoids in regulation of the kidney clocks is discussed in sect. 5.2.

3.3. Vasopressin

Vasopressin (AVP), also known as antidiuretic hormone (ADH), plays a role in osmotic balance, renal Na⁺ handling, renal function, and BP regulation. AVP is synthesized in magnocellular neurons located in the hypothalamus and released from the posterior pituitary into the circulation in response to hypovolemia or hypernatremia. AVP-positive neurons are also present in the SCN (∼20%, the second largest peptide population), but this does not appear to be a source of circulating AVP. For more in-depth review on the role of AVP in the SCN, see Refs. 80, 81. AVP exerts its effects on the distal nephron, where AVP receptors (V_1aR and V₂R) are expressed, stimulating insertion of AQP2 to the apical membrane to induce water conservation (82). Circadian rhythm of circulating AVP concentrations remains in debate, with some reporting rhythms peaking at late inactive phases in humans (83) and rats (84) but others reporting no rhythmicity (85, 86). AVP is unstable in blood, largely attached to platelets, and rapidly cleared (87), which could reflect inconsistencies. A study by Zuber et al. (88) illustrated the gene expression levels of AVP receptors (Avpr1a and Avpr2) every 4 h over a 24-h period in distal portions of the nephron, distal convoluted tubule (DCT)/connecting tubule (CNT), and cortical collecting duct (CCD), of male C57BL/6J mice. For Avpr1a, the peak of expression was at the transition from the active period into the inactive period in both segments of the nephron. The peak expression of Avpr2 is halfway through the active period in the CCD, and this was similar in the DCT/CNT but the amplitude of the curve is smaller. This expression pattern of both Avpr1a and Avpr2 was disrupted in Clock KO mice (background strain C57BL/6J), resulting in a significantly increased plasma osmolality during the active phase. This suggests that CLOCK plays a role in the rhythmicity of renal AVP receptors, but whether the renal clocks are specifically responsible is unknown. Altered plasma osmolality has been reported to control sympathetic activity and heart rate (HR) in both humans and mice (89). However, implications of altered plasma osmolality rhythm remain to be explored. Circadian gene expression of Avpr1a, Avpr2, and Aqp2 has been observed in the inner medulla in the PER2::LUC knockin mouse model (90). These rhythms were altered in Bmal1-deficient mice. Cortico-medullary osmotic pressure gradient rhythm peaked during the active phase, but osmotic pressure gradient was disrupted in Bmal1-deficient mice. Overall, these data suggest that BMAL1 plays a role in the cortico-medullary osmotic pressure gradient rhythm, which is mediated, at least in part, by V_1aR, V₂R, and AQP2.

AVP also exerts its actions on the vascular smooth muscle, causing V_1aR-mediated vasoconstriction and thus increasing BP (91). V_1aR is also present in the cortical and medullary vasculature of the kidney (92). A clear gap in our knowledge pertains to expression rhythms in the renal vasculature and how this could impact rhythms in renal function.

The clock has implications for AVP signaling, but AVP signaling has also been shown to impact the clock. Okamura and colleagues (93) conducted an exciting study showing that Avpr1a/Avpr1b double-KO mice are resistant to jet lag, as measured by 24-h locomotor activity, clock gene oscillations, and body temperature rhythms. These findings suggest that AVP signaling is involved in circadian rhythm alignment and a potential therapeutic target for circadian rhythm misalignment seen in chronic jet lag and shift work.

Together, these studies suggest that expression of AVP receptors in the kidney is rhythmic, with the AVP rhythm itself unclear. The renal clocks may play a role, but further studies are needed to explore this.

3.4. Endothelin-1

The endothelin system, first described by Yanagisawa et al. in 1988 (94), is a family of 21-amino acid peptides (ET-1, ET-2, and ET-3) with potent vasoconstrictor properties. Edn1 encodes a 212-amino acid peptide that is further cleaved into ET-1. ET-1 exerts its effect through the G protein-coupled receptors ET_A and ET_B, which are expressed in neurons, fibroblasts, endothelial cells, vascular smooth muscle cells, and different cell types within the kidney. ET-1 action depends on receptor activation as well as cell type-specific localization. For example, the ET_A receptor induces vasoconstriction and fibrosis in smooth muscle cells of the renal vasculature but inhibits Na⁺ reabsorption in the CD of the kidney via the ET_B receptor (reviewed in Refs. 95, 96). ET-1 has profibrotic effects in the PT as well (97, 98). The clinical use of ET_A, as opposed to dual ET_A/ET_B receptor antagonists, is effective for BP management in CKD patients as well as slowing down CKD progression. ET_A antagonist use is not yet widespread because of concern over fluid retention side effects (99). However, optimism remains for the use of endothelin receptor antagonists for treatment of diabetic kidney disease and CKD, given the promising results from the SONAR trial, which used an enrichment phase to identify patients with increased risk for fluid retention (100, 101). Indeed, a number of clinical trials testing ET_A antagonists in kidney disease, including glomerular disease as well as combinatorial treatment with SGLT2 inhibitors, are ongoing (102).

Circadian rhythms in Edn1 mRNA, ET-1 peptide levels, and ET receptor function in rodents and urinary ET-1 excretion in humans are well established (reviewed in Refs. 96, 103). Interestingly, the mRNA circadian rhythms for ET-1 and ET_A in the kidney of male C57BL/6J mice are quite similar and both are antiphase to ET_B (FIGURE 3). We became interested in ET-1 with our investigation into the short-term effects of aldosterone treatment in mouse IMCD-3 cells. We identified circadian clock gene Per1 and Edn1 as early aldosterone-regulated transcripts, with both transcripts increasing after 1 h of aldosterone treatment (105). Subsequent studies using chromatin immunoprecipitation (ChIP) and DNase I hypersensitivity assays found that in rat kidney tissue, isolated rat IMCD cells, and various murine kidney cell lines the regulation of Edn1 was at a transcriptional level via ligand-dependent transcription factors MR and glucocorticoid receptor (GR) (106, 107). When activated by aldosterone, other mineralocorticoids, or glucocorticoids, a conformational change in MR leads to translocation of the protein to the nucleus, where it can bind to hormone response elements (HREs) in target genes. Because of the significant homology between MR and GR, aldosterone release can also induce this process in GR. HREs were identified in the murine Edn1 gene, and both MR and GR were found to be bound to an HRE in Edn1 by ChIP and DNA affinity precipitation assay (DAPA). Additionally, this interaction was associated with dimethylated H3 lysine 4 residues and found to attract RNA polymerase II, indicative of a transcriptionally active promoter.

FIGURE 3.

Circadian expression of endothelin axis genes in the kidney. Relative circadian expression of endothelin-1 (ET-1) and ET-1 receptors ET_A (ETRA) and ET_B (ETRB); data adapted with permission from CircaDB (http://circadb.hogeneschlab.org/) (104) and fit to cosine curves. Gray shading indicates the active phase, and white background indicates the inactive phase.

Global Per1 KO 129/sv mice showed a decreased BP phenotype and significant elevation of ET-1 mRNA levels in the renal cortex compared with WT mice (108). Additionally, ET-1 mRNA showed inverted circadian oscillation to PER1 in WT mice, indicative of possible regulation of Edn1 by the PER1 protein. In mpkCCD_c14 cells, inhibition of PER1 nuclear localization by a casein kinase 1δ/ε inhibitor showed a fourfold increase of ET-1 mRNA levels. DNA affinity purification showed PER1 interacting near the aldosterone response elements on the Edn1 promoter; however, since the PER1 protein lacks a DNA binding domain, it most likely complexed with other proteins to bind the Edn1 promoter elements. Analysis of PER1 regulation of ET-1 expression, as well as its receptors ET_A and ET_B, was performed in the lung, heart, liver, renal inner medulla, and renal cortex of WT and Per1 heterozygous 129/sv mice (109). Real-time qPCR showed that the night/day patterns of expression for ET_A and ET_B were inverse to each other in the renal inner medulla and liver, but there was no apparent night/day difference in mRNA in the lung. Notably, ET-1 mRNA levels were significantly decreased in the lung in the Per1 heterozygous mice. Additionally, only ET-1 and ET_A were expressed with a diurnal pattern in the heart. These data indicate that PER1 does regulate ET-1 mRNA expression in a tissue-specific and time-dependent manner.

