|Home | About | Journals | Submit | Contact Us | Français|
Sleep apnea, defined as intermittent respiratory arrest during sleep, is associated with increased incidence of hypertension and peripheral vascular disease. Exposure of rodents to brief periods of intermittent hypercarbia/hypoxia (H-IH) during sleep mimics the cyclical hypoxia-normoxia of sleep apnea. Endothelin-1, an upstream activator of nuclear factor of activated T cells (NFAT), is increased during H-IH. Therefore, we hypothesized that NFATc3 is activated by H-IH and is required for H-IH-induced hypertension. Consistent with this hypothesis, we found that H-IH (20 brief exposures per hour to 5% O2-5% CO2 for 7 h/day) induces systemic hypertension in mice [mean arterial pressure (MAP) = 97 ± 2 vs. 124 ± 2 mmHg, P < 0.05, n = 5] and increases NFATc3 transcriptional activity in aorta and mesenteric arteries. Cyclosporin A, an NFAT inhibitor, and genetic ablation of NFATc3 [NFATc3 knockout (KO)] prevented NFAT activation. More importantly, H-IH-induced hypertension was attenuated in cyclosporin A-treated mice and prevented in NFATc3 KO mice. MAP was significantly elevated in wild-type mice (Δ = 23.5 ± 6.1 mmHg), but not in KO mice (Δ = –3.9 ± 5.7). These results indicate that H-IH-induced increases in MAP require NFATc3 and that NFATc3 may contribute to the vascular changes associated with H-IH-induced hypertension.
sleep apnea (SA), defined as intermittent respiratory arrest during sleep, affects up to 20% of the adult population. Among the major consequences of SA are decreased O2 saturation (hypoxia) and increased CO2 saturation (hypercapnia). In SA patients, incidence of hypertension, peripheral vascular disease, stroke, and sudden cardiac death is increased (for review see Refs. 21 and 44). Thus SA appears to directly initiate vascular changes that predispose individuals to cardiovascular disease.
Recently, it was demonstrated that exposure of rodents to periods of intermittent hypoxia with CO2 supplementation [intermittent hypercarbia/hypoxia (H-IH)] during sleep mimics the cyclical hypoxia-normoxia of SA. In rats exposed to H-IH, blood pressure (34, 35) and circulating endothelin-1 (ET-1) are elevated (35), vasoconstrictor sensitivity to ET-1 is increased (2), and blood pressure is normalized with ET antagonists (35). These studies, together with clinical studies showing increased circulating ET-1 in SA (10, 55), suggest that augmented ET-1 vasoconstriction contributes to SA hypertension. The sustained increased blood pressure observed in SA patients and animal models also appears to involve activation of the sympathetic nervous and renin-angiotensin systems and diminished activity of nitric oxide synthase (for review see Refs. 44 and 52).
ET-1 acts through Gαq-coupled receptors as a Ca2+-mobilizing agent and as a potent activator of the nuclear factor of activated T cells (NFAT) (1, 25, 57). ANG II, glucose through UTP release, and α1-adrenergic agonists are also potent stimuli of NFAT activation (1, 25, 27, 46, 57).
NFAT, in turn, serves as a “master transcriptional regulator,” interlinking Ca2+ signaling with other signaling pathways to induce specific genetic programs. NFAT activation is regulated primarily through control of its subcellular localization (54). Ca2+-elevating stimuli activate the Ca2+-dependent phosphatase calcineurin, which dephosphorylates NFAT, leading to nuclear translocation. NFAT activation is generally associated with an increase in gene expression (31). However, several reports have shown that NFAT may also function as a repressor of gene expression (3, 29, 45).
Of the four members of the NFAT family, NFATc3 is specifically implicated in vascular development (28) and maintenance of the contractile phenotype in vascular smooth muscle cells (VSMC) through regulation of α-actin expression (11, 26, 59). NFATc3 also modulates VSMC contractility by regulating voltage-dependent and large-conductance K+ channel expression in cerebral arteries (3, 45).
Although many NFAT transcriptional partners are induced by hypoxia [activator protein-1, GATA, and cAMP response element-binding protein (CREB)-binding protein/p300] (4, 34, 38, 53, 61), NFAT activation in SA patients or animal models has not been described.
