Abstract
Tetrahydrobiopterin (BH4) is an essential cofactor for nitric oxide synthases (NOS). This study investigated the effect of increasing BH4 levels on hypoxia-induced pulmonary vasoconstriction (HPV).
Sprague Dawley rats and hph-1 (BH4 deficient) mice were given BH4 before and during HPV in an isolated perfused lung preparation. BH4 inhibited HPV in a concentration-dependent manner and increased NO metabolites in the perfusate. Bradykinin-induced reductions in HPV were blunted in hph-1 mice and pre-administration of BH4 restored the response. The effect of BH4 was attenuated by l-NAME (NOS inhibitor), PTIO (NO scavenger), and catalase (H2O2 catalyser) administered prior to HPV but enhanced by MnTMPyP (superoxide dismutase mimetic). The effect of BH4 on HPV was partially recapitulated by NH4, a stereoisomer that shares antioxidant properties with BH4 but is not a NOS cofactor.
The bioavailability of BH4 is an important determinant of the pulmonary vascular response to hypoxia. Its effects are mediated via nitric oxide, hydrogen peroxide and its antioxidant properties, and are attenuated by oxidant stress. Pharmacological administration of BH4 may have therapeutic potential in pulmonary hypertension.
- Animal models
- hypoxia
- pulmonary vasoconstriction
- pulmonary hypertension
- nitric oxide
- superoxide
- tetrahydrobiopterin
Nitric oxide synthases (NOS), particularly endothelial NOS (eNOS), play a major role in maintaining normal pulmonary vascular tone and structure through the production of nitric oxide (NO) from l-arginine 1–3. Critical to NO synthesis by NOS is the bioavailability of tetrahydrobiopterin (BH4). When BH4 is deficient through inadequate synthesis or oxidation, NOS becomes uncoupled and generates superoxide instead of NO 4.
The hph-1 mouse, generated by N-ethyl-N-nitrosourea (ENU) mutagenesis, has low tissue BH4 levels because of constitutively reduced expression of GTP cyclohydrolase-1, the rate-limiting enzyme in BH4 biosynthesis 5. Congenital deficiency of BH4 in the hph-1 mouse leads to pulmonary hypertension, distal pulmonary vessel muscularisation and right ventricular hypertrophy 6, 7. The pulmonary vascular pathology in hph-1 mice is more marked than that reported in NOS-deficient mice and suggests that loss of NO production alone is not the sole reason for the vascular pathology 1, 8. In addition to reduced NO synthesis, and consistent with uncoupling of NOS, biochemical analysis of tissue homogenates from hph-1 mice shows increased local pulmonary vascular superoxide production 7. Vascular superoxide production has a number of important actions on vascular tissue, such as scavenging of NO, peroxynitrite formation and modulation of redox-sensitive signalling pathways 9, which may adversely affect vascular structure.
There is increasing evidence of oxidative stress in pulmonary hypertension 10, 11, including reports of nitrotyrosine formation (a marker of peroxynitrite production) in the lungs of patients with the disease 12. Superoxide and peroxynitrite oxidise BH4, and enhanced oxidative degradation is thought to be a major cause of reduced BH4 bioavailability and eNOS uncoupling in endothelial dysfunction states 13–15. As a result, there is interest in using exogenous BH4 therapeutically in pulmonary hypertension to restore tissue levels and promote NO production from NOS. Recent studies suggest that BH4 may also act independently of NO, generating hydrogen peroxide, another molecule with vasorelaxant properties 16, 17.
The purpose of this study was to investigate the pharmacological effects of BH4 on pulmonary vascular tone, specifically hypoxia-induced vasoconstriction (HPV) in an in situ isolated perfused lung preparation. Employing the BH4 deficient hph-1 mouse, we explored further the underlying mechanism of BH4 in balancing NO and superoxide production in pulmonary vasculature.