In contrast to the lower BP phenotype observed in global Per1 KO mice on a 129/sv background, global Per1 KO male mice on a C57BL/6J background exhibited an increased BP phenotype and nondipping hypertension when given HS/DOCP compared with WT mice (110). Aldosterone analog DOCP was used to increase renal Na⁺ reabsorption and model salt-sensitive BP changes that evade RAAS suppression. This nondipping hypertensive phenotype is associated with heightened levels of ET-1 peptide. Interestingly, this phenotype was only present in male mice, and no significant change was seen in female mice. After HS/DOCP treatment, ET-1 mRNA levels were significantly increased in the renal cortex of male Per1 KO mice compared with control conditions, whereas ET-1 mRNA levels did not change in the WT mice in response to HS/DOCP (60). ET_A was also upregulated in the WT male mice but not in the Per1 KO, whereas ET_B was upregulated in both. Female mice showed different gene expression, with WT and Per1 KO mice increasing renal cortex ET-1 expression after HS/DOCP treatment, but there were no significant changes in the expression of the receptors. The mechanism for these expression changes was investigated in mpkCCD_c14 cells transfected with a PER1-specific siRNA, which showed a 3.5-fold increase in ET-1 expression compared with cells treated with nontarget siRNA in the vehicle. Addition of aldosterone to the PER1-specific siRNA-treated cells increased the fold change to 4.5 compared with nontarget siRNA-vehicle-treated cells. ELISA of the apical and basolateral media of these cells also confirmed increased ET-1 peptide levels in aldosterone and PER1-specific siRNA-treated cells. Altered ET-1 peptide levels were mirrored in urine samples of male Per1 KO mice, with decreased night-to-day ratios of Na and ET-1 peptide excretion. These data suggest that in response to HS/DOCP treatment PER1 is a negative regulator of ET-1 expression, repressing aldosterone’s positive ET-1 regulation, but in a sex-dependent manner (60). Consistent with the findings in male global Per1 KO mice, we recently obtained similar results using a kidney-specific KO model (62). After HS/DOCP treatment, KS-Per1 KO mice exhibited higher levels of ET-1 in the urine and in the kidney, further supporting a key role for PER1 in repressing ET-1. Questions that remain are: does the increased ET-1 that results from Per1 KO act via ET_A, ET_B, or both receptors, and does the resulting receptor activation contribute to beneficial or detrimental ET-1 actions? To date, work from our laboratory has focused on the acute HS/DOCP model to study the PER1/ET-1 connection in the setting of salt-sensitive hypertension. More chronic studies employing a double Per1/Edn1 KO model are needed to define the PER1-ET-1 interaction, which may have implications for the choice of ET_A versus dual ET receptor antagonists in disease states.

Pollock and colleagues have conducted rigorous studies investigating ET-1 in the kidney with an emphasis on dietary salt as well as sex and time of day as key biological variables. Johnston et al. (111) carried out studies using male and female control and ET_B-deficient rats. The animals in this study were given an oral Na⁺ load at ZT0 (lights off) or ZT12 (lights on) to determine the role of time of day in the kidney’s ability to excrete Na⁺. ET_B-deficient rats of both sexes exhibited a delayed natriuretic response when the salt load was given at ZT0. Male ET_B-deficient rats had a delayed response to the ZT12 Na⁺ load compared with female ET_B-deficient rats. In general, the rats in this study were better able to excrete the Na⁺ load when it was given at the start of the active phase, demonstrating the importance of time of day in how the kidney and ET-1/ET_B help maintain Na⁺ homeostasis.

Speed et al. (112) investigated the effect of high dietary salt on the kidney clock in ET_B-deficient rats and transgenic control rats. This elegant study demonstrated circadian variation in plasma aldosterone, ET-1 levels, and mRNA expression of αENaC, BMAL1, CLOCK, CRY1, CRY2, PER1, and PER2 in the renal cortex and inner medulla. Under normal diet conditions, the rhythms in gene expression were synchronized between the cortex and inner medulla. Interestingly, the high-salt diet led to desynchrony between the cortex and the inner medulla, with BMAL1 expression undergoing a phase shift in the inner medulla on a high-salt diet. This effect was not observed in the ET_B-deficient rats, suggesting a role for this receptor in mediating synchrony within the kidney.

In a provocative study using stroke-prone spontaneously hypertensive (SHR) rats, Hill et al. (113) modeled shift work by subjecting the animals to chronic circadian disruption by advancing the light phase by 6 h every 7 days for 6 wk. Rhythms in urine output and ET-1 excretion were apparent in stroke-prone SHR rats under control LD conditions, but over the course of the 6-wk study these rhythms were suppressed in the environmental circadian disruption rats. This effect was associated with increased levels of renal injury markers Nephrin and KIM-1 in the urine. These results suggest that circadian disruption over just 6 wk can dampen circadian renal function and possibly damage the kidney.

Because of the significant physiological effects of ET-1, the regulation of the Edn1 gene by both aldosterone and PER1 in the kidney, and noted sex differences in ET-1 gene regulation, continued investigation into the circadian regulation of ET-1 and the impact on kidney function and BP has the potential to lead to the identification of novel therapeutic targets in the treatment of CKD and hypertension. Furthermore, understanding the interplay between the ET-1 system and the circadian clocks across tissues and cell types is likely to inform the use of ET_A or ET_A/ET_B receptor antagonists in various disease states.

4. MECHANISM OF THE KIDNEY CLOCKS

4.1. Development of the Kidney Clocks

In 1966, Mills noted in his landmark review on human circadian rhythms that urine flow rhythms are apparent in infants by 4 wk of age, hinting at a possible role for circadian rhythms in renal function in early life (14). Compelling data from both mice and rats show that the kidney clock begins keeping time during embryonic development (114–116). Mészáros et al. (114) used rats to investigate the developmental origin of the clock mechanism within the kidney. Fetal kidneys were collected every 4 h for a 24-h period at embryonic day (E)20 and at postnatal weeks 1, 4, and 12. Rev-erbα and Per2 gene expression exhibited significant rhythmicity at E20, and this effect persisted at the subsequent time points. In contrast, Bmal1 and Cry1 gene expression did not exhibit significant oscillation until postnatal week 1, and this effect persisted at weeks 4 and 12. Expression of several renal Na⁺ handling genes was also evaluated. Circadian rhythms in mRNA levels of epithelial sodium channel alpha subunit (αENaC; Scnn1a), Sgk, Na⁺/hydrogen (H⁺) exchanger 3 (NHE3; Slc9a3), and vasopressin receptor 2 (V₂R; Avpr2) were all apparent at E20 and at postnatal week 1. Interestingly, these four genes were expressed in phase with each other at E20, with peak expression occurring in the middle of the dark phase. By postnatal week 1, these genes were still in phase with each other, but the rhythm had inverted, reflecting nutritional cues from the nursing dam.

Previous studies in mice demonstrated that PER2::LUC rhythms were evident in the mouse kidney as early as E18 (115). This result was confirmed more recently in a detailed examination of rhythms in the developing mouse kidney by Sampogna and colleagues (116), who observed PER2::LUC rhythms at E17.5. These investigators performed circadian transcriptomic analysis in the fetal kidney by collecting tissues every 4 h for a 48-h period, spanning E18–E20. Strikingly, >4,000 oscillating transcripts were detected by RNA sequencing (RNA-seq), representing ∼18% of all expressed genes in the late embryonic kidney. Enrichment analysis demonstrated that the top pathways exhibiting circadian rhythmicity included cell cycle and cell division, DNA replication and repair, RNA splicing and mRNA processing, transmembrane transport, regulation of Na⁺ ion transport, and also several hits for circadian rhythms. Select genes identified as rhythmic transcripts were further investigated by ChIP to test whether CLOCK interacted with the promoters at two time points, circadian time (CT)6 and CT18. The CLOCK ChIP signal was higher at CT6 versus CT18 for the clock genes differentiated embryo chondrocyte 2 (Dec2), Rev-erba, and Per2. Remarkably, similar results were seen for the well-known kidney development genes Hoxb7 and Pax2. The HoxB7 Cre driver was used to generate ureteric bud-specific BMAL1 KO in the developing kidney. Careful examination of the developing kidney was done in the Hoxb7-Bmal1 KO, global Bmal1 KO, and control mice. These elegant experiments demonstrated that the embryonic kidney clock regulates branching rate and nephron number, clearly highlighting the important role of the clock in renal development.

4.2. Circadian Clock-Mediated Regulation of Cell Type-Specific Functions along the Nephron

The nephron comprises a glomerulus, acting as a filtration barrier, and the tubular system, where solutes are reabsorbed back into the bloodstream and waste products excreted (FIGURE 4) (reviewed in Ref. 7). Renal blood flow, GFR, renal cortico-medullary osmotic gradient, and tubular water and electrolyte transport all display circadian rhythms and have been hypothesized to be driven, at least in part, by the intrinsic kidney clocks (6, 117, 118). In this section, we discuss circadian clock-mediated regulation of renal functions along the nephron.

FIGURE 4.

Circadian clocks regulate gene expression in a cell-type specific manner in the kidney. Blood enters the nephron and is filtered at the glomerulus. A: podocytes cover the outside of the glomerular capillary to support the structure and function of the glomerulus. BMAL1-target genes, G protein-coupled receptor class C group 5 member A (Gprc5a), cathepsin L (Ctsl), transcription factor 21 (Tcf21), N-ethylmaleimide sensitive factor (Nsf), and G protein subunit alpha 12 (Gna12), are rhythmic (depicted with clocks) and involved in podocyte development. From the glomerulus, filtrate moves to the proximal tubule. B: proximal tubule cells express the sodium/hydrogen exchanger isoform 3 (NHE3), sodium-phosphate cotransporter 2 (NPT2A), phosphate transporter (PIT1/2), and the sodium-glucose linked transporter isoform 1 (SGLT1), all of which are located on the apical membrane (facing the filtrate) and are regulated by the circadian clock. Sodium reabsorbed from the filtrate is pumped back into the blood by the basolateral Na⁺-K⁺ pump, found on each of the cell types illustrated here. Filtrate then moves to the loop of Henle. C: cells in the thick ascending limb contain the sodium-potassium-chloride cotransporter isoform 2 (NKCC2), which is positively regulated by the nuclear estrogen-related receptor ERRβ and with-no-lysine 1 (WNK1)-oxidative stress responsive kinase 1 (OSR1)/Ste20-related proline-alanine rich kinase (SPAK) pathway. NKCC2 is negatively regulated by Alström syndrome 1 protein (ALMS1). Expressions of these regulators, as well as NKCC2, have a circadian rhythm (depicted by clocks). Again, these cells have a basolateral Na⁺-K⁺ pump and are regulated by the clock. D: the distal convoluted tubule cells contain apically located sodium-chloride cotransporters (NCC), which reabsorb Na⁺ and Cl⁻ from the filtrate. This cotransporter is regulated by the circadian clock and WNK1-OSR1/SPAK pathway. The final segment of the nephron that the filtrate enters is the collecting duct. E: in principal cells, the epithelial sodium channel (ENaC) reabsorbs Na⁺ from the filtrate and is regulated by the circadian clock. WNK1 positively regulates ENaC, with endothelin-1 (ET-1) and serum and glucocorticoid-inducible kinase 1 (SGK1), via neuronal precursor cell-expressed developmentally downregulated 4-2 (Nedd4-2), which negatively regulates ENaC. Expressions of all of these regulators in the kidney, as well as ENaC but not Nedd4-2, are regulated by the circadian clock. FXYD5 is also regulated by the circadian clock and in turn positively regulates the Na⁺-K⁺ pump. Expression of vasopressin receptors (V_1aR/V₂R) and aquaporin receptors (AQP2/4), involved in water reabsorption in the collecting duct, has been shown to have a circadian rhythm. Together, this highlights the extensiveness of the circadian clocks’ role in the regulation of genes throughout the cells of the kidney and its potential implications on renal function. Adapted from Ref. 6. Diagram created with BioRender.com, with permission.