These intriguing observations led us to hypothesize that H-IH activates NFATc3 in systemic arterial smooth muscle, where it contributes to the development of hypertension.
The aims of the present study were to determine whether 1) H-IH induces hypertension in mice, 2) H-IH induces ET-1 upregulation; 3) H-IH induces NFATc3 activation in systemic arteries; and 4) NFATc3 is required for H-IH-induced hypertension.
In support of the hypothesis, our data demonstrate that H-IH indeed causes hypertension and increases NFATc3 transcriptional activity in mouse arteries. More importantly, pharmacological inhibition of calcineurin activation of NFAT [cyclosporin (CsA)] or the genetic ablation of NFATc3 prevents H-IH-induced hypertension. In addition, upregulation of ET-1 in the lung is correlated with NFATc3 activation in the aorta. These results suggest that NFATc3 plays a role in the vascular and blood pressure changes observed during H-IH.
All protocols were reviewed and approved by the Institutional Animal Care and Use Committee of the University of New Mexico Health Science Center.
Adult male 9×NFAT-luciferase reporter (NFAT-luc) background strain FVBN, NFATc3-knockout (NFATc3 KO), Balb/C wild-type (WT), and FVBN mice (25–30 g) were used. In addition, NFAT-luc mice were backcrossed with NFATc3 KO mice for at least two generations, and NFAT-luc/NFATc3+/+ (WT) and NFAT-luc/NFATc3–/– (KO) mice were used (11). NFAT-luc mice were provided by Dr. Jeffery D. Molkentin (Department of Pediatrics, Children's Hospital Medical Center, Cincinnati, OH) (7), and NFATc3 KO mice were kindly provided by Dr. Laurie Glimcher (Harvard University) (49). Heterozygous mice were bred to obtain age-matched WT (Balb/C) and KO mice. Importantly, it has been shown that loss of NFATc3 does not induce a compensatory upregulation of other NFAT isoforms (60). Cyclosporin A (Calbiochem) or Cremophor EL (vehicle) was administered subcutaneously at 25 mg·kg–1·day–1 for 3 or 10 days to FVBN and NFAT-luc mice.
The animals were housed in regular cages with snug-fitting Plexiglas lids. During the normal sleep period, air in the cage was cycled between a low (5%) O2-high (5%) CO2 environment and room air. In rats, this protocol causes hypoxia but maintains constant plasma CO2 levels, while exposure to 5% O2 alone induces profound hypocapnia (56). The mice were exposed to 20 H-IH episodes per hour for 7 h/day for 2, 7, or 21 days. Control animals were housed in similar cages with a constant flow of air. All animals were maintained on a 12:12-h light-dark cycle. SA patients experience 5–100 apneic episodes per hour of sleep.
Telemeter catheters were implanted using a modification of the procedure of Butz and Davisson as described previously (13). FVBN mice were given buprenorphine hydrochloride (Buprenex, 0.5 mg/kg sc) 20 min before surgery and anesthetized using isoflurane (5% supplemented with O2). The catheter tip was inserted and secured in the carotid artery, and the transmitter body (model PA-C20, Data Sciences International) was secured subcutaneously above the right flank. Warmed sterile 0.9% NaCl solution (0.5 ml sc) was given after surgery, and the mice were allowed to recover for 5–7 days before recording was started. Mean arterial pressure (MAP), heart rate (HR), and activity were recorded daily for up to 36 days and analyzed using a telemetry analyzer (JCL Consultants). Each animal was used as a single n value.
As reported previously by Nieves-Cintron et al. (45), NFATc3 KO mice did not survive carotid artery catheter implantation surgery. We also observed that these mice do not survive femoral artery catheterization. Because breeding of this colony is very slow, systolic, diastolic, and mean blood pressures were monitored using the Coda 2 noninvasive blood pressure system (Kent Scientific, Torrington, CT) in the WT and NFATc3 KO mice immediately after the daily cycling. Animals were acclimated for 5 days before blood pressure was recorded. The Coda system utilizes volume-pressure-recording sensor technology for measurement of the mouse tail blood pressure parameters: systolic blood pressure, diastolic blood pressure, HR, mean blood pressure, tail blood flow, and tail blood volume. Volume-pressure recording has been clinically validated to provide close to 100% correlation with telemetry and direct blood pressure measurements for systolic and diastolic blood pressure (41, 50).
