Cyclooxygenase-1–Deficient Mice Have High Sleep-to-Wake Blood Pressure Ratios and Renal Vasoconstriction
We used cyclooxygenase-1 (COX-1)–deficient mice to test the hypothesis that COX-1 regulates blood pressure (BP) and renal hemodynamics. The awake time (AT) mean arterial pressures (MAPs) measured by telemetry were not different between COX-1+/+ and COX-1−/− (131±2 versus 126±3 mm Hg; NS). However, COX-1−/− had higher sleep time (ST) MAP (93±1 versus 97±2 mm Hg; P<0.05) and sleep-to-awake BP ratio (+8.6%; P<0.05). Under anesthesia with moderate sodium loading, COX-1−/− had higher MAP (109±5 versus 124±4 mm Hg; P<0.05), renal vascular resistance (23.5±1.6 versus 30.7±1.7 mm Hg · mL−1 · min−1 · g−1; P<0.05) and filtration fraction (33.7±2.1 versus 40.2±2.0%; P<0.05). COX-1−/− had a 89% reduction (P<0.0001) in the excretion of TxB2, a 76% reduction (P<0.01) in PGE2, a 40% reduction (P<0.0002) in 6-ketoPGF1α (6keto), a 27% reduction (P<0.02) in 11-βPGF2α (11β), a 35% reduction (P<0.01) in nitrate plus nitrite (NOx), and a 52% increase in metanephrine (P<0.02). The excretion of normetanephrine, a marker for sympathetic nervous activity, was reduced during ST in COX-1+/+ (6.9±0.9 versus 3.2±0.6 g · g−1 creatinine · 10−3; P<0.01). This was blunted in COX-1−/− (5.1±0.9 versus 4.9±0.7 g · g−1 creatinine · 10−3; NS). Urine collection during ST showed lower excretion of 6keto, 11β, NOx, aldosterone, sodium, and potassium than during AT in both COX-1+/+ and COX-1−/−, and there were positive correlations among these parameters (6keto versus NOx; P<0.005; 11β versus NOx; P<0.005; and NOx versus sodium; P<0.005). In conclusion, COX-1 mediates a suppressed sympathetic nervous activity and enhanced NO, which may contribute to renal vasodilatation and a reduced MAP while asleep or under anesthesia. COX-1 contributes to the normal nocturnal BP dipping phenomenon.
Cyclooxygenase (COX) converts arachidonate to prostaglandin (PG) H2 (PGH2),1,2⇓ which can activate thromboxane-PGH2 receptors. However, in vivo, PGH2 is promptly metabolized to several biological active PGs and thromboxane A2 (TxA2). COX produces both vasoconstrictor and vasodilatator metabolites. Therefore, its role in the regulation of BP and hemodynamics is difficult to predict and may explain the positive and negative studies regarding the effect of nonselective COX inhibitors (nonsteroidal anti-inflammatory drugs [NSAIDs]) on blood pressure (BP) in human. However, a metaanalysis showed a 5.0-mm Hg increase in the BP in NSAID users.3
COX-2 is inducible,4–6⇓⇓ but in the kidney it is expressed constitutively in vascular endothelial cells, afferent arterioles, cortical thick ascending limbs, macula densa cells, and glomeruli.7,8⇓ COX-2 regulates renin secretion.7 COX-1 is expressed in the extraglomerular and intraglomerular mesangium, the terminal distal convoluted tubule, collecting duct, and vascular endothelial cells.9,10⇓ Previous studies by Athirakul et al11 have shown that COX-1 gene–deficient (COX-1−/−) mice fail to conserve salt, leading to an exaggerated decline in the mean arterial pressure (MAP) during salt restriction. They also have diminished pressor and renal vasoconstrictor responses during an acute infusion of angiotensin II.12 These studies suggest that COX-1 may be required to maintain BP during states of high angiotensin II. However, less is known concerning the role of COX-1 during states of normal or low angiotensin II. This study was conducted in mice consuming a normal salt diet.