4.2.1. Glomerulus.

Human studies have shown that GFR circadian rhythm is independent of circadian oscillations in BP, cardiac output, and sympathetic regulation (6, 117, 118), leading to the hypothesis that kidney clocks are involved. Firsov and colleagues (119) generated the first podocyte-specific KO by crossing Nephrin-Cre with floxed Bmal1 mice. Notably, GFR was measured in these mice every 4 h over a 24-h period. The results demonstrated the expected circadian rhythm in GFR for control mice, as has been shown for humans (reviewed in Ref. 120). In contrast, the podocyte-specific Bmal1 KO mice exhibited a 12-h rhythm in GFR, with two peaks occurring over the 24-h period, at ZT4 and ZT16. The authors also found that these peak GFR measurements negatively correlated with trough levels of plasma aldosterone, a marker of RAAS activity, suggesting a connection to tubuloglomerular feedback (TGF). However, to date, whether TGF has a circadian rhythm remains unknown. An unbiased transcriptomic approach revealed several BMAL1-specific target genes that are related to podocyte development (FIGURE 4A). The mechanism underlying the 12-h rhythm in GFR was unclear, yet these findings support a key role for BMAL1 in glomerular function. Understanding the kidney clock mechanisms underlying the circadian rhythm of GFR is an important area for investigation, as alterations in this rhythm are seen in patients with nephrotic syndrome (117) as well as in aging (121).

4.2.2. Proximal tubule.

The PT is responsible for the bulk of reabsorption of Na⁺, phosphate, and other solutes. The PT can accomplish this feat because of the high density of transporters on the highly invaginated apical surface of PT cells. Phosphate homeostasis is critical for overall health because of the functional role of phosphate in nucleic acid as well as cell structure, signaling pathways throughout the body, regulation of acid/base and calcium balance, and bone health (122). The PT is the final regulatory site for determining phosphate levels and is responsible for reabsorption of ∼75–85% of filtered phosphate depending on dietary intake as well as the body’s needs.

The circadian rhythm of serum and urinary phosphate levels is well known. Using the freely available circadian database CircaDB (104), we found that the genes encoding Na⁺-phosphate cotransporters PIT1 and PIT2 (Slc20a1 and Slc20a2, respectively) exhibit circadian rhythmicity in the mouse kidney (FIGURE 4B). Interestingly, these gene expression patterns are antiphase to each, such that the mRNA for Slc20a1 peaks early in the active phase while Slc20a2 is at trough levels at this same time. Slc34a1, which encodes the Na⁺-phosphate cotransporter NPT2a, also exhibits a significant circadian rhythm that is distinct from both Slc20a1 and Slc20a2, with a peak occurring at the very start of the active phase (FIGURE 5). It is tempting to speculate that these rhythmic patterns may allow for at least one of these Na⁺-phosphate transporters to be expressed at all times.

FIGURE 5.

Circadian expression of renal Na⁺-phosphate cotransporters. Relative circadian expression of Na⁺-phosphate cotransporters Pit1, Pit2, and NPT2A. Data adapted with permission from CircaDB (http://circadb.hogeneschlab.org/) (104) and fit to cosine curves. Gray shading indicates the active phase, and the white background represents the inactive phase for mice.

In an elegant study, Miyagawa et al. (123) found that protein levels of NPT2a and NPT2c exhibit circadian variations in the kidney. Importantly, Western blots were performed on isolated apical brush border membrane vesicles, so that the reported results represent apically localized and presumably active transporter proteins. Peak expression of NPT2a occurred at the transition of the active to rest phase, suggesting that there is a lag of ∼12 h between the peak in mRNA expression, as seen in FIGURE 5, and the peak in protein levels. In contrast, NPT2c expression peaked just before the transition from rest to active phase. This study also confirmed circadian rhythms in plasma and urinary phosphate in male WT C57BL/6 mice, with the peak of plasma phosphate occurring just before the onset of activity and the peak of urinary phosphate ∼3 h later. Demonstrating a critical role for NPT2a and c in the circadian control of phosphate balance, these investigators showed that double KO of Npt2a and Npt2c results in flattening of the curves for both plasma and urinary phosphate. These results are consistent with a model in which NPT2a and NPT2c have higher activity during the rest phase, when food intake and thus phosphate intake is low, serving to maintain plasma phosphate levels within the normal range. During the active phase when food intake and phosphate intake is high, protein levels of NPT2a and c are lower to maintain phosphate balance.

The PT is also responsible for reabsorbing ∼65% of the filtered Na⁺ load. In the brush border of the proximal convoluted tubule, NHE3 is one of the primary transporters involved in reabsorbing Na⁺ (124). In addition to NHE3, the low-affinity Na⁺-glucose cotransporter 2 (SGLT2) in the proximal convoluted tubule and high-affinity SGLT1 in the proximal straight tubule both help to reabsorb Na⁺, with SGLT1 and SGLT2 reabsorbing Na⁺ with glucose that is filtered through the glomerulus (125, 126).

Several studies have linked the circadian clock mechanism to transporters in the PT (FIGURE 4B). The CLOCK:BMAL1 heterodimer binds directly to the E-box binding domain on the NHE3 promoter (127). This transactivation of NHE3 by CLOCK and BMAL1 can be repressed by PER2 and CRY1. In a different study involving male 129/sv mice, the pharmacological blockade of PER1 with the casein kinase 1δ/ε inhibitor PF670462 resulted in decreased mRNA expression of NHE3 and SGLT1, while SGLT2 was left unchanged. The same results in expression were confirmed with RNA interference targeting PER1 in HK-2 cells (128). The sites of decreased transcription on NHE3 and SGLT1 were exon 7 and exon 3, respectively, and led to decreased occupancy of CLOCK and PER1 on the promoter regions of both NHE3 and SGLT1.

In both human PT cells and rodent renal cortex, pharmacological blockade of PER1 reduced the expression levels of NHE3 and SGLT1 while leaving SGLT2 expression intact. Whereas NHE3, SGLT2, and SGLT1 regulate electrolyte transport on the apical membrane of PT cells, the organic anion transporter 1 (OAT1) and OAT3 act on the basolateral membrane and are involved in the tubular secretion of creatinine (129). Using a mouse model with an inducible conditional KO of BMAL1 in renal tubular cells (Bmal1^lox/lox/Pax8-rtTA/LC1), Nikolaeva et al. (130) found that these mice had lower mRNA and protein expression of OAT3, along with an impaired natriuretic response to the loop of Henle diuretic furosemide and decreased excretion of furosemide. Together these results show the great level of influence the circadian molecular clock has on PT function.

4.2.3. Loop of Henle.

In the loop of Henle, Na⁺, K⁺, and chloride (Cl⁻) ions are actively reabsorbed from the lumen while remaining impermeable to water through the Na⁺-K⁺ -2Cl cotransporter (NKCC2) (131). Renal NKCC2 is encoded by the Slc12a1 gene and is expressed exclusively in the apical membrane of the thick ascending limb (TAL), whereas the Slc12a2 gene encodes for NKCC2 expressed on the basolateral membrane in both epithelial and nonepithelial cells. Crambert and colleagues (132) demonstrated that the estrogen-related receptor b is expressed with a circadian pattern in the kidney and has a role in the regulation of NKCC2 (FIGURE 4C). NKCC2 is negatively regulated by the Alström syndrome 1 protein (ALMS1) via endocytosis (133). Alms1 KO rats exhibit decreased NKCC2 endocytosis in the TAL and display salt-sensitive hypertension, demonstrating the importance of this protein in BP regulation. Regrettably, BP from this study was not evaluated for differences in circadian rhythm. However, interrogation of CircaDB (104) revealed that, at least in the mouse kidney, Alms1 gene expression exhibits a significant circadian rhythm. These findings raise the intriguing possibility that ALMS1 is a circadian clock target and could play a role in time of day-dependent regulation of NKCC2.

Tokonami et al. (134) studied the effects of Bmal1 KO in male mice that have a conditional KO of Bmal1 in cells that contain the endogenous renin promoter (Bmal1^lox/lox/Ren1^dCre). After each section of the nephron was examined, these mice showed a significant reduction of BMAL1 in the TAL, along with the cortical and outer medullary collecting duct (CCD and OMCD). Despite this difference in Bmal1 mRNA expression, the effect of Bmal1 KO on NKCC2 expression or activity is unclear. NHE3 expression appears to have rhythmic activity in the thin descending limb and TAL of the loop of Henle, as evidenced by mRNA abundance in male C57BL/6 mice that were placed in conditions of constant darkness (135). The interaction of the circadian clock with NHE3 in the loop of Henle is not clear and warrants further study.

