High dietary sodium causes dyssynchrony of the renal molecular clock in rats

Abstract

Speed JS, Hyndman KA, Roth K, Heimlich JB, Kasztan M, Fox BM, Johnston JG, Becker BK, Jin C, Gamble KL, Young ME, Pollock JS, Pollock DM. High dietary sodium causes dyssynchrony of the renal molecular clock in rats. Am J Physiol Renal Physiol 314: F89–F98, 2018. First published September 27, 2017; doi:10.1152/ajprenal.00028.2017.—Dyssynchrony of circadian rhythms is associated with various disorders, including cardiovascular and metabolic diseases. The cell autonomous molecular clock maintains circadian control; however, environmental factors that may cause circadian dyssynchrony either within or between organ systems are poorly understood. Our laboratory recently reported that the endothelin (ET-1) B (ET_B) receptor functions to facilitate Na⁺ excretion in a time of day-dependent manner. Therefore, the present study was designed to determine whether high salt (HS) intake leads to circadian dyssynchrony within the kidney and whether the renal endothelin system contributes to control of the renal molecular clock. We observed that HS feeding led to region-specific alterations in circadian clock components within the kidney. For instance, HS caused a significant 5.5-h phase delay in the peak expression of Bmal1 and suppressed Cry1 and Per2 expression in the renal inner medulla, but not the renal cortex, of control rats. The phase delay in Bmal1 expression appears to be mediated by ET-1 because this phenomenon was not observed in the ET_B-deficient rat. In cultured inner medullary collecting duct cells, ET-1 suppressed Bmal1 mRNA expression. Furthermore, Bmal1 knockdown in these cells reduced epithelial Na⁺ channel expression. These data reveal that HS feeding leads to intrarenal circadian dyssynchrony mediated, in part, through activation of ET_B receptors within the renal inner medulla.

INTRODUCTION

High salt (HS) intake is associated with hypertension and cardiovascular disease (CVD), and a recent global study attributed one out of five premature deaths caused by CVD to HS intake (19). The kidney is the organ ultimately responsible for maintaining Na⁺ balance, and, in chronic, steady-state situations, renal excretion of Na⁺ follows a diurnal rhythm that is independent of the timing of Na⁺ intake (18). Studies from the past 15–20 yr have clearly demonstrated that HS intake stimulates renal endothelin 1 (ET-1) production and activation of the ET_B receptor to inhibit Na⁺ reabsorption, especially in the inner medullary collecting duct, to promote natriuresis. In rats lacking functional ET_B receptors, we recently demonstrated a delayed natriuresis following an acute NaCl load (15). However, the degree of delay was dependent on the time of day when salt was given, suggesting a relationship between ET_B-dependent natriuresis and diurnal control of Na⁺ excretion.

Intrinsic, cell autonomous molecular clocks consisting of a system of oscillating transcription factors mediate physiological rhythms (25). Molecular clocks are controlled by a number of negative and positive feedback loops (25). The suprachiasmatic nucleus of the hypothalamus contains the central clock, which is controlled by external light cues transmitted from the retina. The central clock is thought to synchronize the timing of the local or peripheral clocks; however, peripheral clocks are controlled by a variety of other zeitgeber (cues) independent of light, such as food intake or activity. Core components of the cell autonomous clocks in mammals are two transcription factors, aryl hydrocarbon receptor nuclear translocator-like protein (Arntl1, also called Bmal1) and circadian locomotor output cycles kaput (CLOCK). These proteins heterodimerize and drive the transcription of Period (Per) and Cryptochrome (Cry), which subsequently feed back to inhibit their own transcription by interfering with the binding of Bmal1 and CLOCK to E-box response elements located in the promoter region of each gene (21). Local or peripheral clocks exist in every cell type and help regulate tissue-specific function by affecting transcription of output genes. Within the kidney, clock genes modulate renal excretory function by controlling the expression and/or activity of Na⁺ channels and transporters (6, 7, 24). For example, Bmal1 knockout mice have a complete loss in diurnal variability in Na⁺ excretion (37). This is in part due to loss of Bmal1 function in the kidney because knockout of Bmal1 from renin-secreting cells results in a loss of a diurnal pattern of Na⁺ excretion (32). Taken together, these data indicate the importance of renal Bmal1 in the regulation of renal excretory function. However, it is unclear how external cues, such as HS diet, may regulate Bmal1 expression and renal salt handling.

Little is known about the relationship between salt intake and the renal molecular clock. Studies from the Gumz laboratory have demonstrated that Per1 facilitates the ability of aldosterone to increase Na⁺ reabsorption (7). Two recent studies from the Firsov group demonstrate a role for Bmal1 in regulating renal tubular function (20, 32). Overall, these findings suggest that activity of the renal molecular clock genes serves primarily to conserve Na⁺. However, neither of these groups has explored the role of the clock in response to an HS diet. Therefore, the present study was designed to test the hypothesis that HS intake impairs Bmal1 function through an ET-1-dependent mechanism to facilitate Na⁺ excretion and maintain long-term balance. Experiments utilized in vivo studies with a novel ET_B receptor-deficient and transgenic control rat model on normal salt (NS) and HS diets as well as in vitro studies with cultured inner medullary collecting duct cells.