METHODS
Animals
Male Sprague-Dawley rats (250–360 g) from Charles River (Margate, UK) and hph-1 mice aged 3–5 months generated by ENU mutagenesis were used for experiments. Wild-type (WT) animals, hph-1 heterozygous (+/-) and hph-1 homozygous (hph-1) littermates on a C57BL/6 background were obtained by interbreeding hph-1 heterozygotes 5. All studies were conducted in accordance with UK Home Office Animals (Scientific Procedures) Act 1986.
Reagents
6R-BH4 (sapropterin dihydochloride or BH4) was supplied by BioMarin Pharmaceuticals, Inc. (Novato, CA, USA). U46619 (9,11-dideoxy-11α,9α-epoxymethanoprostaglandin F2α), L-NAME (Nω-nitro-l-arginine methyl ester hydrochloride), Catalase, MnTMPyP (Mn(III) tetrakis(1-methyl-4-pyridyl) porphyrin pentachloride), PTIO (2-phenyl-4,4,5,5-tetramethylimidazoline-1-oxyl 3-oxide), bradykinin (BK) and chemical components of the perfused solutions were obtained from Sigma-Aldrich (Poole, UK). NH4 (6R, S-5,6,7,8-tetrahydro-d-neopterin) was from Schircks Laboratories (Jona, Switzerland). All drugs were prepared fresh every day before experiments. 10 μg·uL−1 BH4 or NH4 were dissolved in HEPES buffered saline (composition: NaCl 137 mM, KCl 4.0 mM, CaCl2 1.8 mM, MgCl2 1 mM, HEPES 10 mM, glucose 10 mM; pH adjusted to 7.4 with NaOH).
Isolated perfused rat lung preparation
The pulmonary vascular response to BH4 was studied using the isolated rat lung preparation previously described 18. Briefly rats were anaesthetised with Hypnorm (fentany-fluanisone). The lungs were left in situ after the trachea had been cannulated, ventilated with air (21% O2, 5% CO2 balanced N2) at constant end expiration pressure of 4 mmHg with 60 breaths·min−1; and perfused with a mixture (1:1) of DMEM (Dubelcco's Modified Eagles Medium) with 4% Ficoll, and blood withdrawn from the experimental animal (pH 7.4) at a flow rate of 18 mL·min−1 using a non-pulsatile pump (Masterflex model 7519; Cole-Parmer Instrument Co. Ltd, London, UK). Pulmonary artery pressure (Ppa) was measured using a pressure transducer connected to PowerLab Data Acquisition system (ADInstruments Limited, Chalgrove, UK). HPV was produced by ventilating the lung with 2% O2, 5% CO2 and 93% N2 for 15 min. The ventilation was then returned to air for 10 min to resume the PAP baseline before the next hypoxic challenge. Successive hypoxic challenges (usually 3) were repeated until the HPV response was consistent before vehicle or BH4 treatment was given to examine the effect on HPV.
BH4
BH4 (final perfusate concentration 0.3 μg·mL−1) was added to the perfusate ∼4 min before the hypoxic challenge. Once the pressor response had returned to baseline, the hypoxic challenge was repeated with or without a second aliquot of BH4 (calculated to produce perfusate concentrations of 0.3 or 1 μg·mL−1) added to the perfusate (fig. 1). Perfusate samples (100 μL) were collected after passage through the lungs 1 min prior to and 4 min after BH4 administration and before the hypoxic challenge for measurement of nitrate/nitrite (NOx) concentration. The effect of BH4 was calculated as the per cent change in HPV response and NOx levels after BH4 administration compared to the HPV response and NOx levels before BH4 administration in the same animal.