4.2.4. Distal convoluted tubule and connecting tubule.

The distal tubule is located just downstream of the macula densa and is the shortest portion of the nephron (136). The initial portion of the distal tubule, the DCT, is split up into the early (DCT1) and late (DCT2) segments. The major transporter that is expressed in the DCT1 is the thiazide-sensitive Na-Cl cotransporter (NCC). NCC is an electroneutral transporter that reabsorbs the majority of Na⁺ and Cl⁻ seen in the DCT into epithelial cells across the apical membrane. NCC is encoded by the gene Slc12a3 and is part of the SLC12 family of electroneutral cation-Cl⁻ coupled cotransporters. NCC is also expressed in the DCT2, along with ENaC. The DCT2 leads into the CNT and CCD, and these three regions together make up the region known as the aldosterone-sensitive distal nephron (FIGURE 4D). What is known about circadian clock-mediated regulation of gene expression in the CNT comes from a landmark study by Zuber et al. (88), which is discussed in more detail in sect. 4.3.

Numerous studies have examined the influence of circadian rhythms in the distal tubule. In male C57BL/6J mice, Susa et al. (137) found that downstream targets of the With No Lysine (K) (WNK) signaling pathway phosphorylate NCC, which include oxidative stress responsive kinase 1 (OSR1) and Ste20-related proline-alanine rich kinase (SPAK), which display circadian rhythms in phosphorylated OSR1, SPAK, and NCC. The circadian variation observed in this signaling cascade could be the result of a direct interaction with the circadian clock. Richards et al. (138) showed that in male WT and Per1 heterozygous mice on a 129/sv background reduced expression of PER1 led to decreased NCC and WNK1 mRNA expression and increased WNK4 mRNA expression. In murine DCT cells, pharmacological blockade of PER1 with the casein kinase 1δ/ε inhibitor PF670462 also resulted in the same change in expression for NCC, WNK1, and WNK4. These decreases or increases in mRNA expression correlated with decreased or increased abundance of PER1 and CLOCK, icates that the mRNAs for WNK1 and SPAK (Stk39) exhibit significant circadian rhythmicity in the whole mouse kidney.

The loss of pT53 NCC diurnal variation as a result of corticosterone treatment phenotypically manifested as nondipping BP (139). When looking at pT53 NCC protein after administration of a combination of chronic (vehicle, spironolactone, or RU486) and acute (vehicle or corticosterone) treatments, only RU486 blocked the increase in pT53 NCC that is seen after administration of corticosterone (140).

With recent research focused on developing novel molecules to inhibit components of the WNK-SPAK/OSR1-NCC kinase signaling pathway for treatment of salt-sensitive hypertension (reviewed in Refs. 141, 142), understanding the role of the renal clocks in regulation of this pathway is likely to yield information with the potential to be leveraged for optimizing the treatment of hypertension in humans.

4.2.5. Collecting duct.

The CD is the last major segment of the nephron and is split up into three sections: CCD, OMCD, and inner medullary (IMCD). Although ∼1% of Na⁺ filtered through glomerulus reaches the CD (143), this segment is highly regulated by different physiological pathways to balance Na⁺ intake with output. The major Na⁺ channel in this region is ENaC, an amiloride-sensitive Na⁺ channel that is made up of three main homologous subunits: α, β, and γENaC (144). A fourth subunit, δENaC, has been identified in humans (145). The ENaC subunits are encoded by the genes Scnn1a, Scnn1b, Scnn1g, and Scnn1d. ENaC, along with the renal outer medullary K⁺ (ROMK) channel and the AQP2 channel, are the major channels reabsorbing Na⁺, K⁺, and water on principal cells in the CD (146). Cl⁻ and acid/base balance is maintained in cotransporters located on intercalated cells (147) (FIGURE 4E).

Numerous studies have examined the role of the circadian clock and circadian rhythms in the CD. We found that PER1 reduces the effectiveness of aldosterone on ENaC mRNA expression by >50% in each portion of the CD examined (148). We conducted these studies in cell models representing all three segments of the CD and produced similar results on αENaC in each, expanding our knowledge of what we know about the interaction between the circadian clock and the kidney. This study opened a new area of research into the kidney clock. This study also showed that ENaC expression has a circadian rhythm, and that mutations in PER1 abrogate that rhythm.

Per1 KO mice have a reduced level of αENaC expression in the renal cortex compared with WT mice. This reduction of PER1 expression inhibited αENaC expression after aldosterone treatment in mpkCCD₁₄ cells, a murine renal CCD cell line. This interaction between PER1 and αENaC is due to the interaction of PER1 with E-box response elements on the αENaC promoter, with E-box 3 showing the strongest PER1 signal (149). We also went on to demonstrate that CRY2 expression was increased when PER1 expression was reduced and that CRY2 and PER1 appear to mediate opposing action on αENaC expression in the mpkCCDc14 cell line (150).

More recently, Zhang et al. (151) examined the role of Bmal1 in the CD, using a principal cell-specific Bmal1 KO (CD-Bmal1) mouse model. Male CD-Bmal1 KO mice had decreased mean arterial pressure (MAP) under normal-salt diet, high-salt diet, and high-salt diet with ET_B receptor antagonist. Female CD-Bmal1 KO mice, however, showed no change in MAP from control mice under any diet/treatment. Neither sex showed any difference in the diurnal rhythm of urinary Na⁺ excretion, K⁺ excretion, or aldosterone excretion, suggesting that Bmal1 in the CD has a sex difference in the regulation of BP and this regulation is independent of electrolyte excretion at different times of the day. This independent regulation of BP and electrolyte excretion mediated by Bmal1 was also suggested by Johnston et al. (152), involving a study that utilized the first known whole body clock gene KO rat model.

These studies highlight the benefit of using diverse models, including cell, mouse, and rat models, in understanding the regulation and function of the kidney clocks. Diverse cell and animal models should be utilized in order to better approximate human physiology and pathophysiology.

4.3. Transcriptomic Approaches to Understanding the Kidney Clocks

Our laboratory first became interested in circadian clock genes in the kidney after a microarray study we performed in mouse IMCD-3 cells (105). Cells were grown past confluence on Transwell dishes to mimic the intact tubule as much as possible and then treated with vehicle or aldosterone for 1 h. The most highly induced gene in the entire study was Period homolog, now known as Per1. We confirmed this result with Northern blots and qPCR. In subsequent studies we confirmed that aldosterone treatment increased PER1 protein levels (149).

There have been numerous microarray studies that have explored the influence of circadian clock physiology in the kidney. In a study by Kita et al. (153), the effect of time of day and feeding state on gene expression was examined in a number of genes identified by microarray analysis in the kidney and liver of male Dahl salt-sensitive rats. These animals were maintained on a 12:12 LD cycle and had their tissues harvested at two time points: ZT2 and ZT14. The rats that were fasted were only given water for 24 h before organ collection. The analysis revealed 597 differentially expressed genes or expressed sequence tags (ESTs) out of a total of 8,448 unique genes. Twenty-four of these differentially expressed genes met the criteria of central clock or clock-controlled genes, because of their time-of-day difference in expression that was independent of feeding state. In the kidney, this included Bmal1, Per1/2/3, and Dbp.

Zuber et al. (discussed above in sect. 3.3) used male C57BL/6J mice to perform microarray experiments to analyze the temporal expression of numerous transcripts (88). In this landmark study, RNA was isolated from the DCT/CNT or CCD. These animals consumed a standard chow diet and adapted to a 12:12 LD cycle for 2 wk before tissue extraction. The mice were euthanized for microdissection every 4 h starting at ZT0 (lights on), for a total of six time points. Microarray analysis of these tissues resulted in 5,031 transcripts in the DCT/CNT and 2,765 transcripts in the CCD that exhibit differential expression. These transcripts were linked to 3,814 genes in the DCT/CNT and 2,112 genes in the CCD. To test whether these transcripts met circadian criteria, the mRNA expression patterns were fitted to a cosine curve with a 24-h period. With these criteria, 356 DCT/CNT circadian transcripts and 504 CCD transcripts were designated as circadian expressed genes. Among these, 96 transcripts were found to be expressed in a circadian manner in both the DCT/CNT and the CCD. Not surprisingly, the overlapping circadian transcripts included the core clock genes Bmal1, Clock, Per2/3, Cry1/2, Dbp, Npas2, and Nr1d1/2. Furthermore, the Clock−/− mouse model used in this study exhibited impairments in water and Na⁺ excretion at different times of the day.

Oike et al. (154) showed that feeding young adult male BALB/c mice high-salt diet for >2 wk decreased Bmal1 and increased Dbp mRNA expression in the kidney. These changes in expression under high-salt diet occurred along with a phase advance, by just under 2 h for Dbp and ∼3 h for Bmal1. Despite these results in the kidney, cDNA microarray analysis for this study was performed only in the liver.

Vedell et al. (155) examined transcript abundance variation in the kidneys, along with adipose tissue, heart, and liver, in young adult male C57BL/6J mice. Several genes known to exhibit a circadian rhythm, including core clock genes Per1 and Per2, displayed variations in transcript abundance between different mouse samples used in the kidneys. However, the authors noted that this difference could be expected if the mice are in slightly different phases in their diurnal cycles, despite controlling for the time of day and feeding/fasting cycle.