METHODS

The Institutional Animal Care and Use Committee at the University of Alabama at Birmingham approved all protocols in accordance with the NIH Guide for the Care and Use of Laboratory Animals. Male ET_B-deficient rats (ET_B def) or transgenic control (control) littermates 12–14 wk of age were obtained from our in-house colony (5). The ET_B def strain was developed from a rat that has a naturally occurring mutation within the ET_B receptor gene rendering it nonfunctional (5). Without the receptor, the enteric nervous system does not develop, leading to megacolon; therefore, the strain was “rescued” by inserting a functional receptor transgene into sympathetic tissues under the control of the dopamine beta hydroxylase promoter. The control littermates have functional ET_B receptors expressed throughout the body and also express the transgene. Animals were fed a normal (0.2% Na⁺, TD.96208), or high (1.6% Na⁺, TD.92034) Na⁺ diet (Envigo, Indianapolis, IN) with free access to water. Rats were housed in temperature- and humidity-controlled, 12-h:12-h light/dark cycles.

Circadian clock gene study.

Control or ET_B def rats were maintained on NS or placed on HS diet for 14 days. Rats were euthanized in 4-h intervals starting at ZT0, which is when the lights come on at 7:00 AM. Blood was collected in heparinized tubes, and plasma was immediately separated by centrifugation at 1,000 g for 10 min and snap frozen. Kidneys were removed, and cortex and inner medulla were then dissected and snap frozen in liquid nitrogen. Samples were stored at −80°C until analyses were performed. Plasma aldosterone was measured by EIA (Cayman Chemical, Ann Arbor, MI). Plasma ET-1 was measured by ELISA (QuantiGlo; R&D Systems, Minneapolis, MN).

Urinary ET-1 excretion.

Separate groups of control and ET_B def rats were placed on an NS or HS diet for 14 days, before being placed in metabolic cages for urine collection. Animals were acclimated to the cages for 2 days before the urine collection. Collections and measurements were made from time periods ZT0-ZT12 (lights on, inactive phase) and ZT12-ZT24 (lights off, active phase). Urinary ET-1 concentration was measured by ELISA (QuantiGlo; R&D Systems). ET-1 concentration was multiplied by urine volume to determine excretion rate.

Quantitative real-time PCR.

A sample (50 mg) of cortical tissue or one renal inner medulla was homogenized in Trizol reagent, and RNA was extracted according to manufacturer’s instructions (ThermoFisher, Waltham, MA). Reverse transcription was carried out with 1 μg of mRNA using Bio-Rad cDNA synthesis kit (no. 1708891; Hercules, CA). Quantitative real-time PCR was performed on Bio-Rad CFX96 real-time apparatus using iTaq Universal Probes Supermix (no. 1725131; Bio-Rad). TaqMan gene expression assays (ThermoFisher) or custom primer and probe sets were purchased from IDT (Coralville, IA). All primers and TaqMan gene assays are listed in Table 1. For rat clock gene expression, standard curves were generated using a known concentration of the specific amplicon, and copy number/ng RNA is reported (36). For αENaC (Scnn1a) and Klf15 in which standards were not available, 2^-ΔΔcT method was used to calculate relative mRNA expression from the global mean cycle threshold of the specific gene of interest.

Table 1.

Primer and probe sequences for real-time PCR

	Forward Primer (5′→3′)	Reverse Primer (5′→3′)	Probe (56-Fam-5′→3′-TAMTSp)
Bmal1	TCCGATGACGAACTGAAACAC	CTCGGTCACATCCTACGACAA	CAAAAATCCATCTGCTGCCCTGAGAAT
Cry1	CATCAACAGGTGGCGATTTT	CCCGAATCACAAACAGACGA	TTTAATTTTCGTAGATTGGCATCAAGATCCTCAAGA
Cry2	GTGTGAATGCAGGCAGCTG	ACAGGGCAGTAGCAGTGGAA	ATGTGGCTGTCCTGCAGTGCTTTCTTC
Per1	GGTTCAGGATCCCACGAAG	AAGAGTCGATGCTGCCAAAG	AGCACCTCAGCCAGCATCACCC
Per2	GCAGCCTTTCGATTATTCTTC	GCTCCACGGGTTGATGAAG	ATTCGATTCCGCACACGCAACG

	ThermoFisher TaqMan Assay Number
ET-1	Rn00561129_m1
αENaC	Rn00580652_m1
KLF15	Rn00585508_m1

mIMCD-3 experiments.

