Long-Term Chronic Intermittent Hypobaric Hypoxia in Rats Causes an Imbalance in the Asymmetric Dimethylarginine/Nitric Oxide Pathway and ROS Activity: A Possible Synergistic Mechanism for Altitude Pulmonary Hypertension?

Chronic intermittent hypoxia (CIH) and chronic hypoxia (CH) are associated with high-altitude pulmonary hypertension (HAPH). Asymmetric dimethylarginine (ADMA), a NO synthase (NOS) inhibitor, may contribute to HAPH. This study assessed changes in the ADMA/NO pathway and the underlying mechanisms in rat lungs following exposure to CIH or CH simulated in a hypobaric chamber at 428 Torr. Twenty-four adult Wistar rats were randomly assigned to three groups: CIH2x2 (2 days of hypoxia/2 days of normoxia), CH, and NX (permanent normoxia), for 30 days. All analyses were performed in whole lung tissue. L-Arginine and ADMA were analyzed using LC-MS/MS. Under both hypoxic conditions right ventricular hypertrophy was observed (p < 0.01) and endothelial NOS mRNA increased (p < 0.001), but the phosphorylated/nonphosphorylated vasodilator-stimulated phosphoprotein (VASP) ratio was unchanged. ADMA increased (p < 0.001), whereas dimethylarginine dimethylaminohydrolase (DDAH) activity decreased only under CH (p < 0.05). Although arginase activity increased (p < 0.001) and L-arginine exhibited no changes, the L-arginine/ADMA ratio decreased significantly (p < 0.001). Moreover, NOX4 expression increased only under CH (p < 0.01), but malondialdehyde (MDA) increased (up to 2-fold) equally in CIH2x2 and CH (p < 0.001). Our results suggest that ADMA and oxidative stress likely reduce NO bioavailability under altitude hypoxia, which implies greater pulmonary vascular reactivity and tone, despite the more subdued effects observed under CIH.


Background
Working at high altitude and resting at sea level for many years expose humans to an unusual labor-related condition called long-term chronic intermittent hypobaric hypoxia (CIH) [1]. Exposure to high altitude causes reduced arterial oxygen saturation that, in turn, may elicit various pathophysiological sequelae depending on whether this exposure is acute or chronic. Both acute and chronic exposure cause pulmonary arterial hypertension and an increase in blood hemoglobin levels [2]. Thus, hypobaric hypoxia-induced pulmonary arterial hypertension (HAPH) is a relevant problem that affects populations living and working at high altitudes, such as those in the Andean region and on the Himalayan plateau, with a prevalence varying between 10 and 15% [3].
CIH might lead to the same pulmonary changes as CH [4], but CIH exposure associated with intermittent labor at high altitude has not been studied as thoroughly. A prevalence of pulmonary arterial hypertension up to 4% has been reported among subjects exposed to long-term CIH [5]. Therefore, attempting to understand the complexity of the molecular mechanisms involved in long-term CIH-related pulmonary hemodynamic changes may lead to clarifications in the pathophysiology underlying this and other forms of hypoxia-associated pulmonary diseases.
One of the molecules that plays a key role in regulating vasomotor function under hypoxic conditions is nitric oxide (NO) [6]. NO, which is derived from endothelial cells, dilates almost all types of vessels by stimulating soluble guanylyl cyclase (sGC), resulting in increased cyclic GMP in smooth muscle cells [7]. Asymmetric dimethylarginine (ADMA) is a competitive nitric oxide synthase (NOS) inhibitor that has been identified as a regulator of NO production in vivo [5]. ADMA is formed by the dimethylation of L-arginine residues by arginine methyltransferases and is released by subsequent proteolysis [8]. Leone et al. [9] first reported that endogenous ADMA inhibits endothelium-dependent vasodilation in vitro. Almost 80% of ADMA is degraded by a group of hydrolases called dimethylarginine dimethylaminohydrolases (DDAHs) [8]. Two subtypes of DDAHs, DDAH1 and DDAH2, are known and differ in their tissue distribution and their capacity to degrade ADMA. Currently, DDAH is under investigation as a novel therapeutic target to directly regulate ADMA concentrations and indirectly regulate NO [10].
It is generally accepted that hypoxia is associated with a high burden of oxidative stress. In addition to the many interactions between the L-arginine/NO pathway and oxidative stress, interactions also exist between the ADMA/DDAH pathway and oxidative stress [11]. However, the role of the ADMA/NO pathway in hypoxia-associated chronic respiratory diseases has remained controversial [12]. We recently observed a dramatic increase in ADMA in plasma from volunteers exposed to CIH and high-altitude dwellers [13].
In order to gain a broader view of the changes in the ADMA/NO pathway during long-term CIH compared with CH, this study aimed to assess the changes in the ADMA/NO pathway, the underlying pulmonary molecular mechanisms involved, and the potential interaction with other molecules, such as ROS, in lung tissue as a possible explanation for hypoxia-induced pulmonary hypertension.