A study by Langen et al. (156) looked to characterize the impact that diurnal variation may have on iodine-131 (¹³¹I)-induced effects in peripheral mouse tissues, including renal cortex and medulla sections. ¹³¹I was used to simulate the effects of ionizing radiation, and tissue samples from female BALB/c nude mice were collected for RNA microarray analysis to examine ¹³¹I-induced genomewide transcriptional regulation. Tail vein injections of ¹³¹I or physiological saline as a control were administered at three time points: 9:00 AM, 12:00 PM, and 3:00 PM. After injection, on the same day, these mice were kept in the dark from 5:00 PM until 8:30 AM the following day. Tissues were collected 24 h after the original injection time. The number of significantly regulated transcripts in the renal cortex after ¹³¹I injection was highest at 12:00 PM, with 105 combined up- and downregulated transcripts, compared with 21 transcripts at 9:00 AM and 35 at 3:00 PM. Of these transcripts, only one was differentially regulated at all three time points (Lap). The renal medulla also had the highest number of significantly regulated transcripts at 12:00 PM, with 444 total, compared with 358 transcripts at 9:00 AM and 17 at 3:00 PM. Of these transcripts, six were differentially regulated at all three time points (Cldn11, Hmgcs2, Klk1b5, Psca, Slpi, Ubd). The authors concluded that in radiation research circadian rhythms should be considered as an experimental variable.

Nikolaeva et al. (130) used RNA-seq to investigate the kidney transcriptome at two time points, ZT4 and ZT16, using whole kidneys from control and Bmal1^lox/lox/Pax8-rtTA/LC1 conditional KO mice, where BMAL1 is conditionally inactivated in renal tubular epithelial cells. Gene analysis revealed 721 differentially expressed transcripts between genotypes at ZT4, 765 transcripts at ZT16, and 552 at both time points. Exploring the 552 transcripts further, gene ontology analysis showed that these transcripts contributed to biological processes including carboxylic acid metabolism, organic anion transport, and chemical homeostasis. This was illustrated by a significant difference in renal metabolism, as KO mice displayed a reduced NAD⁺-to-NADH ratio in the kidney compared with control mice, despite having similar levels of renal mitochondrial DNA.

Circadian transcriptome analysis was also performed in a conditional KO mouse model in which Bmal1 was specifically knocked out in podocytes (Bmal1^lox/lox/Nphs2-rtTA/LC1) (119). Glomeruli were isolated every 4 h starting at ZT0 to profile differentially regulated genes across the 24-h cycle. This approach identified a number of novel BMAL1-target genes. These included several genes encoding proteins that function in podocyte development, including G protein-coupled receptor class C, group 5, member A (Gprc5A), transcription factor 21 (Tcf21), sulfatase 2 (Sulf2), Cathepsin L (Ctsl), N-Ethylmaleimide Sensitive Factor (Nsf), and G Protein Subunit Alpha 12 (Gna12) (FIGURE 4A). Confirming that deletion of BMAL1 in podocytes altered the circadian clock mechanism, Cry1, Rora, Rorc, and Npas2 genes were all differentially expressed in glomeruli from the podocyte-BMAL1 KO compared with control mice. Importantly, these changes in mRNA were associated with functional changes: podocyte-Bmal11 KO mice excreted higher levels of creatinine during the rest phase compared with control mice.

Nakashima et al. (157) examined the role of the circadian clock mechanism in the regulation of BP, using the clock protein DEC. ChIP-on-chip analysis was used to examine whether there were BP related genes targeted by DEC1. Atp1b1, the gene encoding the β1-subunit of the Na⁺-K⁺-ATPase, was one of the transcripts identified with E-box elements in its promoter. In human mesenchymal cells DEC1 and CLOCK were shown to bind to the E-box binding domain of the Atp1b1 promoter (157). This regulation of Atp1b1 by DEC1 and CLOCK manifested in a diurnal variation of Atp1b1 in the kidney, along with aorta and heart, with mRNA expression levels highest at ZT10. Male Dec1−/− mice on a C57BL/6J background showed an exaggerated difference in diurnal expression, with mRNA expression significantly higher than WT at ZT10 in all three tissues. In contrast, male Clock-mutant mice (Ck/Ck) on a BALB/c background had reduced levels of Atp1b1 compared with their WT controls, abolishing the diurnal variation in mRNA expression (157). Interestingly, Dec1−/− mice maintained the circadian variation in BP, at significantly reduced levels, whereas Ck/Ck mice had their BP rhythms abolished because of elevated values during the inactive period. This study helped provide direct evidence of circadian clock proteins regulating BP via Na⁺-K⁺-ATPase signaling.

The transcriptomic studies discussed above were performed with microarray technology or bulk RNA-seq. One of the most exciting developments in the kidney field in the last several years has been the use of single-cell and single-nucleus RNA-seq (158–160). These powerful methods have been performed on normal and pathological (i.e., lupus) human kidney samples, organoids grown in culture, and mouse kidneys, yielding a tremendous amount of information and allowing the identification of distinct cell types in these samples. The expense of performing these studies together with the vast computational power required for the analysis has so far precluded inclusion of time of day as a biological variable. An important issue to consider is that the samples that have been the subject of these powerful studies, whether human or mouse, were likely collected during “normal business hours.” Moreover, the time of day of collection is rarely, if ever, noted in the primary literature. Given that the circadian clock controls the expression of thousands of genes in the kidney, future transcriptomic studies, especially those done in rodent models, will have more translational power if time of day is at least noted so that it can be considered in the interpretation of these data. This is especially critical for relating data from nocturnal rodents to that of diurnal humans.

5. REGULATION OF THE KIDNEY CLOCKS

Although there is growing knowledge of the kidney clocks’ role in regulation of processes contributing to renal function, there is still a gap in our knowledge of how the kidney clocks are regulated and how that regulation changes in pathophysiological conditions. In this section, we review evidence and postulate potential mechanisms behind the regulation of the kidney clocks.

5.1. Central Clock and Behavior

It is widely accepted that the central clock in the SCN is entrained by light cues to act as a pacemaker, driven by information from the retino-hypothalamic tract and neuronal signals to synchronize peripheral clocks, including the kidney clocks, throughout the body (161, 162). A study by Wu et al. (163) looked directly at the effect of light cues on the synchronization of renal circadian clock gene expression in male Wistar rats, using reversal of the LD cycle for 7 days. Per1 displayed a 4-h phase delay following reversed LD cycle, with no changes in other clock gene expression in the kidney, suggesting clock gene-specific light-induced regulation. However, evidence suggests that peripheral clocks, including the kidney clocks, are self-sustained oscillators. This was demonstrated by Yoo et al. (164), who showed that lesioning the SCN, where the central clock resides, in PER2::LUC mice did not abolish circadian rhythms in peripheral tissues, including the kidney, but caused phase desynchrony. This phase desynchrony involved the period, or the time taken for a full circadian cycle, from one peak of the oscillation to the next, being shortened in the kidney but lengthened in the pituitary in SCN-lesioned mice. This would suggest that cell- and organ-specific synchronization mechanisms exist and that the SCN acts as a “phase synchronizer” instead of a “pacemaker,” driving peripheral oscillations. Interestingly, a recent study has initial evidence to suggest kidney clock-to-SCN feedback (165). This study used an adenine-induced model of CKD in mice and found impaired behavioral rhythms: activity was disorganized and decreased specifically in the active period, with an altered circadian period in CKD mice. Since the SCN is intact in these mice, they proposed that kidney clock-induced changes in fluid balance may impact the central clock and alter circadian rhythms at the behavioral level.

5.1.1. Feeding.

The timing of food intake is an important external timing cue (zeitgeber) that can synchronize peripheral circadian clocks, contributing to regulation of physiological functions (166). Evidence has shown that timing of food intake has implications for BP homeostasis. Daytime-restricted feeding (with a diet for maintenance of body weight) in human studies showed BP-lowering effects (167, 168), Interestingly, restriction of feeding to late afternoon/evening caused increases in BP in a separate study (169). These studies used a single time point to measure BP; the effect of time-restricted feeding on BP rhythms in humans has not been assessed.

Zhang et al. (65) investigated the timing of food intake on BP rhythm in mice, where mice were subjected to restricted feeding during their inactive phase (reversed feeding). This resulted in inverted BP rhythm, with no change in average 24-h BP and no change in night/day patterns of renal Na⁺ excretion. These findings were shown to be independent of BMAL1. The effect of time-restricted feeding on PER2 was then assessed in PER2::LUC mice. PER2 in the SCN was not affected, but restricted feeding induced a phase shift in the renal inner medulla, liver, and adrenal gland. The role of the other clock components has not been explored. As mentioned above, Wu et al. (163) investigated 7-day reversed LD cycles on kidney clock gene rhythms in male Wistar rats. They also assessed 7-day reversed feeding alone, as well as reversed feeding with reversed LD cycles together. Reversed feeding alone showed no significant changes to circadian phases of the clock genes in the kidney of rats. However, with both feeding and LD cycle reversed, phases of Per1 and Clock were shifted rapidly by 12 h in only 3 days. BP was not measured in the Wu et al. rat study. It should be noted that, in contrast to the Zhang et al. findings in mice, reversed feeding did not alter the BP rhythm in male Sprague-Dawley rats (170). Strikingly, Rhoads et al. (170) showed that rats subjected to reverse feeding of a high-salt diet exhibited loss of the normal night/day pattern in renal Na⁺ excretion. Effects on clock gene expression were not assessed. These studies suggest that the role of feeding in regulating the kidney clocks, kidney function, and BP is complex, with species-specific effects.