For the in vitro ET-1 studies, mIMCD-3 (ATCC, Manassas, VA) were grown to confluency on 12-well plates in Dulbecco’s Modified Eagle Medium nutrient mixture F-12 (DMEM/F-12) with 10% fetal bovine serum and 1% penicillin/streptomycin. Only passages 3–6 were used in this study. At confluency, cells were exposed to DMEM/F-12 with 25% FBS for 1 h. Immediately following, media was changed to DMEM/F-12 with 1% FBS and 50 μM phosphoramidon to inhibit endogenous ET-1. Cells were treated with either vehicle or 100 nM ET-1 for 4 h. Cells were lysed using buffer from RNAqueous microRNA kit (ThermoFisher Scientific), and RNA was isolated according to manufacturer’s protocol. Reverse transcription was carried out with cDNA synthesis kit (Bio-Rad).

Mouse Bmal1 knockdown experiments were performed with commercially available small interfering RNA (siRNA, sc-38166; Santa Cruz Biotechnology, Dallas, TX) and control scramble siRNA (sc-37007). mIMCD-3 cells were transfected using the reverse siRNA protocol of Invitrogen (Carlsbad, CA) with Lipofectamine RNAiMax in opti-MEM reduced serum medium. Twelve-well plates for confirmation of knockdown or mRNA analyses were transfected using 10 and 60 pmol of siRNA and 1 μl of RNAiMax. At 100% confluence, cells were collected for Western blot and mRNA analyses. All cell culture experiments were replicated on three different days with three different passages.

Western blot analysis.

Western blots of mIMCD-3 cells from the siRNA experiments were performed to confirm Bmal1 knockdown. Nuclear fractions of control and Bmal1 siRNA-treated cells were extracted using the Cayman Chemical nuclear extraction kit. Ten micrograms of nuclear extraction were separated by 8% SDS-PAGE and transferred to PVDF membranes. Bmal1 was detected by a goat anti-human Bmal1 (0.2 μg/ml, sc-8550; Santa Cruz Biotechnology) with donkey anti-goat IgG-Alexa Fluor 680 (ThermoFisher) by the Odyssey CLx (Li-Cor, Lincoln, NE).

Statistics and data analysis.

Circadian data were analyzed by cosinor analysis using Graph Pad Prism with the assumption of a 24-h period. With cosinor analysis, mesor represents the midline of a rhythmic curve, amplitude represents the difference between the peak of the rhythm and the mesor, and acrophase represents the time of day (zeitgeber time) in which the rhythm peaks. These values were compared between diets by extra sum of squares F test. A separate cosinor analysis was performed to compare genotypes of the same diet. Urinary excretion data were compared by two-way ANOVA followed by Tukey’s post hoc test. In vitro data were compared by Student’s t-test. Statistical significance was set at P < 0.05.

RESULTS

Renal and extrarenal ET-1 follows a diurnal rhythm.

We determined whether renal ET-1 production varies by the time of day. Urinary excretion of ET-1 is often used as an index of renal ET-1 production, so we measured ET-1 excretion during the 12-h light (inactive) and the 12-h dark (active) periods. In control rats, urinary excretion of ET-1 was significantly greater during the active period vs. the inactive period in both NS- and HS-fed rats (Fig. 1A, p_interaction = ns, p_time < 0.05, p_diet = ns). In contrast, ET_B def rats on NS exhibited a loss of the day-night difference in renal ET-1 production. However, a diurnal pattern of ET-1 excretion was observed with HS feeding in ET_B def rats similar to controls (Fig. 1B, p_interaction = ns, p_time < 0.05, p_diet = ns). We also determined renal inner medullary Edn1 mRNA levels in 4-h time periods. Renal inner medullary Edn1 mRNA was lower during the inactive period compared with the active period in control rats on NS (Fig. 1C). HS intake had no significant effect on Edn1 mRNA in either genotype (Fig. 1, C and D).

Fig. 1.

Renal and extrarenal endothelin-1 (ET-1) production follows a circadian rhythm. A and B: 12-h urinary ET-1 excretion (pg/12 h) in control or ET_B-deficient (ET_B def) rats fed either normal salt (NS n = 5 and 4, respectively) or high salt (HS, n = 6 and 4, respectively) diet for 2 wk. C and D: circadian expression pattern (relative to average global Ct) of ET-1 by the renal inner medulla in control and ET_B def rats fed either NS or HS diet for 2 wk (n = 5). E and F: plasma ET-1 (pg/ml) in control and ET_B def rats in 4-h intervals (n = 6). Symbols represent means ± SE, and lines represent cosinor analysis. *P < 0.05 vs. NS inactive; †P < 0.05 vs. HS inactive. See Table 2 for data from cosinor analysis.