Isolated perfused mouse lung preparation
Animals were anaesthetised and the lungs ventilated with air (volume controlled ventilation, 230 μL tidal volume, 80 breaths·min−1 and 2 cmH2O positive end-expiratory pressure) via the trachea. The lungs were perfused through the main pulmonary artery accessed via the right ventricle and perfusate collected from the left atrium for recycling in a closed circuit (type 839; Hugo Sachs Electronik-Harvard Apparatus GmbH, March-Huggstetten, Germany). Using a nonperistaltic pump, the perfusate (DMEM with 4% Ficol, HEPES-buffered, pH 7.4) was delivered at an initial flow of 0.2 mL·min−1, increasing in increments after 10 min to 2 mL·min−1. The reservoir volume was set at 10 mL and pH monitored by the pH sensor placed in the reservoir. Ppa was measured using a pressure transducer connected to the inflow catheter. Left atrial pressure was measured using a pressure transducer connected to the outflow catheter. Pressure tracings were collected and analysed using PowerLab Data Acquisition system. U46619 (thromboxane analogue; 5 μM) was added to reservoir to induce basal pulmonary vasoconstriction. 2 min after U46619 administration, a stable Ppa was achieved and recording was continued for a further 6 min before HPV was induced by ventilating the lung with 2% O2, 5% CO2 and 93% N2. Ventilation was returned to air until Ppa returned to baseline (fig. 2a).
BH4 and NH4
Vehicle (fig. 2a), BH4 (1, 10 and 100 μg·mL−1; fig. 2b) or NH4 (100 μg·mL−1) was added to the perfusate during stable HPV. The effect of BH4 and NH4 was calculated as the Ppa reduction (δPpa) 3 min after drug administration expressed as a percentage of the maximum HPV response (HPVmax) in the same hypoxia challenge.
Bradykinin
BK (10 ng·mL−1) was administered during stable HPV in WT, +/- and hph-1 mice (fig. 3). Futhermore, in separate sets of hph-1 mice, BH4 (100 μg·mL−1) was added to reservoir 3 min before 10 ng·mL−1 of BK. The response to BK was calculated as the maximum reduction in Ppa (usually within 1 min of BK administration) expressed as a percentage of the HPVmax in the same hypoxia challenge.
Inhibition of NO and superoxide
In separate experiments, L-NAME (100 μM), catalase (200 U·mL−1), MnTMPyP (1 μM) and PTIO (1 μM) was added to the perfusate 3 min before inducing HPV, followed by BH4 (100 μg·mL−1) administration during stable HPV (fig. 4). BH4 effect was calculated as Ppa reduction (δPpa) 3 min after BH4 administration expressed as a percentage of the HPVmax in the same hypoxia challenge.
Measurements of nitrate/nitrite (NOx)
Perfusate samples were centrifuged at 1000×g for 4 min. The supernatant was immediately removed and stored at -20°C until analysis. NOx concentration was measured in duplicate using an ozone-based chemiluminescence method 19. Briefly, prior to analysis the samples were treated with a 1:1 volume of cold ethanol and centrifuged at 14,000×g for 5 min. NOx levels were measured by injecting 50 μL of the supernatant in a glass purge vessel containing vanadium(III) in 1 N hydrochloric acid at 90°C, which reduced NOx to NO gas. A nitrogen stream was bubbled through the purge vessel containing vanadium(III), first through 1 N NaOH, and then into a NO analyser (Sievers Model 280 NO Analyser; Sievers Instruments, Boulder, CO, USA), which detected NO released from NOx by chemiluminescent detection. Results are presented as the percentage of the NOx levels after BH4 administration in the perfusate compared with the levels prior to administration.
Statistical analysis
In all experiments, results are expressed as mean±sem; n equals the number of mice or rats per experiment. Statistical analysis of the data was performed by using unpaired t-test when appropriate, nonparametric test with Man–Whitney modification, or one-way ANOVA when applied to multiple group comparisons. A value of p<0.05 was considered statistically significant.