A loss of the normal reduction in BP during sleep (“nondipping” status) increases the risk for cardiovascular disease,13–16⇓⇓⇓ but the mechanism is not clear. The decline in nocturnal BP is associated with a reduced sympathetic nervous system tone.17 Loss of the nocturnal decline in BP is accompanied by reduced physical activity,18 increased NaCl reabsorption,19,20⇓ and impaired endothelial function.13 Because COX-1 may regulate NaCl reabsorption and endothelial function, we assessed the sleep-to-awake BP pattern in these mice using telemetry to test the hypothesis that COX-1 products mediate the normal circadian fluctuation in BP.
The first aim of this study was to determine the role of COX-1 in the regulation of BP and renal hemodynamics. The MAP was measured by telemetry in unanesthetized mice and by intra-arterial recording under anesthesia during clearance studies in COX-1 gene–deficient (COX-1−/−) and wild (COX-1+/+) mice. The second aim of this study was to evaluate the role of COX-1 on the circadian regulation of several bioactive markers. Mice urine samples were collected separately during day and night. The renal excretion of sodium, potassium, metanephrine, normetanephrine, aldosterone, nitrate plus nitrite (NOx), and prostaglandins or their metabolites (TxB2, 6-ketoPGF1α, PGE2, and 11-βPGF2α) were measured to assess their potential roles in the functional changes observed.
The generation and genotyping of COX-1 gene–deficient mice (Black6/129 background) have been reported.21 Male homozygous (COX-1−/−) and wild (COX-1+/+) mice were housed in a quiet room at 25°C with a 12-hour light/dark cycle with free access to food and water. This study was approved by the Georgetown University Animal Care and Use Committee.
The surgical procedure to implant telemetry is detailed in the online data supplement (available at http://www.hypertensionaha.org). The hourly BP, heart rate (HR), and physical activity data were collected, and the mean and standard error of the mean (SEM) values for groups of mice in synchronized time were calculated and plotted as shown in Figure 1. The mean of the lowest 4 sets of MAPs, HRs, and physical activities during the 6:00 am to 6:00 pm (day time) within the same day (sleep period) and the highest 4 sets of MAPs, HRs, and physical activities during the next 6:00 pm to 6:00 am (nighttime: awake time period) were calculated in individual COX-1+/+ (n=10) and COX-1−/− mice (n=9).
Renal Function Studies
The protocol of the renal function study is detailed in the online data supplement.
Renal Excretion of Sodium, Potassium, TxB2, 6-ketoPGF1α, PGE2, 11-βPGF2α, Nephrines, and NOx
Mice (n=6 each group) were housed in mouse metabolic cages (Nalgene Nunc International, Rochester, NY) with NOx-free synthetic diet (Na+ content 0.4 g · 100 g−1). Urine was collected from 6:00 am to 6:00 pm for sleep (daytime) sample and the next 6:00 pm to 6:00 am for awake (nighttime) sample. Measurement of each parameter is detailed in the online data supplement.
Chemical Methods and Statistics
Chemical methods were those published previously.22 Statistics are detailed in the online data supplement.
Hematocrit, Total Body and Kidney Weights, Plasma Electrolytes, and Urinary Volume of COX-1−/− and COX-1+/+ Mice
There were no differences in these parameters between COX-1+/+ and COX-1−/− mice (supplemental Table I, available online at http://www.hypertensionaha.org).
MAP, HR, and Physical Activity Measured by Telemetry
Figure 1 shows the hourly mean±SEM values of MAP (Figure 1A), HR (Figure 1B), and physical activity (supplemental Figure I) in groups of COX-1−/− and COX-1+/+ mice measured by telemetry for 5 days. In both groups, MAP, HR, and activity were elevated at night (awake) and reduced during the day (sleep). Figure 2 shows the calculated sleep and awake MAPs (Figure 2A) and sleep-to-awake BP ratios (Figure 2B). Although there was no significant difference in awake MAP (COX-1+/+: 131±2 mm Hg; COX-1−/−: 126±3 mm Hg), the sleep MAP (COX-1+/+: 93±1 mm Hg; COX-1−/−: 97±2 mm Hg) and the sleep-to-awake BP ratio was higher in COX-1−/− mice by 8.6%. This was accompanied by a 5.7% increase in sleep-to-awake HR ratio (Figure 3), but there were no differences in physical activity (supplemental Figure II).