We also measured plasma ET-1, which is highly sensitive to ET_B receptor expression because of its ability to clear ET-1 from the circulation (22). Similar to previous publications, we found an ~10-fold higher concentration of ET-1 in the plasma of ET_B def rats vs. control (11). Plasma ET-1 levels were analyzed by cosinor analysis to determine the mesor (midline of cosine curve), amplitude (peak minus trough), and acrophase (time of peak) (Table 2). In control rats, circulating ET-1 followed a circadian rhythm in which the amplitude was suppressed by HS intake (Fig. 1E and Table 2). The amplitude and mesor in plasma ET-1 were higher in ET_B def rats compared with control, yet HS did not suppress the circadian rhythm of circulating ET-1 (Fig. 1F and Table 2).

Table 2.

Cosinor analyses of plasma parameters from Figs. 1 and 5

	Normal Salt			High Salt
	Mesor	Amplitude	Acrophase	Mesor	Amplitude	Acrophase
Control
Plasma ET-1	0.66 ± 0.02	0.10 ± 0.03	9.70 ± 0.30	0.67 ± 0.02	0.04 ± 0.03*	10.60 ± 0.70
Plasma Aldo	966.00 ± 80.00	594.00 ± 107.00	13.30 ± 0.20	618.00 ± 51.00*	247.00 ± 71.00*	16.10 ± 0.30
ETB Def
Plasma ET-1	6.80 ± 0.10†	0.70 ± 0.20	9.90 ± 0.30	6.70 ± 0.20†	1.00 ± 0.02	10.20 ± 0.20
Plasma Aldo	1,174.00 ± 99.00	565.00 ± 138.00	14.30 ± 0.30	701.00 ± 60.00*	328.00 ± 93.00	17.70 ± 0.20*

Data are presented as means ± SE. Units are pg/ml for endothelin-1 (ET-1) and aldosterone (Aldo). Units for acrophase are zeitgeber time. ET_B def, ET_B deficient.

^*P < 0.05 vs. NS.

^†P < 0.05 vs. control.

HS intake desynchronizes renal clocks.

To examine the potential impact of high Na⁺ intake on circadian rhythms of the renal molecular clock, we determined mRNA expression of the major clock genes in the renal inner medulla and cortex in response to HS intake over the course of 24 h. Male rats were euthanized in 4-h intervals, and kidneys were dissected to measure clock gene expression. Molecular clock gene expression patterns were analyzed by cosinor analysis (Table 3). In response to 2 wk of HS intake, control rats had a significant 5.5-h phase delay in Bmal1 expression in the inner medulla (Fig. 2A) but not in the renal cortex (Fig. 2C). In contrast, ET_B def rats had similar Bmal1 expression when fed NS or HS diets (Fig. 2, B and D). There were no significant differences in the expression of CLOCK in response to HS in either genotype (Fig. 2, E–I).

Table 3.

Cosinor analyses of renal inner medullary gene expression Figs. 1–5

	Normal Salt			High Salt
	Mesor	Amplitude	Acrophase	Mesor	Amplitude	Acrophase
Control
Bmal1	275.00 ± 12.00	232.00 ± 30.00	23.70 ± 0.10	302.00 ± 16.00	308.00 ± 32.00	5.30 ± 0.10*
Clock	10,113.00 ± 331.00	453.00 ± 493.00	17.00 ± 1.00	11,097.00 ± 397.00	1,402.00 ± 588.00	6.90 ± 0.90
Cry1	226.00 ± 19.00	268.00 ± 56.00	19.60 ± 0.20	157.00 ± 10.00*	130.00 ± 28.00*	20.30 ± 0.20
Cry2	1,490.00 ± 56.00	872.00 ± 152.00	13.50 ± 0.20	1,440.00 ± 70.00	666.00 ± 198.00	14.40 ± 0.30
Per1	197.50 ± 18.00	202.00 ± 50.00	14.80 ± 0.30	204.00 ± 18.00	214.00 ± 48.00	13.40 ± 0.20
Per2	230.00 ± 11.00	304.00 ± 32.00	14.80 ± 0.10	176.00 ± 8.00*	206.00 ± 24.00*	15.10 ± 0.10
Klf15	1.20 ± 0.08	0.59 ± 0.12	15.70 ± 0.20	0.77 ± 0.06	0.29 ± 0.08	13.80 ± 0.30
Edn1	1.06 ± 0.06	0.20 ± 0.08	3.30 ± 0.40	1.10 ± 0.07	0.20 ± 0.09	20.30 ± 0.50
αENaC	1.38 ± 0.12	0.63 ± 0.18	16.80 ± 0.30	0.80 ± 0.07*	0.28 ± 0.10*	16.20 ± 0.30
ETB Def
Bmal1	295.00 ± 17.00	496.00 ± 42.00	23.90 ± 0.10	315.00 ± 15.00	354.00 ± 40.00	23.60 ± 0.10
Clock	10,411.00 ± 379.00	1,090.00 ± 553.00	19.90 ± 0.50	11,418.00 ± 379.00	1,661.00 ± 491.00	0.10 ± 0.40
Cry1	236.00 ± 18.00	126.00 ± 36.00	20.50 ± 0.20	170.00 ± 12.00*	126.00 ± 36.00*	20.20 ± 0.20
Cry2	1,515.00 ± 70.00	692.00 ± 192.00	14.70 ± 0.30	1,482.00 ± 51.00	526.00 ± 138.00	14.70 ± 0.30
Per1	221.00 ± 20.00	250.00 ± 66.00	15.00 ± 0.20	197.00 ± 18.00	252.00 ± 50.00	14.90 ± 0.20
Per2	253.00 ± 9.00	278.00 ± 24.00	14.90 ± 0.10	190.00 ± 7.00*	220.00 ± 20.00	14.80 ± 0.10
Klf15	1.77 ± 0.13	1.20 ± 0.18	14.40 ± 0.20	0.99 ± 0.06	0.21 ± 0.08	14.10 ± 0.40
Edn1	0.87 ± 0.04	0.30 ± 0.05	21.00 ± 0.20	1.00 ± 0.07	0.30 ± 0.09	23.50 ± 0.40
αENaC	1.96 ± 0.18†	1.23 ± 0.26	16.40 ± 0.20	1.14 ± 0.11*†	0.43 ± 0.17*	17.20 ± 0.30