RESULTS
BH4 attenuates HPV in the isolated perfused rat lung
The effect of increasing concentration of BH4 on HPV and NOx levels was first examined in the isolated perfused rat lung. Consistent HPV responses were established (11.6±2.8 mmHg, n = 15) in each lung preparation before exposure to exogenous BH4. BH4 at 1 μg·mL−1, but not 0.3 μg·mL−1, attenuated the pressor response to hypoxia (-31.9%, HPV 8.5±2.7 mmHg versus 12.4±2.8 mmHg with or without BH4 1 μg·mL−1, respectively; p<0.01, n = 5; fig. 1). Perfusate NOx levels were increased at both BH4 concentrations (27±6% and 47±15%, respectively; fig. 1b).
BH4 attenuates HPV in the isolated perfused mouse lung
To explore the effect of restoring BH4 in the context of impaired endogenous synthesis of the cofactor, studies were conducted in WT, hph-1 heterozygous (+/-) and hph-1 homozygous (hph-1) mice. Hypoxia induced similar maximum rises in Ppa in each of the mouse genotypes: 17.1±1.3 mmHg in WT, 14.7±1.1 mmHg in heterozygous and 16.1±1.0 mmHg in hph-1 (n = 12–15). Compared to vehicle (fig. 2a), BH4 1, 10 and 100 μg·mL−1 administered during stable HPV induced concentration-dependent reductions in Ppa (fig. 2b). Data from all three genotypes are summarised in figure 2c. BH4 produced greater falls in WT than in hph-1 mice at doses of 10 and 100 μg·mL−1 (fig. 2c).
BH4 restores the response to BK in hph-1 mouse
In order to examine endothelial function in hph-1 mice, the vasorelaxant response to BK was investigated in the isolated perfused lung (fig. 3). BK administration to achieve a perfusate concentration of 10 ng·mL−1 during stable HPV reduced Ppa in WT (-12.1±1.8%) and heterozygous (-9.7±1.2%) mice, but had no detectable effect in hph-1 mice (0±1.7%) (table 1). BH4 (100 μg·mL−1) given prior to BK restored the response to BK in hph-1 mice (-11.7±2.4%).
The effect of L-NAME, catalase, MnTMPyP and PTIO on the response to BH4 in mice
To further dissect the mechanism by which BH4 regulates pulmonary vascular tone, the roles of NO and reactive oxygen species were investigated. In this set of experiments, the maximum HPV response before treatment was in the range of 8–10 mmHg in all the three genotypes (see initial HPV, table 2). L-NAME, catalase, MnTMPyP and PTIO had minor effects on the baseline normoxic Ppa (-0.5 to 1.1±0.5 mmHg) and these were similar among the three genotypes (see table 1, online supplementary material).
L-NAME and BH4
NOS inhibitor L-NAME (100 μM) administration enhanced the HPV response in all three genotypes (table 2), and reduced the effect of BH4 (100 μg·mL−1) in WT (-20±3% versus -11±1.7%, in the absence and presence of L-NAME, respectively; p<0.05; fig. 4a).
Catalase and BH4
Pretreatment with catalase increased the HPV response significantly in heterozygous and hph-1 mice, but not in WT mice (table 2). It attenuated the effect of BH4 in WT mice significantly (-20±3% versus −7.8±2%; p<0.01; fig. 4b), and to a lesser degree in heterozygous and hph-1 mice.
MnTMPyP and BH4
The intracellular superoxide dismutase mimetic MnTMPyP did not affect HPV (table 2). It enhanced the vasodepressor effect of BH4 significantly in all the genotypes and almost returned Ppa to baseline levels. The net change was greater in hph-1 animals (-32.5±3% versus -7.9±2.9%, in the presence and absence of MnTMPyP, respectively; p<0.01) than in WT mice (-34.8±2.8% with MnTMPyP versus -20±3% without; p<0.01; fig. 4c).
PTIO and BH4
Pretreatment with an NO scavenger, PTIO, increased HPV significantly in WT mice (table 2) and attenuated the subsequent BH4-induced reduction in Ppa (-20.5±3% versus -7.3±0.9%; p<0.01; fig. 4d).