Rat Model of CIH.
Twenty-four adult Wistar rats (3 months old) were used for the experiments. The rats were randomly assigned to three groups: CIH2x2 ( = 8; 2 days of hypobaric hypoxia/2 days of normoxia), CH ( = 8; sustained hypoxia), and NX ( = 8; permanent normoxia, for 30 days; this rat model of long-term CIH exposure has been previously used in studies of high-altitude changes [14,15]). To simulate geographical high altitude, the rats were kept in a hypobaric chamber at 428 Torr, which is equivalent to an altitude of 4,600 m, at Universidad Arturo Prat facilities. After 30 days of exposure, the rats were anesthetized with 0.8 mg ketamine/kg and were subjected to a fatal thoracotomy under anesthesia. Lung tissue was snap-frozen for mRNA and protein extraction as well as measurement of L-arginine, ADMA, and enzyme activities. All the measurements, with the exception of physiological parameters, were performed in lung tissue. This protocol was approved by the Ethics Committee of Universidad Arturo Prat, and the animals were handled according to international protocols and under veterinary care.

Physiological
Measurements. The extent of right ventricular hypertrophy (RVH) was estimated by calculation of the ratio of the right ventricle over the left ventricle plus the septum (RV/LV + S) according to [16]. Weight was measured using an Acculab V-12005 (Chicago, IL, USA) electronic balance 1 h after the chamber descended. For blood pressure and heart rate measurements, an inflatable tail-cuff and a pressure sensor (RTBP1003-220; Kent Scientific5, Torrington, CT, USA) were used following a previously validated method [17]. The data were collected and analyzed using the RTBP-001-DS worksheet (Kent Scientific, Torrington, CT, USA). The hematocrit was measured from blood collected by tail vein puncture at baseline and after 30 days of intervention. An Eppendorf AG5 (Hamburg, Germany) microcentrifuge was used to measure the hematocrit level, and the hemoglobin level was measured using a Coulter Electronics Counter, Cell Dyn 37005, and Abbott5 (Santa Clara, CA, USA).

Measurement of L-Arginine and ADMA.
Mass spectrometric determination of L-arginine and ADMA in the lung tissue was performed as previously described using a fully validated high-throughput liquid chromatography/tandem mass spectrometry (LC-MS/MS) assay [18,19]. In brief, samples were analyzed using 96-well 0.20 m microfiltration plates precoated with 40 pmol of [ 2 H 6 ]-ADMA and 800 pmol of L-[ 2 H 7 ]-L-arginine (internal standards). After conversion to their butyl ester derivatives, the analytes were evaluated on a Varian 1200L Triple Quadrupole MS5 (Varian, Walnut Creek, CA, USA) in the positive electrospray ionization (ESI+) mode. The L-arginine/ADMA ratio, which is considered a strong indicator of NOS substrate availability, was calculated [20].