Together, these studies suggest that feeding could influence synchronization of the kidney clocks, having implications on BP rhythm. However, the mechanisms behind this remain unclear. Studies assessing the role of light and food as zeitgebers of peripheral clocks tend to focus on the liver and other extrarenal tissues (166). Therefore, future studies are needed to better understand how food and light affect the kidney clocks, with an emphasis on differences between cell types in the kidney. Given the observed differences between mice and rats and the lack of human studies in this area, the use of diverse animal models could provide deeper understanding of how the kidney clocks are regulated by food intake.

5.1.2. Melatonin.

Melatonin, a key regulator of the sleep/wake cycle, is synthesized in the pineal body and is suppressed by light (171). Melatonin communicates with the central clock, to relay information on the LD cycle and regulate various physiological functions. Melatonin plays a major role in regulating sleep/wake cycles, but it is also a powerful antioxidant that limits oxidative stress and maintains bioavailability of nitric oxide. Interestingly, low melatonin levels have been associated with a gradual decrease in renal function (172, 173). This also seems to be accompanied by altered melatonin rhythm, as melatonin circadian rhythm has a reduced amplitude and altered acrophase (the time at which the peak of a rhythm occurs) in CKD patients, and it was reported that patients receiving nocturnal dialysis experienced a normal peak in melatonin and recorded better sleep quality compared with daytime dialysis patients (174). Interestingly, melatonin treatment decreased intrarenal RAAS activity in a rat model of CKD and decreased the burden of kidney injury (175). However, the mechanisms behind this association are not fully understood. This is challenging to investigate with animal models, as many mouse strains (C57BL/6J, BALB/c, and 129/sv) are deficient in melatonin, with only CBA and C3H strains considered melatonin proficient (176). Furthermore, melatonin peaks at night, whether active or sleeping, so with mice being nocturnal this may be difficult to relate to humans, although the melatonin receptors in mice appear to have a pharmacological profile similar to human receptors (177). Together, these studies show a link between melatonin and the kidney, but data on the relationship between melatonin and kidney clocks are lacking. Work investigating the kidney-SCN clock cross talk could be important for understanding this behavioral circadian adaptation in CKD and whether this impacts melatonin homeostasis. However, care will need to be taken in the rodent model used to investigate this, given the striking mouse strain melatonin differences.

5.1.3. Core body temperature.

Core body temperature rhythmicity is dependent upon the circadian clock and is also affected by the sleep/wake cycle in humans and rodents (178). Changes in core body temperature have also been shown to cause phase shifts of peripheral clocks. Specifically, when PER2::LUC knockin mice were placed in a water bath for 1–2 h during the inactive period over 2 days, core body temperature increased to 40–41°C and resulted in a phase advance in the kidney clock (179). Thus, body temperature appears to be both an input signal to the clock, particularly to the kidney clock, as well as an output of the circadian clock.

5.2. HPA Axis

Glucocorticoids, as mentioned in sect. 3.2, are regulated by the clock and are important for maintenance of renal function by regulating renal blood flow, GFR, renal vascular resistance, and water and electrolyte metabolism (70, 71). Glucocorticoids, the hypothalamic-pituitary-adrenal (HPA) axis end-effectors, are also crucial for the normal function of the HPA axis, a highly conserved regulatory network vital for survival (180). Glucocorticoids have been suggested to regulate the kidney clocks (139, 140), and therefore alterations in the HPA axis may also have a profound effect on renal function and the kidney clocks. The HPA axis responds to both circadian and stress-related stimuli. It is activated by environmental, physical, or physiological stressors, including circadian input from the SCN and stress inputs from the brain stem and limbic regions (including the hippocampus) in the brain, resulting in the stimulation of the hypothalamic paraventricular nucleus (PVN). The parvocellular neurons of the PVN project to the median eminence, which is at the base of the hypothalamus and serves as connection between the neural and peripheral endocrine systems, to stimulate the release of corticotropin-releasing hormone (CRH) and AVP into the hypophyseal portal vessels. CRH acts on CRH type 1 receptors on anterior pituitary corticotropic cells, which stimulates the activation of proopiomelanocortin (POMC) transcription for the secretion of POMC-encoded adrenocorticotropic hormone (ACTH). ACTH levels exhibit a circadian rhythm, reaching a peak in the morning and declining throughout the day in humans (181). ACTH enters the systemic circulation and binds to the G protein-coupled receptor melanocortin (MC) receptors (MC2R/MC5R) in the adrenal cortex, where a cascade of processes occurs to promote the synthesis and secretion of glucocorticoids (reviewed in Refs. 180, 182).

Interactions between clock components, glucocorticoids, and their targets have been reported, but little is known about their impact on the kidney, which is a major site of excretion of cortisol and its metabolites (70). Decades ago, Moore-ede et al. (183) conducted studies in adrenalectomized (ADX) diurnal squirrel monkeys that demonstrated a role for cortisol in the regulation of renal K⁺ excretion. Interestingly, Poulis et al. later studied ADX rats and found that urine rhythm persisted in the absence of corticosterone but renal function became desynchronized from activity rhythms (219). Corticosterone replacement in ADX rats caused a phase shift in urine flow rhythms, indicating a role for glucocorticoids in phase setting. This was supported with studies showing glucocorticoids regulating Per1 and Per2 through the glucocorticoid response element (GRE) and E-box in the promoter region resulting in advanced and delayed circadian phases in mice, respectively (184, 185) (FIGURE 6). Additionally, it was shown that glucocorticoid-mediated Per2 induction was reliant on BMAL1-dependent binding of its receptor, GR, to the GRE/E-box region (185). Interestingly, glucocorticoid-induced phase-shifting has not been reported in Per1 KO mouse models. This would suggest that Per1 plays a major role in mediating the phase-shifting effects of glucocorticoids. Unlike Per, Rorα and Reverbα contain a functional negative GRE (186, 187). Studies in cells and mice have also demonstrated that clock components may interact directly with the GR. CLOCK has been reported to regulate transcriptional activation by GR by inhibiting its ability to bind DNA (188). Lamia et al. (189) demonstrated a role for CRY1 and CRY2 in opposing GR-mediated transcriptional activation via direct interaction with GR. Together, these data highlight cross talk between glucocorticoids and the circadian clock, affecting the status of glucocorticoid rhythm and GR activity. This cross talk could regulate the renal clocks with implications for renal function, including renal Na⁺ transport, as glucocorticoid binding to GR leads to activation of NHE3 in the PT (190) and influences the diurnal pattern of variation in NCC activity (139, 140).

FIGURE 6.

Integration of glucocorticoid signaling and circadian clock-mediated transcriptional regulation. Glucocorticoids (GCs) bind to their receptor, the glucocorticoid receptor (GR), leading to translocation of GR into the nucleus. GR regulates the transcription/translation feedback loops of the circadian clock mechanism through glucocorticoid response elements (GREs) of Per1/2 and negative GREs (nGREs) of Reverbα and Ror. Both positive and negative arms of the circadian clock inhibit GR activity. Diagram created with BioRender.com, with permission.

There is a plethora of pathological conditions associated with chronic glucocorticoid excess that have negative impacts on renal rhythms, affecting overall renal function. Furthermore, disruption in the rhythmicity of glucocorticoid release has implications for BP rhythm and renal function and has been suggested to dysregulate circadian clock expression (140, 191, 192). Ivy et al. demonstrated this, as the infusion of corticosterone, which clamps circulating corticosterone levels at midphysiological range with diurnal variation attenuated, in male C57BL/6J mice resulted in nondipping BP. This phenotype was due to GR-mediated elevation in phosphorylated renal NCC during the inactive period, and these mice exhibited increased whole kidney expression of Per1, Per2, Cry1, and Bmal1 genes (139, 140). Whether this impacts activity of the renal clocks and the implications of increases in renal core clock gene expression remains unknown. Dysregulation of glucocorticoid rhythm has been illustrated in end-stage CKD patients, with an observational study showing that these patients had increased night plasma and salivary cortisol levels, blunting the normal diurnal rhythm of cortisol (193).

Glucocorticoids regulate the activity of the HPA axis through feedback mechanisms by binding to its receptors, GR and/or MR, in the anterior pituitary, hypothalamus, and hippocampus (reviewed in Refs. 68, 69). Renal handling of glucocorticoids may affect their production through disruption of these negative feedback loops. Thus, decreased renal function and/or disruption of the kidney clocks negatively affects glucocorticoid handling and may subsequently alter its production inappropriately. Dysregulation of the HPA axis can exert an array of adverse outcomes and impact the organism’s ability to respond to stressful stimuli. Responses to a stressor are also circadian in nature, and acute external stress, in the form of a restraint stress test, acted as a synchronizer of peripheral circadian clocks, including the kidney clocks, which was more responsive to stress compared with the liver, in PER2::LUC knock-in mice (194). This entrainment was independent of sleep/wake and body temperature cues. Daily restraint stress induced a phase advance of the kidney clock that persisted for 2 wk of stress testing, suggesting that the kidney clock was completely entrained by this chronic stress. Clock gene expression in the kidney, as well as other tissues such as the liver and hippocampus, was also entrained to the stressor but not in the SCN. The authors suspected that this was due to the lack of GR in the SCN. An interesting observation in this study was that when the restraint stress was performed at ZT0–2, PER2 oscillations were completely abolished in the kidney. This phenomenon is known as singularity behavior, i.e., suppression of circadian rhythms by a critical stimulus (195). This could have implications on renal rhythms that rely on the kidney clocks and could be detrimental to patients with declining renal function.

Glucocorticoids are also partly involved in the regulation of the circadian rhythm of the immune response. Generally, glucocorticoids suppress inflammation, but they have the capability of having immune-enhancing effects, to regulate the circadian oscillation of the innate and adaptive immune response; this topic was recently reviewed by Shimba and Ikuta (196). This could have implications for renal rhythms and function, especially for CKD patients, although if glucocorticoid rhythms are impaired, the effect this has on the immune response remains in question, and whether it has a role in initiating/progressing renal fibrosis would be of interest.