Blood samples were obtained by direct cardiac puncture at the time of tissue collection for measurement of hematocrit. After isolation of the heart, the atria and major vessels were removed from the ventricles. The right ventricle (RV) was dissected from the left ventricle (LV) and septum. The degree of RV hypertrophy was expressed as the ratio of RV weight to total ventricular weight (RV + LV) plus septum weight, and LV hypertrophy was expressed as the ratio of LV weight to body weight.
After NFAT-luc mice were euthanized with an overdose of pentobarbital sodium (200 mg/kg ip), the aorta, brain and small intestines were removed and placed in ice-cold HEPES-physiological saline solution. The aorta and cerebral and mesenteric arteries were isolated and lysed using tissue lysis buffer (Promega) according to the manufacturer's protocol. The lysate was centrifuged for 10 min at 10,000 rpm. Luciferase activity and protein content were determined in the supernatant. Luciferase activity was measured using a luciferase assay system kit (Promega), and light was detected with a luminometer (model TD20/20, Turner). Protein content was determined by the Bradford method (Bio-Rad) and used to normalize luciferase activity.
LV and isolated arteries were stored in RNAlater (Ambion). Total RNA was isolated using the RNeasy Mini Kit (Qiagen) following the manufacturer's protocol, with the addition of DNase Treatment (DNase Turbo, Ambion) and RNeasy MinElute Cleanup Kit (Qiagen) for removal of possible genomic DNA contamination. The High Capacity Reverse Transcription Kit (Applied Biosystems) was used for reverse transcription of total RNA to cDNA. For real-time PCR detection, SYBR Green Master Mix (Applied Biosystems) was used with the following sets of primer pairs: 5′-AGGTGCTGTCCCAGATGATTCTGT-3′ and 5′-TTGTGAGGCCTTGGTCCTTCAAGA-3′ for brain natriuretic peptide (BNP) in LV, 5′-GACATCATCTGGGTCAACACTC-3′ and 5′-CATCTAACTGCCTGGTCTGTG-3′ for prepro-ET-1, and 5′-TCAACTTTCGATGGTAGTCGCCGT-3′ and 5′-TCCTTGGATGTGGTAGCCGTTTCT-3′ for 18S RNA (endogenous control). PCR was performed using the model 7500 Fast Real-Time PCR System (Applied Biosystems). The normalized gene expression method (2–ΔΔCT) was used for relative quantification of gene expression (40).
Values are means ± SE. Statistical significance was tested at 95% (P < 0.05) confidence level using unpaired t-test, one-way ANOVA followed by Tukey's multiple comparisons test, or two-way ANOVA followed by the Holm-Sidak test.
Animals exposed to H-IH showed a significant increase in MAP (Fig. 1A) on day 1, and this elevation was sustained up to the end of the experiment (day 17). This increase was observed during H-IH exposure and also during hours of air exposure. Interestingly, the normal fall in blood pressure during the day (sleep period) was inverted in H-IH-exposed mice. Activity also followed an inverted pattern (Fig. 1B), although no significant changes were observed in the level of activity.
Because of the change in diurnal blood pressure patterns, blood pressure during the hours of air exposure was no longer different between groups at the same time of day after 5 days of H-IH. However, it was different between groups when the comparison was made during the active hours (light for H-IH and dark for air) and during the inactive hours (dark for H-IH and light for air).
HR was not different at any time point (data not shown). Therefore, 7 h of H-IH daily during sleep appears to induce systemic hypertension in mice and disrupt normal activity patterns.
Additional parameters were determined in mice exposed to air and H-IH for up to 21 days. H-IH-exposed mice did not develop polycythemia over the course of 21 days (Table 1). Normalized LV weight did not increase over the period of H-IH exposure, suggesting no LV hypertrophy (Table 1). To further confirm that elevations in MAP were not associated with LV hypertrophy, BNP transcript levels were measured in LV from air- and H-IH-exposed mice at every time point. As shown in Table 1, no significant changes were observed in LV BNP mRNA compared with LV from air-exposed animals. In addition, no changes in the ratio of RV to RV + LV + septum weight (Table 1) were observed over the course of the experiment, suggesting no RV hypertrophy and pulmonary hypertension development.