DDAH Activity.
The DDAH activity in the lung tissue was determined as previously described [21]. In brief, tissue samples were homogenized in PBS with protease inhibitors to minimize possible interference from endogenously formed ADMA. The homogenate was centrifuged in a precooled (4 ∘ C) centrifuge for 5 min at 12,000 ×g. Fifty-microliter aliquots of the resulting supernatant were added to 50 L aliquots of PBS containing 20 M [ 2 H 6 ]-ADMA and were incubated at 37 ∘ C for 60 min. The reactions were stopped, and [ 2 H 6 ]-ADMA was determined by LC-MS/MS as described above.
2.5. Arginase Activity. Arginase activity was determined using urea measurement by the conversion of L-arginine to urea. A lung tissue sample (30 mg) was homogenized using RIPA buffer (Homogenizer Glas-Col5, Terre Haute, IN, USA) and was then centrifuged at 4,000 rpm for 2 min over ice at 4 ∘ C. To activate arginase, the tissue was heated Pulmonary Medicine 3 Table 1: Weight and hematocrit in rats exposed to normoxia (NX), chronic intermittent hypoxia (CIH2x2), and chronic hypoxia (CH) at day 0 (baseline) and day 30 (after exposure). Data are expressed as means ± standard error (SE), and differences were obtained by repeated measures ANOVA; * < 0.01; * * < 0.001 compared with the baseline.

Data Analysis.
The results were analyzed using SPSS 17.0 (SPSS, Inc.5, Chicago, IL, USA). Means, standard deviations, standard errors, and confidence intervals were calculated for each parameter. Normality was established using the Kolmogorov-Smirnov test, and all data were normally distributed. Statistical analysis of the differences across all testing conditions was performed using one-way analysis of variance (ANOVA) and after obtaining an ratio <0.05. A Bonferroni post hoc test was performed to identify differences between the three groups. Additionally, for differences in time, a repeated-measures ANOVA was performed. For correlation analysis of ADMA and physiological outcomes, one-tailed parametric Pearson correlation was used. Results were considered significant when the value was less than 0.05.

Rat Physiological Measurements.
Both hypoxic groups lost weight compared with the NX group ( < 0.01), and the weight loss was greater in the CH rats (Table 1). Residual food was left only in the first days of the experiment. As expected, the hematocrit increased in the hypoxic groups ( < 0.01) and was greater in the CH rats ( < 0.01) ( Table 1).

Right Ventricle Hypertrophy Ratio (RVH). Both CH and
CIH2x2 exhibited an RVH, which was expressed as the RV/LV + S ratio. Rats exposed to CIH2x2 showed a slightly lower ratio than CH rats ( < 0.05; Figure 1).

Formation and sGC-Dependent Activity of NO.
Despite a progressive increase (up to four-fold in CIH2x2 and eightfold in CH) in endothelial NOS (eNOS) mRNA expression ( < 0.001; Figure 2(a)) in the lung tissue, the ratio of pVASP to total VASP was unchanged in both hypoxic groups ( NS; Figure 2(b)). The NOS substrate availability, which relates to the mRNA expression and activity of arginases, the enzymes responsible for degrading L-arginine, was also affected by hypoxia. The relative ARG1 mRNA expression level in lung tissue was increased during CH ( < 0.05; Figure 2(c)), whereas ARG2 exhibited a nonsignificant increasing trend for rats under both forms of hypoxia (Figure 2(d)). Although there were only marginal differences in mRNA expression, the arginase activity was strikingly increased during hypoxia ( < 0.001), without differences between CIH2x2 and CH (Figure 2(e)). However, the L-arginine tissue concentration in the lung was not affected by hypoxia (Figure 3(a)), whereas the L-arginine/ADMA ratio decreased according to the level of hypoxia ( < 0.001; Figure 3(b)).