The abundance of this work has been done only in males, with a substantial gap in our knowledge regarding circadian clock regulation in females. Glucocorticoid rhythm is influenced by the estrus cycle, where glucocorticoid rhythm amplitude is lowest during estrus (peak estrogen levels) (197). This suggests that not only time of day matters but so too does time in the estrus cycle for women. Future work to address this gap in our knowledge is vital to understanding the implications of glucocorticoid rhythms throughout the estrus cycle and postmenopause for renal rhythms and renal function.

Together, these data suggest that day/night patterns in glucocorticoid concentrations, regulated at least in part by the adrenal clock, play a major role in the regulation of renal rhythm, including the renal clocks, with alterations in glucocorticoid rhythm having a negative impact on renal function. Furthermore, central and peripheral circadian clocks certainly contribute to the regulation of the HPA axis (198, 199) (FIGURE 7), but the role of the kidney clocks in these processes remains poorly understood and should be a focus of future investigations.

FIGURE 7.

Integrative physiology of the kidney clocks. The central clock in the suprachiasmatic nucleus (SCN) directly receives photic cues for entrainment to the light-dark cycle to synchronize peripheral clocks including the kidney clocks. The kidney clock is responsive to entrainment from nonphotic cues including feeding, core temperature, and stress. Hormone signaling, including adrenal hormones (aldosterone and cortisol), vasopressin (AVP), and endothelin-1, has been shown to regulate the circadian clock. With stress and glucocorticoids involved in entrainment of the kidney clock, the hypothalamic-pituitary-adrenal (HPA) axis plays a key role in control of the kidney clock, but the kidney clocks’ role in the regulation of the HPA axis remains unclear. The clocks in the kidney have been suggested to communicate with the central clock and adrenal clock, but the mechanisms are not fully understood. Diagram created with BioRender.com, with permission.

5.3. Autonomic Nervous System

The autonomic nervous system contributes to regulation of BP rhythm and renal functions, including GFR, water and electrolyte handling, and hormone release (200, 201). Therefore, it could play a role in the regulation of the kidney clocks. There is a large gap in our knowledge of autonomic control of the circadian clock, and this should be a focus of future investigations. In an elegant review by Becker et al. (202), the authors suggested the potential role of autonomic control of renal function in circadian clock gene expression. They hypothesized that timing of afferent and efferent nerve activity influences clock gene expression. Future studies to explore this should be a focus. Renal sympathetic overactivity and disruption in BP rhythm are more likely in patients with hypertension and CKD (201, 203, 204). Renal denervation has been shown to reduce BP and improve renal function in animal models of hypertension and CKD (205). Early trials in renal denervation have had conflicting outcomes, but there remains promise in ongoing clinical trials (reviewed in Ref. 206). Interestingly, in a small pilot study, renal denervation increased dipping status in CKD patients, suggesting a role for renal nerves in regulating BP rhythm (207). These studies highlight the need to better understand the relationship between the autonomic nervous system and the kidney clocks because these findings could have clinical implications for improving circadian rhythms in pathophysiological conditions such as CKD.

6. REMAINING QUESTIONS

6.1. Does Knockout of One Clock Gene in Kidney Cell Types Truly Disrupt the Kidney Clocks?

Perhaps one of the most surprising findings in recent studies of the kidney clock is that KO of BMAL1 in the kidney does not dramatically affect rhythmic renal function. Firsov and colleagues (134) were the first to generate a BMAL1 KO within the kidney using renin-Cre. This led to the deletion of BMAL1 in renin-producing cells of the juxtaglomerular apparatus and the CD as well as partial knockdown in the medullary TAL. Only male mice on normal chow diets were used for this study, which demonstrated that renin-Cre-driven KO of BMAL1 resulted in lower BP without a change in BP rhythm, subtle changes in urinary Na⁺ excretion during the active phase, and increased GFR. The 24-h variation in plasma aldosterone was altered in renin-Cre Bmal1 KO mice, with the KO mice exhibiting higher plasma aldosterone at several time points compared with control mice. Nikolaeva et al. (130) generated a conditional kidney Bmal1 KO using Pax8-rtTA/LC1 Cre, which resulted in BMAL1 deletion in adulthood (8- to 10-wk-old mice) in renal tubule cells. In contrast to the results observed in renin-Cre Bmal1 KO mice, the Pax8-inducible Bmal1 KO mice did not have alterations in GFR, Na⁺ excretion rhythms, or plasma aldosterone levels. Like the previous study using renin-Cre Bmal1 KO mice, the Pax8-Bmal1 KO study was limited to male mice on a normal chow diet. RNA-seq analysis of the whole kidney was performed in this mouse model, coupled with plasma metabolomics. Samples were collected from control and KO mice at two time points, ZT4 and ZT16. This unbiased “omics” approach revealed a role for tubular BMAL1 in metabolism and drug disposition, as discussed above. The Pax8 Cre Bmal1 KO mice did exhibit a lower BP phenotype, but the BP rhythm remained intact.

Firsov and colleagues (119) went on to generate an inducible, podocyte-specific KO of BMAL1 using nephrin-Cre as discussed above. The circadian rhythm of plasma aldosterone was also altered in the podocyte-Bmal1 KO mice, with aldosterone levels significantly lower in the KO at ZT4 and ZT20. Subtle changes in urinary Na⁺ excretion rhythms were seen in the podocyte-specific Bmal1 KO mice. Transcriptomics profiling of glomeruli revealed novel BMAL1 target genes related to podocyte function, as discussed above. BP was not significantly different between control and KO in these studies, nor was the rhythm of BP affected by the KO. Like the previous kidney-specific Bmal1 KO studies, this study was also limited to male mice on a normal chow diet.

Pollock and colleagues generated a CD-specific Bmal1 KO mouse model. Zhang et al. (151) used Aqp2-Cre to generate BMAL1 deletion in the CD. As others have recently shown (208, 209), Aqp2-Cre was also active in the brain, so this CD-Bmal1 KO is not strictly limited to the kidney. Consistent with the previous tubule cell-specific KO of Tokonami et al. (134) and Nikolaeva et al. (130), the Aqp2-Cre-Bmal1 KO male mice exhibited lower BP than control mice without any changes to the BP rhythm. Zhang et al. also studied female mice. Genotype-dependent differences in BP were not observed in female Aqp2-Cre Bmal1 KO mice. Assessment of night/day differences in urinary Na⁺ excretion showed that all mice in the study (males, females, controls, and Aqp2-Cre Bmal1 KO) excreted more Na⁺ during the night compared with the day, as expected. To perturb homeostasis in these mice, Zhang et al. administrated a high-salt diet for 6 days followed by 6 additional days on the high-salt diet plus treatment with the ET_B receptor antagonist A-192621. BP was unchanged in female control and Aqp2-Cre Bmal1 KO mice in response to the high-salt diet, but treatment with the ET_B receptor blocker raised BP in both genotypes, as expected. Thus, the effect of high salt or high salt plus ET_B receptor blockade did not tease out a genotype-dependent effect in females. Male mice did not undergo an increase in BP in response to a high-salt diet but did exhibit an increase in response to the ET_B receptor blocker. Throughout the high salt and high salt plus ET_B receptor antagonist treatment, the genotype effect of Aqp2-Cre Bmal1 KO to lower BP was maintained.

In an independent study, we generated a distal nephron/CD KO using Ksp-cadherin Cre to drive BMAL1 deletion (210). Similar to the Aqp2-Cre Bmal1 KO phenotype, these kidney-specific (KS) Bmal1 KO male mice exhibited lower BP compared with control mice. Female KO mice did not exhibit a difference in BP compared with control. Genotype-dependent night/day differences in urinary Na⁺ excretion were not observed in males or in females. We challenged the male control and KS-Bmal1 KO mice with HS/DOCP. We previously showed that this HS/DOCP treatment led to nondipping hypertension (110) and decreased night-to-day ratio of urinary Na⁺ excretion (61) in male global Per1 KO mice. This treatment also led to increased Na⁺ balance in KS-Per1 KO male mice (62). In contrast, HS/DOCP treatment did not affect the BP rhythm in KS-Bmal1 KO mice, nor did it alter the night/day urinary Na⁺ excretion or overall Na⁺ balance (210). These findings highlight the need to better understand the mechanism of the clocks within the kidney. Current paradigms describe BMAL1 as the only truly essential element of the clock mechanism, yet the results from kidney-specific KO models demonstrate a detrimental phenotype in KS-Per1 KO mice but not KS-Bmal1 KO mice.

A question that remains from all of these studies is whether KO of BMAL1 in specific cell types results in true disruption of the kidney clocks. BP rhythms were not altered in any of these kidney BMAL1 KO models (FIGURE 8). Work from the Firsov group (130) demonstrated that Cry1 circadian expression was altered in the kidney of Pax8 Cre Bmal1 KO mice, as Cry1 mRNA levels appeared to stay around the highest level of expression across 24 h. The amplitude of expression for Rev-erbα and Dbp was visibly dampened in the KO mice as well. Although results for these specific genes suggest that the clock was disrupted, significant changes were not found in Per1 (absent on the RNA-seq data), and the change to Per2 expression was subtle. Clock gene expression was not assessed in the other kidney Bmal1 KO models, raising the question of whether the molecular clock may still be functioning in the kidney when BMAL1 is only missing from a subset of cells in this tissue. This is a difficult problem to tackle in the kidney given the heterogeneous cell types in this tissue. There is not a single Cre transgenic mouse line that could be used to generate a whole kidney KO. The available Cre mice can generate KO in specific cell types, specific nephron segments, or the entire tubule but not the vasculature or glomerulus. Alternatively, BMAL1 in the regions of the kidney tested so far may not affect BP rhythms. Additional studies are necessary in order to understand how cells in the kidney may compensate when BMAL1 and/or clock function is lost in a subset of renal cell types.