Data are presented as mean ± SE; n = 5 or 6. Units for mesor and amplitude are copy number/ng RNA for Bmal1, Clock, Cry1, Cry2, Per1, and Per2. Relative expression for Klf15, Edn1, and αENaC is shown. Units for acrophase: zeitgeber time. ET_B def, ET_B deficient.

^*P < 0.05 vs. NS.

^†P < 0.05 vs. control.

Fig. 2.

High salt intake desynchronizes kidney clocks via ET_B receptors. mRNA expression (copy count) in 4-h intervals of renal medullary Bmal1 (A and B, n = 5–6), renal cortical Bmal1 (C and D, n = 5), renal inner medullary Clock (E and F, n = 5–6), and renal cortical Clock (G and H, n = 5) of control and ET_B-deficient (ET_B def) rats that have been maintained on normal or high salt diet for 2 wk is shown. Symbols represent means ± SE, and lines represent cosinor analysis. *P < 0.05 for acrophase. See Table 2 for data from cosinor analysis.

HS intake also altered expression of other key clock genes in both ET_B def and control rats. In both control and ET_B def rats fed HS, inner medullary Cry1 (Fig. 3, A and B, respectively) and Per2 (Fig. 3, G and H, respectively), expression was significantly suppressed at ZT16 compared with NS-fed rats (Table 3 for cosinor analysis). Expression of these clock genes was similar in ET_B def rats compared with controls whether fed NS or HS. No difference was found in Cry2 (Fig. 3, C and D) or Per1 (Fig. 3, E and F) expression in either genotype or in response to HS intake. Kruppel-like Factor 15 (Klf15), thought to be under the control of Bmal1 (12), was also suppressed in the inner medulla by HS in both genotypes (Fig. 3, I and J). Expression of all clock genes was similar in the renal cortex between genotype and salt treatment (Fig. 4). Collectively, these data indicate that clock genes are regulated by HS intake specifically in the renal medulla and reveal a clear dyssynchrony of the molecular clock between different regions of the kidney. These results further suggest a role for the ET_B receptor in modulating Bmal1 function during conditions of HS intake.

Fig. 3.

High salt (HS) intake dissociates molecular clock components independent of ET_B receptor activation. mRNA expression (copy count, n = 5–6) in 4-h intervals of Cry1 (A and B), Cry2 (C and D), Per1 (E and F), Per2 (G and H), and Klf15 (I and J) in the renal inner medulla of control and ET_B-deficient (ET_B def) rats maintained on normal salt or HS diet for 2 wk is shown. Symbols represent means ± SE, and lines represent cosinor analysis. †P < 0.05 for amplitude. See Table 2 for data from cosinor analysis.

Fig. 4.

High salt (HS) intake has no significant effect on the molecular clock components of the renal cortex. mRNA expression (copy count, n = 5) in 4-h intervals of Cry1 (A and B), Cry2 (C and D), Per1 (E and F), and Per2 (G and H) in the renal cortex of control and ET_B-deficient (ET_B def) rats maintained on normal salt or HS diet for 2 wk is shown. Symbols represent means ± SE, and lines represent cosinor analysis (Table 2).

HS intake impacts diurnal Na⁺ excretion.