The effect of NH4 on HPV response
To evaluate the antioxidant action of BH4, its stereoisomer NH4 was used for comparison in the isolated perfused lung WT mouse lung. NH4 100 μg·mL−1 produced a smaller fall in Ppa (-6.4±2.3%) than BH4 100 μg·mL−1 (-31%; p<0.01; fig. 4e).
DISCUSSION
In the present study, exogenous BH4 attenuated the acute pulmonary vascular pressor response to hypoxia in a concentration-dependent manner. There appear to be three mediators of the effect: increased NO production, a direct antioxidant action and H2O2. Evidence that the response to BH4 is mediated in part by coupling NOS to increase NO production and improve pulmonary endothelial function comes from: 1) the associated increase in NO (NOx levels) in the perfusate; 2) the observation that BH4 restored the response to BK, an endothelium-dependent vasodilator, in the hph-1 mouse; and 3) that the effect was inhibited by pretreatment of L-NAME, a NOS inhibitor, and PTIO, a NO scavenger. A direct antioxidant effect from BH4 is inferred from the response to NH4, a BH4 stereoisomer that has antioxidant properties but which is not a NOS cofactor. A role for H2O2 is supported by the observation that catalase, which promotes the decomposition of H2O2 to oxygen and water, attenuated the vasorelaxant effect of BH4. Conversely, the effect of BH4 is reduced by increased superoxide production, as seen in the hph-1 mouse. Prior treatment with MnTMPyP, an intracellular superoxide dismutase mimetic, restored the response to BH4 in hph-1 mice and reversed HPV to near baseline Ppa levels.
HPV is a characteristic of the pulmonary circulation, in which resistance vessels constrict in response to alveolar hypoxia in order to maintain perfusion/ventilation matching 20. NO synthesis contributes to this vasoregulatory mechanism 21. For example, reduced NO bioavailability through inhibition of NOS with L-NAME or NO scavenging with PTIO augments HPV. Conversely, increased NO synthesis attenuates HPV. Attenuation of HPV with BH4 in the isolated perfused rat lung is consistent with the known biochemistry of BH4 as an essential cofactor for NOS dimerisation, stability and electron transfer during arginine oxidation 22. Interestingly, in a previous experiment we showed that hph-1 mice, which are relatively deficient in BH4, were more sensitive to hypoxia 7. In this study, we sought to maximally stress the pulmonary vasculature with the vasoconstrictor U46619 as well as hypoxia, and achieved similar maximal responses to HPV in all three genotypes (WT, and heterozygous and homozygous hph-1 mice).
L-NAME and PTIO, at doses described previously to cause full inhibition of the NO system 23, 24, only partially blocked the effects of BH4 on Ppa during the stable phase of HPV in WT mice, indicating that “NOS coupling” is not the sole mechanism by which BH4 modulates HPV. In addition to uncoupled NOS, the generation of superoxide and free radicals from NADPH oxidase and the mitochondrial electron transport system has been implicated in HPV. BH4 is a potent reducing agent 25 and has been shown to prevent ischaemia reperfusion injury through its antioxidant properties rather than acting as a NOS cofactor 26. To address this contribution to its effect on HPV, we examined the effect of NH4, an equipotent antioxidant analogue of BH4 with no cofactor activity. At the same concentration, the reduction in Ppa produced by NH4 was ∼30% of that produced by BH4. Whether the antioxidant effect of BH4 is related mainly to reduction of superoxide or the nitrosyl radical (peroxynitrite), and how these reactions might influence the total effect of BH4 on pulmonary pressure, is unknown and needs further investigation.