MAP and HR Under Anesthesia
When measured by direct intra-arterial catheterization under anesthesia with moderate salt loading, the MAP of COX-1−/− mice (124±4 mm Hg) was higher than COX-1+/+ mice (109±5 mm Hg), but there was no difference in HR (supplemental Figure III).
Renal Function Under Anesthesia
The glomerular filtration rate was not different (COX-1+/+: 0.89±0.06 mL · min−1 · g−1; COX-1−/−: 0.93±0.04 mL · min−1 · g−1), but COX-1−/− mice had a reduced renal blood flow (RBF) (COX-1+/+: 4.7±0.2 mL · min−1 · g−1; COX-1−/−: 4.1±0.2 mL · min−1 · g−1; supplemental Figure IVA and IVB) and increased filtration fraction (FF) (COX-1+/+: 34±2%; COX-1−/−: 40±2%) and renal vascular resistance (RVR) (COX-1+/+: 24±2 mm Hg · mL−1 · min−1 · g−1; COX-1−/−: 31±2 mm Hg · mL−1 · min−1 · g−1; supplemental Figure IVC and IVD).
Renal Excretion of Sodium and Potassium
The excretion of sodium and potassium was reduced during sleep (supplemental Figure V). COX-1 did not affect excretion of sodium or potassium.
Renal Excretion of TxB2, PGE2, 6-ketoPGF1α, and 11-βPGF2α
The renal excretion of TxB2 in the COX-1+/+ awake, COX-1−/− awake, COX-1+/+ sleep, and COX-1−/− sleep groups are 4.0±0.2, 0.4±0.1, 3.8±0.3, and 0.4±0.1 g · g−1 creatinine · 10−6, respectively. The renal excretion of PGE2 in the COX-1+/+ awake, COX-1−/− awake, COX-1+/+ sleep, and COX-1−/− sleep groups are 1.9±0.4, 0.4±0.1, 1.4±0.6, and 0.4±0.1 g · g−1 creatinine · 10−6, respectively. The renal excretion of 6-ketoPGF1α in the COX-1+/+ awake, COX-1−/− awake, COX-1+/+ sleep, and COX-1−/− sleep groups are 1.6±0.1, 1.0±0.1, 1.3±0.1, and 0.8±0.1 g · g−1 creatinine · 10−6, respectively. The renal excretion of 11-βPGF2α in the COX-1+/+ awake, COX-1−/− awake, COX-1+/+ sleep, and COX-1−/− sleep groups are 0.090±0.009, 0.062±0.004, 0.064±0.005, and 0.050±0.003 g · g−1 creatinine · 10−6, respectively. COX-1−/− had a 89% reduction in the excretion of TxB2, a 76% reduction in PGE2, a 40% reduction in 6-ketoPGF1α, and a 27% reduction in 11-βPGF2α. Excretion of 6-ketoPGF1α and 11-βPGF2α were reduced during sleep, but TxB2 or PGE2 were not changed (Figure 4).
Renal Excretion of Catecholamine Metabolites
The renal excretion of metanephrine in the COX-1+/+ awake, COX-1−/− awake, COX-1+/+ sleep, and COX-1−/− sleep groups are 0.037±0.005, 0.049±0.004, 0.032±0.004, and 0.056±0.007 g · g−1 creatinine · 10−3, respectively. The renal excretion of normetanephrine in the COX-1+/+ awake, COX-1−/− awake, COX-1+/+ sleep, and COX-1−/− sleep groups are 6.9±0.9, 5.1±0.9, 3.2±0.6, and 4.9±0.7 g · g−1 creatinine · 10−3, respectively. Excretion of metanephrine was not different during awake time and sleep time but was increased in COX-1−/− mice. Excretion of normetanephrine was reduced during sleep time. COX-1−/− mice had a higher normetanephrine excretion only during the sleep time (Figure 5A and 5B).