ADMA Formation and Degradation.
Clearly, hypoxia increased the ADMA tissue concentrations in the lung (CIH2x2 < 0.05; CH < 0.05, but CIH2x2 produced intermediate values Figure 4(a)). The DDAH lung tissue activity was reduced only in CH rats ( < 0.05; Figure 4(b)). The relative DDAH1 mRNA and protein levels did not change in either hypoxic group ( NS; Figures 4(c) and 4(e)). Conversely, DDAH2 mRNA expression significantly increased ( < 0.001), whereas DDAH2 protein expression slightly decreased in the CH group ( < 0.05; Figures  4(d) and 4(f)). This latter finding correlates with the DDAH activity result.

Correlation of ADMA and Physiological Outcome.
After 30 days of chronic hypoxia ADMA was positively correlated with systolic blood pressure (Table 2), whereas in the status of CIH no correlation was seen. In contrast to that, the heart rate was negatively correlated to ADMA after 30 days of CIH whereas for CH no correlation was seen ( Table 2).

NADPH Activity and Oxidative Stress under Hypoxia.
CH increased NADPH subunit NOX4 mRNA expression ( < 0.01), whereas in CIH2x2, no difference from normoxia was observed (Figure 5(a)). The MDA concentration within the lung increased up to 2-fold, under both hypoxic conditions ( < 0.001) without significant differences between them ( NS) (Figure 5(b)).