Consistent with our previous observations in rats (35), H-IH induced a significant increase in mRNA levels of prepro-ET-1, the precursor of ET-1, in the lung (Fig. 2). Interestingly, prepro-ET-1 increased significantly after 2 and 7 days of H-IH and returned to air values at day 21.
Since NFAT has been shown to be activated in cerebral artery of ANG II-hypertensive mice (3, 45), we determined the degree of NFAT activation in aorta and cerebral and mesenteric arteries isolated from H-IH-exposed NFAT-luc mice. Luciferase activity, which reflects NFAT transcriptional activity, was significantly elevated in aorta and mesenteric artery after 2 and 7 days of H-IH (Fig. 3, A and B). In aorta, luciferase activity returned to air control values after 21 days of H-IH. Interestingly, there was a strong positive correlation between the elevation in aortic NFAT activity and prepro-ET-1 transcript levels in the lung (Fig. 3C). In mesenteric artery at day 21, luciferase activity tended to remain elevated compared with the air-exposed group, although the difference was not statistically significant (P = 0.14). In addition, a trend was observed for a correlation in the timing and degree of mesenteric artery NFAT activation and ET-1 upregulation, but it was not significant (Fig. 3D). These results are consistent with previous reports showing that ET-1 induces an increase in NFATc3 nuclear accumulation in vascular smooth muscle (25, 48). Taken together, these results suggest that ET-1 may be an upstream mediator of H-IH-induced NFAT activation in aorta and mesenteric artery.
NFAT was not significantly activated by H-IH in cerebral arteries (2.60 ± 0.43 in air and 2.87 ± 0.44, 3.22 ± 0.83, and 2.51 ± 1.10 at days 2, 7, and 21, respectively, of H-IH). Interestingly, basal NFAT activity is higher in cerebral artery than in mesenteric artery and aorta: 2.60 ± 0.43 in cerebral artery (P < 0.05 vs. mesenteric artery and aorta), 1.53 ± 0.29 in mesenteric artery (P < 0.05 vs. cerebral artery and aorta), and 0.16 ± 0.02 in aorta (n = 6).
To address whether activation of calcineurin is upstream of NFAT activation, NFAT-luc mice were treated with CsA or vehicle for 3 days before and during 2 days of H-IH exposure. As expected, CsA completely prevented NFAT activation by H-IH in aorta and mesenteric artery (Fig. 4). These findings suggest that H-IH-induced NFAT activation requires calcineurin.
Since it has been previously demonstrated that NFATc3 significantly contributes to VSMC function (3, 26, 27), we sought to determine the contribution of this NFAT isoform to the H-IH-induced increase in NFAT activity. Therefore, we determined the effect of 2 days of H-IH on reporter expression in aorta and mesenteric artery isolated from NFAT-luc/NFATc3 WT and NFAT-luc/NFATc3 KO mice. Similar to results in Fig. 3, reporter activity increased in aorta and mesenteric artery after 2 days of H-IH in WT mice, but the reporter response to H-IH was significantly impaired in tissues isolated from NFAT-luc/NFATc3 KO mice (Fig. 5). NFATc3 nuclear accumulation consistently increased after H-IH exposure in aortic smooth muscle cells. Aortas from animals used in the experiments reported in Fig. 4 were used in Fig. 6 to show that 2 days of H-IH exposure increases aortic smooth muscle cell NFATc3 nuclear accumulation and that the increase is prevented by CsA.
These results support the conclusion that NFATc3 significantly contributes to H-IH-induced increases in NFAT transcriptional activity and further support our focus on this isoform.
To determine the role of calcineurin/NFAT in H-IH-induced hypertension, animals were treated with CsA for 3 days before 7 days of H-IH exposure and MAP was recorded. Interestingly, CsA significantly attenuated the H-IH-induced increase in MAP (Fig. 7A) without affecting basal MAP during air exposure.
Since calcineurin does not exclusively regulate NFAT activity (54), we also measured changes in MAP in WT and NFATc3 KO mice exposed to H-IH for 7 days. Blood pressure was measured for 5 days of air exposure and at the end of the daily H-IH cycling for 7 days (both at the same time of day). Consistent with the results shown in Fig. 1, H-IH induced a significant and immediate increase in MAP in WT mice but failed to elevate MAP in NFATc3 KO mice. These results demonstrate that NFATc3 is required for the development of H-IH-induced systemic hypertension.