Renal Excretion of NOx and Aldosterone
The renal excretion of NOx in the COX-1+/+ awake, COX-1−/− awake, COX-1+/+ sleep, and COX-1−/− sleep groups are 0.33±0.06, 0.22±0.04, 0.19±0.03, and 0.12±0.02 mol · mol−1 creatinine · 10−6, respectively. The renal excretion of aldosterone in the COX-1+/+ awake, COX-1−/− awake, COX-1+/+ sleep, and COX-1−/− sleep groups are 0.017±0.002, 0.017±0.003, 0.012±0.001, and 0.010±0.002 g · g−1 creatinine · 10−3, respectively. Excretion of NOx and aldosterone were reduced during sleep time. COX-1−/− mice had reduced excretion of NOx but not aldosterone (Figure 5C and 5D).
Correlation Analysis Among Tx/PGs Metabolites and NOx
Renal excretion of NOx was positively correlate with excretion of 6-ketoPGF1α and 11-βPGF2α (Figure 6A and 6B), but not with TxB2 or PGE2 (supplemental Figure VI).
Correlation Analysis Between Sodium and NOx
Renal excretion of sodium was positively correlated with NOx (Figure 6C).
A novel finding is that COX-1−/− mice have higher sleep time MAP and increased ratios of sleep:awake BP and HR. The importance of circadian rhythm in BP was recognized by O’Brien et al who introduced the “dipper/nondipper” classification and showed that nondipper status is a risk factor for stroke.14 Staessen et al and Okubo et al15,16⇓ showed further that a 5% increment in the night-to-day BP ratio is associated with a 20% increase in cardiovascular events or death even in normotensive subjects. The normal decrement in night-to-day BP in humans has been related to nocturnal reductions in physical activity18 and sympathetic nervous system tone.17 However, increased NaCl reabsorption19,20⇓ or endothelial dysfunction13 impair the normal night-to-day BP ratio. We found that COX-1 deficiency does not change physical activity, but increases sleep:awake HR ratio. Because HR is regulated by the sympathetic nervous system, a blunting of the nocturnal reduction in the sympathetic nervous system tone could be a cause for the higher sleep:awake BP and HR ratios in COX-1−/− mice. We measured normetanephrine excretion to index sympathetic nervous system activity. The result confirmed that normetanephrine excretion in wild-type mice declines by 60% while the animals are asleep. In contrast, there were no circadian changes in COX-1−/− mice (Figure 5B) (interaction day−night×COX-1; P<0.01). Consequently, COX-1−/−, relative to COX-1+/+ mice, have an increased normetanephrine excretion while asleep, consistent with increased sympathetic nervous system tone. COX-1−/− mice also had increased excretion of metanephrine, suggesting increased epinephrine secretion. The excretion of NOx or aldosterone failed to explain the higher sleep:awake BP ratio in COX-1−/− mice (Figure 5C and 5D) (interaction day−night×COX-1; not significant).
Presently, the cause of the enhanced sympathetic nervous system activity during sleep time in COX-1−/− mice is not clear. PGE2 can activate sympathetic nerves in kidney23 and in central nervous system (paraventricular nucleus of hypothalamus),24 and PGD2 can activate norepinephric neurons in the hypothalamus.25 Whether local COX-1–dependent PGE2/D2 production in the brain or peripheral nerves underlies the circadian rhythm of sympathetic nervous system activity requires further study.
COX-1−/− mice had an increased MAP when studied under anesthesia with moderate saltwater loading (1.5% albumin in saline at a rate of 0.35 mL/h) accompanied by an increased RVR and FF and a decreased RBF. This renal vasoconstriction may have contributed to the increased BP. Because anesthetic drugs blunt the action of catecholamines, the cause of the renal vasoconstriction in COX-1−/− mice under anesthesia is not likely to be attributed to an increase in catecholamines and sympathetic nervous activity. It also cannot be attributed to an increased ratio of vasoconstrictor TxA2 to vasodilatator PGs, because COX-1−/− mice had a more marked reduction in renal excretion of TxB2 than 6-ketoPGF1α, consistent with previous studies that TxA2 is mainly generated by COX-1, whereas both COX-1 and COX-2 contribute to the production of PGI2.26 There was no difference in excretion of aldosterone between COX-1−/− and COX-1+/+ mice.