Because chronic HS intake impacts the circadian clock within the kidney, at least in part, via ET_B receptor activation, we examined the physiological impact of ET_B receptor deficiency on diurnal Na⁺ excretion (Table 4). HS diet stimulated a significant increase in Na⁺ excretion during both the inactive and active period compared with NS-fed animals; however, under these steady-state conditions, both genotypes had a similar diurnal rhythm in response to HS. Importantly, food and water intake followed a diurnal pattern that was similar in rats of both strains (Table 4). Furthermore, 24-h excretion as well as food and water intake were similar between genotypes. Diurnal excretion of K⁺ was similar between genotypes during both active and inactive periods on either salt diet (Table 4). These data suggest that alterations in renal clock gene function in response to HS do not have a steady-state impact on urinary excretion patterns.

Table 4.

Diurnal metabolic cage parameters in control and ET_B def rats fed normal or high salt for 2 wk

	Normal Salt (n = 9)		High Salt (n = 5)
Control Rats	Active	Inactive	Active	Inactive
Na+, µEq/12 h	921.0 ± 54.0	625.0 ± 32.0	8,796.0 ± 534.0*	2,956.0 ± 368.0*†
K+, µEq/12 h	1,528.0 ± 87.0	627.0 ± 29.0†	2,187.0 ± 148.0*	530.0 ± 97.0†
UV, ml/12 h	6.9 ± 0.9	4.1 ± 0.3†	25.6 ± 1.0*	9.1 ± 0.9*†
Food intake, g/12 h	17.8 ± 1.1	2.0 ± 0.4†	17.3 ± 0.6	2.3 ± 0.7†
H2O intake, ml/12 h	29.0 ± 3.5	2.3 ± 0.4†	54.4 ± 2.3*	4.7 ± 1.0†

	Normal Salt (n = 8)		High Salt (n = 5)
ETB def rats	Active	Inactive	Active	Inactive
Na+, µEq/12 h	833.0 ± 57.0	720.0 ± 58.0	10,590.0 ± 575.0*	3,395.0 ± 633.0*†
K+, µEq/12 h	1,563.0 ± 74.0	661.0 ± 52.0†	2,602.0 ± 93.0*	680.0 ± 95.0†
UV, ml/12 h	6.8 ± 0.6	4.7 ± 0.4†	32.3 ± 1.9*	9.3 ± 1.6†
Food intake, g/12 h	19.3 ± 1.7	2.7 ± 0.6†	17.8 ± 0.3	2.2 ± 0.3†
H2O intake, ml/12 h	32.5 ± 2.8	3.1 ± 0.6†	51.0 ± 0.6*	4.6 ± 1.3†

Data are presented as means ± SE. ET_B def, ET_B deficient.

^*P < 0.05 vs. NS.

^†P < 0.05 vs. active.

Suppression of αENaC expression in response to HS is blunted in ET_B-deficient rats.

Activation of ET_B receptors is known to promote Na⁺ excretion by inhibiting epithelial sodium channel (ENaC) expression and activity (2). αENaC expression also appears to be regulated by the molecular clock, specifically Per1, in the collecting duct of the kidney (30). In contrast to what has been shown in response to a low-salt diet, we hypothesized that suppression of αENaC expression during HS intake is at least in part mediated by suppression of core clock genes. In control rats, renal inner medullary αENaC mRNA expression displayed a circadian rhythm as determined by cosinor analysis and, as expected, was suppressed by HS intake (Fig. 5A; Table 3). On the other hand, ET_B def rats on NS also displayed a rhythm in αENaC mRNA expression but with a larger mesor than control rats (Fig. 5A and Table 3). αENaC was also suppressed in ET_B def rats in response to HS intake, but expression was higher than control rats (Table 3). These data suggest that renal αENaC gene expression may be under the control of the molecular clock in vivo, and ET_B receptor activation may serve to suppress basal αENaC expression. The differences in αENaC between ET_B def rats and control do not appear to be from differences in circulating aldosterone, as plasma aldosterone concentration was suppressed by HS and similar between the two genotypes (Fig. 5, C and D, Table 2).

Fig. 5.

High salt (HS) intake suppression of α-epithelial sodium channel (αENaC) is partially mediated by ET_B receptor activation. A and B represent rhythmic expression of renal inner medullary αENaC expression (n = 4–5), and C and D represent plasma aldosterone (n = 4–6) in control and ET_B-deficient (ET_B def) rats fed either normal salt or HS for 2 wk. Symbols represent means ± SE, and lines represent cosinor analysis. See Table 2 for cosinor analysis and statistics.

Elucidating the ET-1/Bmal1/ENaC axis in vitro.

A phase shift in Bmal1 expression was observed in response to HS in control rats but not ET_B def rats, suggesting a mechanism whereby HS intake causes a phase shift in the Bmal1 circadian rhythm via ET-1. To determine whether changes in ET-1 impact Bmal1 directly, we exposed mouse inner medullary collecting duct-3 (mIMCD-3) cells to 100 nM ET-1, a concentration that stimulates nitric oxide release by these cells (10) and reduces ENaC open probability in split-open mouse cortical collecting duct tubules (8). After 4 h of treatment, there was a significant (~26%) decrease in Bmal1 mRNA expression compared with vehicle-treated cells (Fig. 6A). To determine whether a reduction in Bmal1 can suppress αENaC expression, mIMCD-3 cells were treated with Bmal1 or scrambled siRNA. Bmal1 siRNA (60 pmol) led to an ~60% reduction in Bmal1 protein expression (Fig. 6B, inset). Knockdown of Bmal1 significantly reduced αENaC mRNA (Fig. 6B). These data support the hypothesis that activation of the ET-1/ET_B pathway results in reduced αENaC expression in a Bmal1-dependent manner.