Discussion
Our study has seven major findings: (1) Under both hypoxic conditions, rats developed moderate RVH. (2) eNOS expression increased in long-term CIH2x2 and CH, but the activation of sGC by NO did not. (3) The L-arginine concentration was unaffected despite the fact that arginase activity was enhanced under both hypoxic conditions. (4) ADMA concentrations in the lung increased in the longterm CIH2x2 and CH due to a hypoxia-induced reduction of DDAH activity in CH. (5) ADMA concentrations were positively correlated with systolic blood pressure in longterm CH. (6) Both hypoxic conditions led to a strong increase in biomarkers for oxidative stress. (7) Both CIH and CH exhibited similar effects, but they were lower or absent in the CIH group.
Hypoxic pulmonary vasoconstriction (HPV) occurs in response to hypoxia [23]. If this exposure is sustained, a subsequent vascular remodeling of pulmonary arteries develops, causing an increase in pulmonary vascular resistance and leading to RVH as part of hypoxia-induced pulmonary hypertension, resulting in premature death [24][25][26][27]. Consistent with these studies, we observed RVH in both hypoxic groups, although RVH was more moderate in CIH2x2, as previously described [14]. The molecular mechanisms involve  Figure 3: Comparison of (a) L-arginine levels (nmol/g protein) by mass spectrometry and (b) the L-arginine/ADMA ratio in the lungs of rats exposed to normoxia (NX), chronic intermittent hypoxia (CIH2x2), or chronic hypoxia (CH). Values are expressed as means and SE. * * * < 0.001; = 8 in each group. components both upstream of the endothelial NO pathway and mechanisms leading to oxidative stress downstream in the smooth muscle cells [28,29].
Therefore, vascular response to hypoxia is critical, and the endothelium plays a crucial role. Specifically, the endothelium is one major source of the vasodilator NO.
Our current study shows that eNOS mRNA expression is enhanced in hypoxia as expected. However, the increased mRNA expression of eNOS did not result in an elevation of NO-activated sGC-dependent activity as expressed as the pVASP/VASP ratio (an indirect measurement of NOactivated sGC-dependent activity) [30,31]. There are several possible explanations for this result.
First, a reduction in DDAH activity in the CH rat lung was observed, which resulted in an increased ADMA tissue concentration. Interestingly, the reduction in DDAH activity appears to be dependent only on DDAH2. Although we have recently demonstrated that DDAH1 is the isoform genetically associated with ADMA [32], the influence of DDAH2 on ADMA appears to increase in diseased states, which is consistent with our results. Subsequently, as ADMA is known to displace L-arginine (NOS substrate) from the binding site of NOS and thereby leads to competitive NOS inhibition, lower NO substrate availability could occur. This possibility is supported by a hypoxia-induced decrease detected in the L-arginine/ADMA ratio, an indicator of NOS substrate availability [20] and the correlation of ADMA and systolic blood pressure pointing to a systemic endothelial dysfunction.
Second, a hypoxia-induced increase in oxidative stress occurred, as evidenced by the enhanced NOX4 mRNA expression. NOX4 is mainly expressed in vascular smooth muscle cells (VSMCs), and NADPH oxidase activity in VSMCs is largely dependent on NOX4 mRNA expression [33]. Additionally, under hypoxic conditions, superoxide radicals are produced, enhancing lipid peroxidation. It is known that superoxide radical-mediated NO degradation attenuates its capacity to stimulate cGMP and the potential beneficial actions of NO on pulmonary arterial vasodilation [34].
Accordingly, in this study, both CIH and CH led to a strikingly increased MDA level along with increased NOX4 mRNA expression only in CH rats, suggesting a burst of superoxide radicals. These results are coincident with a diminished NO bioavailability and increased superoxide anions in rat pulmonary arteries observed by confocal microscopy [35]. Moreover, MDA might also inhibit DDAH activity, as previously demonstrated for 4-HNE (4-hydroxy-2-nonenal) [36]. This effect could also explain why our results indicate reduced DDAH activity.
L-Arginine is converted into L-ornithine by arginases, thereby diminishing its availability as a NOS substrate. Arginases have recently been reported to play important roles in the pathogenesis of various pulmonary disorders [37]. A steep increase in lung arginase activity is clearly demonstrated in our results, and this increase was not followed by an equal decrease in L-arginine. Interestingly, only arginase isoform I gene expression was significantly increased, with an increasing trend in arginase isoform II expression. This latter is not coincident with a previous report showing an increase in arginase II isoform in CIH, but using different methodology and measurement techniques [38]. Therefore, given the L-arginine/ADMA ratio results, we could postulate that although the L-arginine concentration exhibits almost no changes, the bioavailability of this substrate is imbalanced or impaired.
Our study has some limitations. We used RVH, as a generally accepted indirect marker of increased pressure within the pulmonary circulation. Due to the increases in NOX4 mRNA expression and MDA formation, we inferred a rise in ROS, but we did not measure ROS directly. Instead of measuring NO directly, we used an established indirect measurement by determining the pVASP/VASP ratio. Furthermore, all measurements were made in whole-lung homogenates, which may not represent what specifically occurs in the pulmonary endothelium. Nevertheless, most of our results are consistent with those that have been separately described, whereas we have attempted to describe a wider and more integrated picture of these changes in the lung under long-term CIH exposure.  Figure 5: Comparison of (a) mRNA expression levels of NOX4 determined by qRT-PCR and (b) MDA concentrations (nmol/L) determined by spectrophotometry of lung tissue from rats exposed to normoxia (NX), chronic intermittent hypoxia (CIH2x2), or chronic sustained hypoxia (CH). Values are expressed as means and SE. * < 0.05; * * < 0.01; * * * < 0.001; = 8 in each group.

Conclusions
This observational study demonstrated that long-term CIH and CH in rats led to an imbalance in the NO pathway, where ADMA and oxidative stress could be responsible for impaired pulmonary vascular function by inhibiting or impairing signal transduction (endothelial dysfunction). Both hypoxic conditions resulted in similar effects on NO signaling in the lungs, but CIH had a more limited or no effect on certain parameters. The clinical relevance of ADMA and DDAH and their roles in the development of hypoxic pulmonary hypertension in CIH remain to be determined, opening a new field of research as biomarkers or drug targets in either HAPH or other pulmonary hypoxic conditions.