Our finding that COX-1−/− mice have a reduced excretion of NO metabolites suggests that the cause for the renal vasoconstriction seen in COX-1−/− mice may rather relate to defective generation of NO. Similar to our finding in COX-1−/− mice, NSAIDs reduce NO production.27,28⇓ We found a strong positive correlation between the renal excretion of sodium and NOx. Inhibition of NO generation with l-nitro methyl arginine ester causes salt-sensitivity, accompanied by reduced renal Na+ excretion and renal vasoconstriction with an elevated FF and hypertension.29–33⇓⇓⇓⇓ These are all features of the COX-1−/− mouse in this study. Therefore, a reduction in NO generation may underlie the hemodynamic change seen in COX-1−/− mice.
The role of PGs in NO production needs further study. We found strong correlations between the renal excretion of NOx with PGI2 and PGD2 metabolites, but not with TxA2 or PGE2 metabolites. Zenge34 has shown PGI2-induced pulmonary vasodilatation is mediated by NO release. Niwano35 has shown that the PGI2 analogue, Beraprost sodium, increases endothelial nitric oxide synthase gene expression and releases NO from endothelial cells via cAMP signaling. PGD2 increases endothelial nitric oxide synthase expression and NO generation in the developing choroids.36 The PGD2 metabolite, 15-deoxy-delta12,14⇓-PG J2 (15d-PGJ2), also increases NO generation by endothelial cells,37 but the molecular mechanism is not yet established.
NSAIDs can cause renal vasoconstriction and raise BP in patients with hypertension. The similar profile of effects of COX-1 gene deletion in this study to that of NSAIDs recipients suggest that some of the adverse effects associated with NSAIDs use may originate from inhibition of COX-1. Likely, these potential adverse cardiovascular effects are offset by inhibition of platelet COX-1 that will inhibit thrombogenesis. The profound reduction in NOx excretion and increased nocturnal excretion of normetanephrine in COX-1−/− mice implies a new role for this enzyme, and likely its products, PGI2 and PGD2 or their metabolites in NO generation and sympathetic drive. Defects in NO generation are linked to atherosclerosis and adverse cardiovascular outcome in many clinical studies. These findings expose complex, and perhaps discordant, effects of COX-1 metabolites on potential cardiovascular risk factors that add to the growing concern of unanticipated cardiovascular events in patients using drugs that inhibit cyclooxygenase.
This work was supported by grants from the National Institute of Diabetes, Digestive, and Kidney Disease (DK-36,079 and DK-49,870), the National Heart, Lung, and Blood Institute (HL-68686), and American Heart Association Beginning grant-in-aid (PI: N Kawada; 0465440U), and by funds from the George E Schreiner Chair of Nephrology.
- Received November 22, 2004.
- Revision received December 16, 2004.
- Accepted April 2, 2005.
- ↵DeWitt DL, Smith WL. Primary structure of prostaglandin G/H synthase from sheep vesicular gland determined from the complementary DNA sequence. Proc Natl Acad Sci U S A. 1988; 85: 1412–1416.
- ↵Sheng H, Shao J, Morrow JD, Beauchamp RD, DuBois RN. Modulation of apoptosis and Bcl-2 expression by prostaglandin E2 in human colon cancer cells. Cancer Res. 1998; 58: 362–366.
- ↵Ishaque A, Dunn MJ, Sorokin A. Cyclooxygenase-2 inhibits tumor necrosis factor alpha-mediated apoptosis in renal glomerular mesangial cells. J Biol Chem. 2003; 278: 10629–10640.
- ↵Cutler NS, Graves-Deal R, LaFleur BJ, Gao Z, Boman BM, Whitehead RH, Terry E, Morrow JD, Coffey RJ. Stromal production of prostacyclin confers an antiapoptotic effect to colonic epithelial cells. Cancer Res. 2003; 63: 1748–1751.
- ↵Harris RC, Breyer MD. Physiological regulation of cyclooxygenase-2 in the kidney. Am J Physiol. 2001; 281: F1–F11.