Fig. 6.

Bmal expression in mIMCD-3 cells treated with 100 nM endothelin-1 (ET-1) (A). α-Epithelial sodium channel (αENaC) expression in mIMCD cells treated with 60 pmol of siRNA to Bmal1 is shown (B). Inset: representative Western blot to confirm nuclear Bmal1 knockdown with siRNA. *P < 0.05, n = 3 separate experiments conducted in duplicate.

DISCUSSION

Disruption of normal circadian patterns, whether genetic or environmental, can increase the risk of many disorders, especially within the cardiovascular system (35). In addition, it is well documented that elevated Na⁺ intake contributes to overall mortality due to CVD (19). Our findings implicate several major processes that occur in the kidney in response to HS intake related to the renal circadian clock. First, HS causes a significant phase shift or suppression of clock gene expression in the renal inner medulla of control rats that are resistant to salt-induced hypertension. In addition, HS intake causes dyssynchronization of clock gene rhythms in a region-specific manner within the kidney. Finally, our studies also demonstrate that the ET-1/ET_B receptor pathway contributes to the phase shift in Bmal1 produced by an HS diet, and loss of this phase shift may contribute to salt-dependent effects previously observed in the ET_B def model.

Renal ET-1 regulates blood pressure by activating the ET_B receptors and reducing Na⁺ reabsorption largely within the collecting duct (16, 28, 29). ET-1 inhibits ENaC activity in cortical collecting ducts, which most likely accounts for the natriuretic activity of the peptide. Our data also suggest that ET-1 plays a role in diurnal variation in Na⁺ excretion by modulating clock gene expression (2). It is interesting to speculate that ET-1 could serve as an immediate regulator of ENaC activity in the short term but also as a “sensor” of sodium delivery via feedback to Bmal1 expression and thus modulate a variety of circadian-controlled genes within the inner medulla.

The present findings indicate a diurnal rhythm in the renal production of ET-1 as measured in the urine that is associated with diurnal rhythms in Na⁺ excretion. Furthermore, our laboratory has shown that HS intake increases ET-1 receptor binding and shifts ET-1 binding in the inner medulla from 60/40% ET_B/ET_A to 100% ET_B (9, 13). Finally, Richards et al. (26) reported higher levels of ET_B receptor expression at midnight, or the middle of the active phase, in the inner medulla of mice with a concurrent decrease in ET_A receptor expression compared with the middle of the inactive phase. Taken together, these data suggest that activation of ET_B receptors in response to HS intake promotes a shift in Bmal1 expression in control rats and support our hypothesis that renal ET-1 regulates renal clock genes that may be responsible for diurnal patterns in renal Na⁺ handling.

Within the renal cortex and inner medulla, we observed oscillations in core clock genes that are similar to those reported by other investigators (34), and changes in gene expression are highly correlated to protein expression in vitro (3) and in vivo (27). One of the major findings of the present study is that several clock genes within the inner medulla of the kidney are suppressed or phase delayed in response to HS feeding, but this was not observed in the renal cortex of these same animals. The fact that clock genes within a single organ become desynchronized in terms of the molecular clock is contrary to the dogma that peripheral clock genes function as one synchronous unit. At least part of the mechanism leading to this discrepancy is most likely due to the fact that expression of ET-1 and ET_B receptors is ~10-fold higher in the inner medullary collecting duct compared with the renal cortex (13, 29, 31). Given that disruption of the molecular clock leads to elevated CVD risk, we speculate that risk associated with HS intake may be partially due to dyssynchrony between peripheral clocks of the kidney (17).