The present study demonstrates that H-IH induces systemic hypertension in mice and activates the Ca+2/calcineurin-mediated transcription factor NFATc3 in aorta and mesenteric artery but does not cause concomitant cardiac hypertrophy or polycythemia. More importantly, inhibition of calcineurin or the genetic ablation of NFATc3 (NFATc3 KO) prevents H-IH-induced hypertension. In addition, prepro-ET-1 transcripts are elevated in the lung of mice exposed to H-IH, suggesting that it may be an upstream activator of NFATc3.
Sleep-disordered breathing with recurrent apnea is a major cause of morbidity and mortality. Affected individuals have increased risk of systemic hypertension. To our knowledge, this is the first study reporting the development of systemic hypertension in a mouse model of SA, where mice are exposed to intermittent episodes of low O2 (from 20.9 to 5%) and elevated CO2 (from 0 to 5%). However, previous reports show that intermittent hypoxia (nadir O2 of 5%) without CO2 supplementation also induces systemic hypertension in mice (51, 58).
It is well established that nocturnal hypoxemia in SA patients causes an inversion of the circadian arterial pressure rhythm and triggers nocturnal hypertension (nondipping) (36, 62, 63). Consistently, our results show that mice exposed to H-IH show an inverted pattern in blood pressure and activity. In patients, it has been shown that nondipping hypertension is caused by cardiac preload, sleep disturbance, and hypoxia-induced increases in norepinephrine and ET-1 (62). Accordingly, we found that prepro-ET-1 transcripts are upregulated in the lung. Therefore, it is possible that the same mechanisms underlie the changes observed in our animal model.
Polycythemia is not consistently observed in SA animal models, with or without CO2 supplementation, and SA patients (2, 9, 12, 16, 23, 30, 35, 42, 43, 56). In the present study, H-IH did not induce significant polycythemia in mice. Interestingly, our group recently demonstrated that polycythemia only develops in rats exposed to intermittent hypoxia without CO2 supplementation (56), similar to other rat models (23, 35), where hematocrit does not increase in eucapnic intermittent hypoxia. Recently, Choi et al. (9) reported that SA patients show a very slight increase in hematocrit but do not develop clinical polycythemia, while others (2, 30) also suggested that SA patients do not develop polycythemia. These findings suggest differences in the physiological responses to intermittent hypoxia alone or with CO2 supplementation.
Similar to alterations in hematocrit, the development of cardiac hypertrophy in intermittent hypoxia with or without CO2 is also not consistent. Although several studies report that intermittent hypoxia induces modest-to-severe LV and RV hypertrophy in rodents (2, 8, 22, 23, 58), our study is in agreement with other reports showing that intermittent hypoxia with or without CO2 supplementation does not induce cardiac hypertrophy (5, 17, 56). It is possible that LV hypertrophy did not develop in our mouse model because of CO2 supplementation or that the elevation in systemic blood pressure was below the threshold for activation of the cardiac hypertrophy gene program. Alternatively, proteolysis genes may have been turned on during H-IH to prevent hypertrophy, as suggested by Fan et al. (17).
The lack of RV hypertrophy and polycythemia in our mouse model suggests the absence of pulmonary hypertension, consistent with our recent results in rats exposed to intermittent hypoxia with 5% CO2, which do not show RV hypertrophy, pulmonary arterial remodeling, or pulmonary hypertension (56). These results further support our focus on the systemic side of the circulation.
The present study shows for the first time that H-IH induces NFAT activation in small and large systemic vessels. This result is consistent with our previous report showing that NFATc3 nuclear accumulation increases in VSMC of ex vivo mouse arteries exposed to increased intraluminal pressure (27). Furthermore, Amberg et al. (3) and Nieves-Cintron et al. (45) showed that arterial NFATc3 is activated in ANG II-induced hypertension, independent of elevated blood pressure. Thus the elevated blood pressure during H-IH may not be the cause of the chronically increased NFATc3 activity. Our results suggest that ET-1 is one potential upstream activator of NFAT, since we found that transcripts of prepro-ET-1, the precursor of ET-1 and a potent activator of NFAT (25, 48), are elevated in the lung of mice exposed to H-IH. This tissue is a very important source of circulating ET-1 (14). In addition, it is well established that hypoxia upregulates ET-1 transcription through hypoxia-inducible factor-1α (15, 32, 33, 61) and that this pathway is activated by intermittent hypoxia (37). However, these studies do not exclude other factors, such as ANG II and/or catecholamines (10, 44, 52, 55), in the observed NFAT activation, particularly because heterozygous hypoxia-inducible factor-1α deficiency prevents intermittent hypoxia-induced hypertension and increased plasma norepinephrine (51).