- ↵Shaw JA, Chin-Dusting JP, Kingwell BA, Dart AM. Diurnal variation in endothelium-dependent vasodilatation is not apparent in coronary artery disease. Circulation. 2001; 103: 806–812.
- ↵Staessen JA, Thijs L, Fagard R, O’Brien ET, Clement D, de Leeuw PW, Mancia G, Nachev C, Palatini P, Parati G, Tuomilehto J, Webster J. Predicting cardiovascular risk using conventional vs ambulatory blood pressure in older patients with systolic hypertension. Systolic Hypertension in Europe Trial Investigators. JAMA. 1999; 282: 539–546.
- ↵Ohkubo T, Hozawa A, Yamaguchi J, Kikuya M, Ohmori K, Michimata M, Matsubara M, Hashimoto J, Hoshi H, Araki T, Tsuji I, Satoh H, Hisamichi S, Imai Y. Prognostic significance of the nocturnal decline in blood pressure in individuals with and without high 24-h blood pressure: the Ohasama study. J Hypertens. 2002; 20: 2183–2189.
- ↵Dodt C, Breckling U, Derad I, Fehm HL, Born J. Plasma epinephrine and norepinephrine concentrations of healthy humans associated with nighttime sleep and morning arousal. Hypertension. 1997; 30: 71–76.
- ↵Uzu T, Ishikawa K, Fujii T, Nakamura S, Inenaga T, Kimura G. Sodium restriction shifts circadian rhythm of blood pressure from nondipper to dipper in essential hypertension. Circulation. 1997; 96: 1859–1862.
- ↵Langenbach R, Morham SG, Tiano HF, Loftin CD, Ghanayem BI, Chulada PC, Mahler JF, Lee CA, Goulding EH, Kluckman KD, Kim HS, Smithies O. Prostaglandin synthase 1 gene disruption in mice reduces arachidonic acid-induced inflammation and indomethacin-induced gastric ulceration. Cell. 1995; 83: 483–492.
- ↵Kawada N, Imai E, Karber A, Welch WJ, Wilcox CS. A mouse model of angiotensin II slow pressor response: role of oxidative stress. J Am Soc Nephrol. 2002; 13: 2860–2868.
- ↵Zhang ZH, Wei SG, Francis J, Felder RB. Cardiovascular and renal sympathetic activation by blood-borne TNF-alpha in rat: the role of central prostaglandins. Am J Physiol. 2003; 284: R916–R927.
- ↵Belton O, Byrne D, Kearney D, Leahy A, Fitzgerald DJ. Cyclooxygenase-1 and -2-dependent prostacyclin formation in patients with atherosclerosis. Circulation. 2000; 102: 840–845.
- ↵Ribeiro MO, Antunes E, de Nucci G, Lovisolo SM, Zatz R. Chronic inhibition of nitric oxide synthesis. A new model of arterial hypertension. Hypertension. 1992; 20: 298–303.
- ↵Zanzinger J. Role of nitric oxide in the neural control of cardiovascular function. Cardiovasc Res. 2000; 43: 639–649.
- ↵Zenge JP, Rairigh RL, Grover TR, Storme L, Parker TA, Kinsella JP, Abman SH. NO and prostaglandin interactions during hemodynamic stress in the fetal ovine pulmonary circulation. Am J Physiol. 2001; 281: L1157–L1163.
- ↵Niwano K, Arai M, Tomaru K, Uchiyama T, Ohyama Y, Kurabayashi M. Transcriptional stimulation of the eNOS gene by the stable prostacyclin analogue beraprost is mediated through cAMP-responsive element in vascular endothelial cells: close link between PGI2 signal and NO pathways. Circ Res. 2003; 93: 523–530.
- ↵Dumont I, Hardy P, Peri KG, Hou X, Molotchnikoff S, Varma DR, Chemtob S. Regulation of endothelial nitric oxide synthase by PGD2 in the developing choroids. Am J Physiol. 2000; 278: H60–H66.
- ↵Calnek DS, Mazzella L, Roser S, Roman J, Hart CM. Peroxisome proliferator-activated receptor gamma ligands increase release of nitric oxide from endothelial cells. Arterioscler Thromb Vasc Biol. 2003; 23: 52–57.