The canonical paradigm is that zeitgebers, or cues, initiate the transcription of clock proteins Bmal1 and CLOCK. These proteins heterodimerize and promote the transcription of Cry and Per, which feed back in a negative loop to inhibit Bmal1 and CLOCK. Although this circuit is well documented in the central clock, these new data challenge the pattern of this regulatory scheme between the peripheral clocks within the kidney. In fact, HS affected several clock genes within the inner medulla, whereas Per1 oscillated similarly to NS-fed animals. Gumz et al. (6, 30) reported that aldosterone signals through Per1 in the collecting duct to increase expression of ENaC. A spike in aldosterone expression at zeitgeber time 12 (ZT12, or 7:00 PM/lights off) is observed in NS-fed animals and, to a lesser extent, in HS-fed animals. Therefore, it is reasonable to speculate that aldosterone may be maintaining circadian rhythm of Per1 expression in the collecting duct although other clock factors are suppressed. In addition to hormonal differences that may be regulating the molecular clock, it is also necessary to point out that a majority of what we know about clock gene regulation and function is derived from knockout mouse models. Interestingly, in the mouse, physiological rhythms continue when any single clock gene is knocked out, except for Bmal1 (23). This is possible because of functional overlap between core clock proteins with other isoforms or other transcription factors. However, Bmal1 knockout mice lose diurnal rhythm in both blood pressure (4) and Na⁺ excretion (D. M. Pollock, unpublished observations), suggesting that it is the master circadian clock protein. Our study, along with unpublished observations from our laboratory suggests that this may not be the case in peripheral tissues of the rat; however, more studies in knockout rats will be required to address these issues.

The major consequences of Bmal1 and Per1 appear to be antinatriuretic and prohypertensive effects (7), specifically by increasing transcription of Na⁺ transporters and channels in the kidney. Our findings that knockdown of Bmal1 significantly reduces αENaC mRNA in inner medullary collecting duct cells support this concept. This suggests that a reduction in clock gene expression in response to an HS diet would be beneficiary to excrete excess salt. The absence of a phase shift in Bmal1 expression, as seen in the ET_B def rat, is consistent with a contribution of clock genes to an impaired ability to efficiently excrete a salt load (15) and thus may contribute to salt-sensitive hypertension observed in this model. To date, the nature of clock gene interactions within the renal tubular system has not been established. Two experimental models reported by Firsov’s group have begun to delineate the role of the renal molecular clock in control of renal excretory function (20, 32). First, it was observed that knockout of Bmal1 from renin-secreting cells, including juxtaglomerular cells and collecting duct, reduced blood pressure and altered circadian patterns in Na⁺ excretion (32). Second, knockout of Bmal1 only in the nephron reduced systolic blood pressure with no effect on the circadian pattern of blood pressure or Na⁺ excretion in NS-fed animals (20). To understand how tubular Bmal1 affects Na⁺ handling, more studies involving HS intake or an acute Na⁺ challenge are necessary.

Klf15, a transcription factor that has been shown to be under the control of the molecular clock, specifically in the heart, is abundantly expressed in the kidney (1). In the present study, we observed that Klf15 expression is significantly reduced by HS intake. Work from Jain et al. (38) indicates that, within the heart, Klf15 is responsible for the oscillation of over 1,300 genes while also suppressing the oscillation of 473 genes. Thus we speculate that HS may suppress downstream targets though Klf15 while also allowing other genes to begin oscillating. Little work has been done to identify the downstream targets of Klf15 within the kidney; however, our study provides a potential link between Klf15 and Na⁺ homeostasis.

The impact of dietary salt intake on clock gene expression could provide insights into clinical complications of HS diets. Perturbations in circadian rhythms such as disturbances in sleep patterns or night shift work are risk factors for CVD. Importantly, in human subjects, clusters of single nucleotide polymorphisms (haplotypes) in the Bmal1 gene were associated with hypertension or type II diabetes (33). These studies highlight the importance of circadian regulation at the molecular level in contributing to cardiovascular function. The present study provides evidence of molecular and circadian mechanisms whereby HS intake may contribute to CVD risk, and this is in part through modulation of renal clock gene expression by the endothelin system.

GRANTS

This work was supported by National Institutes of Health grants to J. Speed (K99 HL127178 and T32DK079337), D. Pollock and J. Pollock (P01 HL69999, P01 HL95499, and HL136267), K. Hyndman (K01 DK105038), B. Fox (F30 DK107194), M. Kasztan (American Society of Nephrology, Joseph A. Carlucci Research Fellowship), and B. Becker (T32HL007457).

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

Conceptualization, J.S.S., K.A.H., M.E.Y, J.S.P., D.M.P.; Methodology, J.S.S., K.A.H., J.S.P., D.M.P.; Validation, J.S.S., K.A.H., K.J.R., J.S.P., D.M.P.; Formal Analysis, J.S.S., K.L.G., M.E.Y.; Investigation, J.S.S., K.A.H., K.J.R., J.B.H., M.K., B.M.F, J.G.J., B.K.B, C.J., D.M.P.; Resources, J.S.S., K.L.G., M.E.Y., J.S.P., D.M.P., Data Curation, J.S.S., D.M.P., Writing-Original Draft, J.S.S., D.M.P., Writing-Review and Editing, J.S.S., K.A.H., K.J.R., J.B.H., M.K., B.M.F., J.G.J., B.K.B, C.J., K.L.G., M.E.Y., J.S.P., D.M.P; Visualization, J.S.S, D.M.P.; Supervision, J.S.S, D.M.P.; Project Administration, J.S.S, D.M.P.; Funding Acquisition; J.S.S., K.A.H., J.S.P., D.M.P.