NFAT activation in mesenteric arteries was sustained throughout the 21 days of H-IH, suggesting the importance of NFAT in resistance arteries. It is surprising that NFAT was not significantly activated in cerebral arteries from H-IH-exposed animals, since ET-1, ANG II, and catecholamines, potent stimuli for NFAT activation (1, 25, 27, 57), have been shown to be increased in H-IH (10, 44, 52, 55). However, blood flow control in the cerebral circulation is complex. Increased systemic blood pressure constricts large cerebral arteries with autoregulatory dilation of small vessels to maintain blood flow. In addition, hypoxia is a potent vasodilator in the cerebral circulation (for review see Ref. 19). Thus response to H-IH in the cerebral circulation may be very different from that in other systemic beds such as the mesentery. Alternatively, it is possible that NFAT is already maximally activated in cerebral arteries, since under basal conditions, luciferase levels are higher in cerebral artery than in mesenteric artery and aorta. This could be due to the elevated myogenic tone in this vascular bed under physiological conditions (18, 27), although it seems unlikely on the basis of reports of Amberg et al. (3) and Nieves-Cintron et al. (45), who showed that chronic administration of ANG II activates NFAT in the cerebral vasculature. Another explanation could be that the elevation of NFAT activators (ET-1, ANG II, and catecholamines) in the cerebral circulation is not sufficient to significantly activate NFAT.
NFATc3 is the isoform that seems to predominantly regulate the vasculature (3, 11, 20, 24–27, 45). Consistently, the present study showed that H-IH increases NFATc3 nuclear accumulation in aortic smooth muscle cells and that H-IH-induced NFAT activation is impaired in aorta and mesenteric artery from NFATc3 KO mice, supporting the importance of this isoform in the circulatory system and our focus on NFATc3. In addition, we recently demonstrated that NFATc3 is activated by hypobaric chronic hypoxia (an animal model of pulmonary hypertension) in mouse pulmonary arteries (11) and is required for the associated arterial remodeling. However, an elegant study by Bonnet et al. (6) shows that NFATc3 and NFATc2 are activated in pulmonary arterial smooth muscle cells from patients with secondary pulmonary hypertension and that NFATc2 is required for decreased apoptosis and increased pulmonary arterial remodeling in patients and monocrotaline-induced pulmonary hypertensive rats. Therefore, other NFAT isoforms might be important in rat and human pulmonary circulation, but our data strongly suggest that, in the peripheral vascular system of mice, NFATc3 is the most important isoform.
The most significant finding of our study is that H-IH-induced systemic hypertension is prevented by calcineurin inhibition with CsA and the genetic ablation of NFATc3, clearly demonstrating that NFATc3 is required for the development of hypertension in this model. These results are in agreement with the study of Nieves-Cintron et al. (45), which showed that NFATc3 is required for ANG II-induced hypertension, suggesting that this may be common in multiple forms of hypertension. Because previous reports demonstrate that activation of NFATc3 decreases the expression of voltage-dependent (KV2.1) and large-conductance β-subunit K+ channels in cerebral arteries contributing to hypertension, it is reasonable to speculate that NFATc3 may modulate the expression of these same genes to mediate enhanced vasoconstriction and hypertension during H-IH. This is further supported by our previous finding of increased ET-1-dependent vasoconstriction in mesenteric artery of H-IH-exposed rats (2, 56). Also, intermittent hypoxia has been shown by others to increase norepinephrine-dependent vasoconstriction (12, 39).
In conclusion, NFATc3 is activated in the systemic circulation by H-IH and is required for the development of H-IH-induced hypertension. The present study highlights the potential clinical relevance of inhibition of the NFAT pathway to prevent the development of hypertension and its associated vascular pathology.
We thank Miriam Hutchinson for technical assistance.
This work was supported by American Heart Association Scientist Development Grant 0535347N and a postdoctoral fellowship from the Ministerio de Educación y Ciencia (Spain) to S. de Frutos.