H2O2 is also a mediator of the effect of BH4 on Ppa. The NOS family is a major source of vascular H2O2, derived from superoxide by the action of various superoxide dismutases. H2O2 has been proposed to be an endothelium-derived hyperpolarising factor (EDHF). Recent data suggest that H2O2 may be an important regulator of vascular homeostasis in small vessels 27. EDHF-mediated relaxation and hyperpolarisation of mesenteric arteries in response to acetylcholine is reduced progressively in eNOS, n/eNOS, and n/i/eNOS deficient mice as the number of disrupted NOS genes increases 27. The vascular effects of H2O2 are unmasked by BH4 deficiency, where superoxide production is increased and NO synthesis reduced 7. Catalase inhibits endothelium-dependent relaxation in the aorta of hph-1, but not WT mice 28 and significantly augmented HPV in heterozygous and hph-1, but not in WT mice in our study.
Paradoxically, increasing BH4 levels also increases H2O2. It is now accepted that the NOS family produce superoxide under normal conditions, not just when uncoupled. BH4 enables both superoxide and NO production from NOS and increases H2O2 production 24. It may also generate H2O2 directly following oxidation by molecular oxygen in aqueous solution 17. Catalase inhibited BH4 induced vasorelaxation in mouse cerebral arteries 24 and rabbit iliac arteries, and inhibited the vasodilatory effect of BH4 on HPV in WT mice in our study. The mechanism by which H2O2 produces its vasorelaxant effects includes activation of soluble guanylate cyclase 29, stimulating the production of prostanoids 30 and sarcoplasmic endoplasmic reticulum Ca2+ ATPase 16.
Conversely, oxidation of BH4 limits its bioavailability and so efficacy. Peroxynitrite, the product of superoxide and NO, oxidises BH4 to BH2 with 10 times greater potency than it reacts with other cellular oxidants, such as ascorbic acid and glutathione 31. Oxidised BH2 competes with BH4 for NOS but does not act as a cofactor, thereby fostering the uncoupling of NOS 32 and further exacerbating NOS dysfunction. Excess superoxide production previously reported in hph-1 mice 7 and oxidation of BH4 likely impaired its effect on HPV in heterozygous and hph-1 mice. The addition of MnTMPyP to the perfusate prior to the hypoxic challenge enhanced the vasodilatory effect of BH4 in all WT, heterozygous and hph-1 mice. These data are consistent with the observations of Farrow et al. 33, that recombinant human superoxide dismutase treatment increased eNOS activity and expression, elevated BH4 levels, decreased oxidative stress and induced pulmonary vasodilation in the persistent pulmonary hypertension model of neonatal lambs.
A limitation of this study is the lack of direct biochemical measurement of superoxide and H2O2 levels coincident with the biochemical manipulations. There are sufficient data from published work to support the interpretations made 24, 28, 33–35. The absence of these measurements, however, does not permit comment on the extent to which pharmacological manipulation of each contributes to the effects observed.
In summary, BH4 participates in the regulation of the pulmonary vascular response to hypoxia through the generation of NO and superoxide, and consequently the formation of hydrogen peroxide and peroxynitrite (fig. 5). The net effect of pharmacological supplementation with BH4 is to inhibit the pressor response. The effect is attenuated in the presence of oxidative stress and combining BH4 with an antioxidant strategy (for example, a superoxide dismutase mimetic) augments the efficacy of BH4 treatment. Pharmacological administration of BH4 may have therapeutic potential in pulmonary hypertension and warrants further investigation.
Acknowledgments
We thank A. Hobbs (University College London, London, UK) for nitrate/nitrite measurement.
Footnotes
Support Statement
This work is supported by the British Heart Foundation (London, UK), BioMarin Pharmaceuticals, Inc. (Novato, CA, USA) and PULMOTENSION (European Commission under the 6th Framework Programme, contract numbers LSHM-CT-2005-018725, Brussels, Belgium).
Statement of Interest
A statement of interest for M.R. Wilkins and for the study itself can be found at www.erj.ersjournals.com/misc/statements.dtl
This article has supplementary material available from www.erj.ersjournals.com
- Received November 26, 2009.
- Accepted February 28, 2010.
- ©ERS 2010