FIGURE 8.

Blood pressure phenotypes in kidney-specific BMAL1 knockout (KO) mice. Blood pressure tracings over the course of 24 h in mice with various BMAL1 KOs compared with control mice. All panels are based on 12:12-h light-dark conditions while maintained on standard rodent diets. Red shaded areas of nephron demonstrate the location of knockout. A: male global BMAL1 KO mice were the first to be characterized, which lack BMAL1 expression in the entire body including the kidney. MAP, mean arterial pressure. Data from Ref. 51. B: male Nephrin-Cre BMAL1 KO is specific to podocytes in the glomerulus. SBP, systolic blood pressure. Data from Ref. 119. C: male inducible Pax8-Cre BMAL1 KO is specific to the entire nephron. Data from Ref. 130. D: male Renin-Cre BMAL1 KO mice. BMAL1 expression is lost in cells that express renin, which includes juxtaglomerular cells and collecting duct. Data from Ref. 134. E and F: female (E) and male (F) Aqp2-Cre BMAL1 KO (151) is specific to collecting duct cells. Data from Ref. 151. G and H: female (G) and male (H) Ksp-cadherin BMAL1 KO is specific to thick ascending limb, distal tubule, and collecting duct cells. Data from Ref. 210.

Together, these findings raise an important issue, that the clock mechanism is likely more complex than the current paradigm and may function differently across distinct tissues and cell types. Full understanding of the molecular and physiological function of individual clock proteins and how they work together in a cell type- and tissue-specific manner represents a major gap in the field and should be a key area of future research.

6.2. What Makes the Kidney Clocks Tick?

The regulation of the kidney clocks is an area in need of investigation. This review has speculated potential regulators including multiorgan systems, like the HPA axis and autonomic nervous system, feeding, and body temperature. However, unpacking the mechanisms behind how the kidney clocks tick is challenging because if different cell types in the kidney have different clocks, it is reasonable to assume that the zeitgebers may differ as well. Furthermore, how our lifestyle impacts our kidney clocks, what happens to the kidney clocks under pathological conditions, and how this can impact overall health are key questions that need to be explored. Studies have investigated the implications of shift work on risk of development of pathological conditions such as hypertension and CKD (reviewed in Ref. 23). However, the impact on the kidney clocks has not been explored. Another gap in our knowledge is the impact of lack of gravity on the clocks and circadian rhythms. Interestingly, astronauts experience altered circadian rhythms in heart rate and body temperature during spaceflight due to changes in gravity load, lighting, and work schedules (211). Spaceflight is also known to alter renal function, as antinatriuretic effects have been observed (212, 213). One small study of two astronauts indicated that the circadian rhythm of BP was dampened during spaceflight (214). With increased spaceflight and the commercialization of spaceflight, understanding the implications of gravity on the circadian clocks, including the kidney clocks, is of increasing relevance.

6.3. Are There Sex and Age Differences in the Kidney Clocks?

To date, only a handful of studies have used both male and female mice in studies of the kidney clock (60, 151, 210). As noted above, PER1 and BMAL1 clearly mediate different effects on BP and on renal Na⁺ handling in male compared with female mice. Sex differences in renal Na⁺ excretion was also seen in global Bmal1 KO rats, the first known clock gene KO rat model (152). We also discussed above that glucocorticoid rhythm is influenced by the estrus cycle (197). This could apply to other hormones and be altered in aging, especially after menopause, which could have influences on the clock. Martino and colleagues (215) studied female Clock delta 19 mutant mice and showed that they were protected from age-related cardiomyopathy and had lower MAP compared with control mice when the experimental mice were 21 mo old. To our knowledge, this is the only study conducted in a clock gene transgenic mouse model where both age and sex were considered experimental variables. Although BP was not measured, this group also performed ovariectomy in the female Clock delta 19 mice and showed that loss of ovarian hormones was associated with cardiac remodeling and signs of heart disease in 12-mo-old mice. This study as well as studies highlighted throughout this review demonstrate the importance of conducting circadian physiology studies in both male and female animals, so that these important findings can be translated to men and women.

As part of the American Heart Association Go Red for Women Strategically Focused Research Network, Makarem et al. (216) conducted a compelling study in adult (average age ∼33 yr), premenopausal women to assess the relationship between circadian disruption and cardiometabolic risk. The findings revealed that greater day-to-day variability in eating patterns, eating jet lag (weekday/weekend scheduling of meals), and eating late at night were associated with increase in BP, adiposity, and worse glycemic control. Variable or erratic eating patterns are a form of circadian disruption, and this recent study is striking in that it demonstrates a connection between circadian disruption and cardiometabolic risk in women. This further highlights the importance of studying female animals in preclinical studies aimed at better understanding the integration of circadian biology and renal physiology.

7. TIME OF DAY IS A KEY BIOLOGICAL VARIABLE

Over 100 years of documented evidence from human studies and decades of animal studies, together with more recent targeted molecular studies, unequivocally demonstrate the importance of circadian rhythms in kidney function. Despite this, time of day continues to be largely ignored as a key biological variable in the study of renal physiology. This is particularly important for translating rodent studies to humans, given that rodents are nocturnal whereas humans are naturally diurnal (FIGURE 9). Including time of day as a key biological variable can improve the rigor and reproducibility of renal physiology studies (217).

FIGURE 9.

Time of day is a key biological variable. Humans are diurnal, with activity patterns and cardiovascular functions peaking during the daytime, whereas mice and rats are nocturnal, with activity patterns and cardiovascular functions peaking during the nighttime. Circadian rhythms in physiological function are very similar between rodents and humans, except that the peaks and troughs are antiphase to each other. For rodent studies to be as physiologically relevant to humans as possible, time of day should be considered a key biological variable. Diagram created with BioRender.com, with permission.

One example of the benefit of reporting the time of day comes from the work of Thomas et al. (218), where they showed that treatment of mice with a small-molecule NPT2a inhibitor led to phosphaturia, but only during the first 3 h after treatment, which was given at 9:00 AM. Pairing these results with those of Miyagawa et al. (123) suggests that the effects of the NPT2a inhibitor may be limited to the early morning because this is when peak NPT2a protein expression occurs in the mouse. Since time of day is reported so rarely, it is impossible to know if there are other results in the literature that could be explained or reconciled with conflicting findings simply by considering the effect of circadian rhythms on physiological function. This concept also applies to findings that may not be in the literature because apparently negative results were never published. There may be unreported cases in which a gene KO or other experimental maneuver in a rat or mouse did not yield the expected phenotype because studies were limited to “normal working hours” during the rodent rest phase, which coincides with the nadir of many physiological functions. Just like sex and age, time of day is a key biological variable and should be reported in order to maximize rigor and reproducibility.

8. CONCLUSIONS AND FUTURE DIRECTIONS

Overall, it is imperative to better understand the mechanisms behind how the kidney clock ticks, its cross talk with other circadian clocks, and its implications for integrative physiology and pathophysiology. However, this remains challenging, with the heterogeneity of the kidney cell types and the lack of an appropriate Cre recombinase to cover the entire kidney with limited off-target effects.

A plethora of questions remain regarding both the regulation and the function of the kidney clock (see TABLE 2). Very little is known about why or how clock proteins mediate different effects on renal function in males compared with females, and there appears to be nothing known about how the kidney clock ages. Studies of the kidney clock have mostly been limited to normal, steady-state conditions in males, only in relation to dietary salt manipulation in both sexes, or in a limited number of CKD or renal injury models in males only, as discussed above. Clinical and epidemiological data demonstrate a critical role for normal circadian rhythms in renal function with maintaining cardiovascular health. In the last several years, increased interest in the kidney clocks has led to preclinical animal research that has shed significant light on the importance of the circadian clock in the regulation of BP and renal function, yet these studies have raised more questions than they have answered. Furthermore, whether some circadian functions operate independently of the current clock paradigm remains unknown. Detailed studies of the kidney clocks in male and female animals, across the life span, and in a variety of renal pathophysiology models are absolutely necessary to lay the groundwork for translating these findings to humans in order to harness the potential of circadian medicine in the treatment of kidney disease and hypertension.

Table 2.

Remaining questions

Remaining Questions
How does the mechanism of the clock differ between males and females and among different cell types in the kidney?
Are there kidney-specific zeitgebers?
Do the kidney clocks communicate with each other and how?
What happens to the kidney clocks with aging?

GRANTS

The authors are supported by American Heart Association Grants 19EIA34660135 (M.L.G.), 906349 (H.M.C.), and 872573 (J.G.J.); and National Institute of Diabetes and Digestive and Kidney Diseases Grants R01 DK-109570 (M.L.G.), R01 DK-123078 (M.L.G.), F32 DK-121424 (G.R.C.), and T32 DK-104721 (J.G.J.).

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

H.M.C. and M.L.G. analyzed data; H.M.C. and A.J. interpreted results of experiments; H.M.C., J.G.J., A.J., G.R.C., and M.L.G. prepared figures; H.M.C., J.G.J., A.J., G.R.C., and M.L.G. drafted manuscript; H.M.C., J.G.J., A.J., G.R.C., and M.L.G. edited and revised manuscript; H.M.C., A.J., and M.L.G. approved final version of manuscript.