|Home | About | Journals | Submit | Contact Us | Français|
Rationale: Neonatal chronic lung disease (CLD), caused by prolonged mechanical ventilation (MV) with O2-rich gas, is the most common cause of long-term hospitalization and recurrent respiratory illness in extremely premature infants. Recurrent episodes of hypoxemia and associated ventilator adjustments often lead to worsening CLD. The mechanism that causes these hypoxemic episodes is unknown. Hypoxic pulmonary vasoconstriction (HPV), which is partially controlled by O2-sensitive voltage-gated potassium (Kv) channels, is an important adaptive response to local hypoxia that helps to match perfusion and ventilation in the lung.
Objectives: To test the hypothesis that chronic lung injury (CLI) impairs HPV.
Methods: We studied preterm lambs that had MV with O2-rich gas for 3 weeks and newborn rats that breathed 95%-O2 for 2 weeks, both of which resulted in airspace enlargement and pulmonary vascular changes consistent with CLD.
Measurements and Main Results: HPV was attenuated in preterm lambs with CLI after 2 weeks of MV and in newborn rats with CLI after 2 weeks of hyperoxia. HPV and constriction to the Kv1.x-specific inhibitor, correolide, were preferentially blunted in excised distal pulmonary arteries (dPAs) from hyperoxic rats, whose dPAs exhibited decreased Kv1.5 and Kv2.1 mRNA and K+ current. Intrapulmonary gene transfer of Kv1.5, encoding the ion channel that is thought to trigger HPV, increased O2-sensitive K+ current in cultured smooth muscle cells from rat dPAs, and restored HPV in hyperoxic rats.
Conclusions: Reduced expression/activity of O2-sensitive Kv channels in dPAs contributes to blunted HPV observed in neonatal CLD.
Recurrent episodes of hypoxemia and associated ventilator adjustments often lead to worsening of neonatal chronic lung disease (CLD). The mechanism that causes these hypoxemic episodes is unknown.
Blunted hypoxic pulmonary vasoconstriction (HPV) due to reduced expression/activity of O2-sensitive Kv channels may explain hypoxemic episodes. Restoring Kv channel activity restores HPV. Modulation of Kv channels may improve management of neonatal CLD.
Premature infants account for more than 12% of all U.S. births (National Institute of Medicine report, July 13, 2006, http://www.iom.edu/CMS/3740/25471/35813.aspx). Recent advances in obstetrical management and neonatal intensive care have increased survival of extremely premature infants (< 1,000 g). These infants are prone to respiratory failure, for which they receive mechanical ventilation (MV) with O2-rich gas. Prolonged exposure to cyclic stretch and high concentrations of inspired O2 often leads to a chronic form of lung disease (CLD) that remains the leading cause of long-term hospitalization and recurrent respiratory illness in infants born at less than 28 weeks of gestation (1).
Prominent histologic features of CLD include arrested alveolar growth and decreased capillary density (2–4). Extremely premature infants with this condition frequently experience severe episodes of hypoxemia. The mechanism accounting for these hypoxemic events in neonates is poorly understood, but more likely related to either acute changes in airway resistance (5, 6) or increased pulmonary vasoconstriction. It is unknown whether impaired ventilation in CLD may lead to mismatching of ventilation and perfusion, which could contribute to episodic hypoxemia and cyanosis.
Hypoxic pulmonary vasoconstriction (HPV) is a unique and important physiological response that facilitates matching of ventilation and perfusion in the lung. In segmental hypoxia, as occurs in CLD, constriction of distal pulmonary arteries (dPAs), which regulate vascular resistance, diverts blood from poorly ventilated areas to well-ventilated areas of the lung, thereby optimizing oxygenation, usually without increasing pulmonary vascular resistance (PVR) (7).
Although the mechanisms that control HPV have not been completely elucidated (8–10), it appears that hypoxia initiates HPV through a redox mechanism (11, 12) that inhibits K+ channels in pulmonary artery smooth muscle cells (PASMC) (13). O2-sensitive voltage-gated K+ channels (Kv) play an important role in regulating HPV through modulation of cell membrane potential (EM) (14). During HPV, hypoxia closes Kv channels. This leads to membrane depolarization, opening of voltage-gated L-type calcium (Ca2+) channels, and increased influx of extracellular Ca2+, thereby causing pulmonary vasoconstriction. Inhibition of O2-sensitive Kv channels Kv1.5 and Kv2.1 contribute to initiation of HPV (15, 16).
Thus, we hypothesize that HPV is impaired in neonatal CLD and that the basis for impaired HPV is reduced expression and function of O2-sensitive Kv channels.
Some of the results of these studies have been previously reported in the form of an abstract (17).
Detailed methods are provided in the online supplement.
All procedures were approved by the Institutional Animal Care and Use Committee at the University of Utah in Salt Lake City (preterm lambs) and the University of Alberta (newborn rats).
Preterm lambs were mechanically ventilated for 3 weeks and chronically catheterized as previously described (18–21). PVR was measured during a 2-hour steady-state baseline period, followed by a 2-hour steady-state period of hypoxemia, induced by lowering the FiO2 so that PaO2 decreased to less than 40 mm Hg.
Sprague-Dawley rat pups were exposed to normoxia (21% O2, control group) or hyperoxia (95% O2, O2-chronic lung injury [CLI] group) from birth to Postnatal Day (P)14 in sealed Plexiglas chambers with continuous O2 monitoring (22–25). After P14, pups were killed and lungs processed for various analyses.
Exposure of newborn rats to hyperoxia from birth to P14 during the critical period of alveolar development impairs alveolar growth (larger but fewer alveoli) and is associated with right ventricular hypertrophy, an indirect marker of pulmonary hypertension. These results have been previously reported (26).
Pulmonary artery acceleration time (PAAT) (Vevo 770B; VisualSonics, Inc., Toronto, ON, Canada) was measured from the onset of the pulmonary flow to its peak as previously described (26). For the HPV challenge, animals were briefly (~ 30 s) exposed to hypoxia (2.5% O2) until a saturation of approximately 60% (MouseOx; Starr Life Sciences Corp., Oakmont, PA) was obtained.
Third-generation PAs (diameter < 100 μm, length = 2 mm, here referred to as dPAs) were mounted in a wire myograph (MyoDaq; Danish Myo Technology, Aarhus, Denmark) and bathed in Krebs-Henseleit buffer bubbled with 21% O2–5% CO2–balance N2 (normoxia, Po2 ~ 120 mm Hg) or 5% O2–5% CO2–balance N2 (hypoxia, Po2 ~ 50 mm Hg) maintained at 37°C, pH 7.35 to 7.45 (27, 28). Isometric changes in response to phenylephrine (10−5 mol/L), HPV, the nonspecific Kv channel inhibitor 4-aminopyridine (10 mM), a specific Kv1.x channel inhibitor correolide (100 μM; gift from John Obenchain, Merck and Co., Inc., Rahway, NJ), the calcium-sensitive K+ channel (BKCa) inhibitor iberiotoxin (10−7 M), and the ATP-dependent K+ channel (KATP) inhibitor glyburide (5 × 10−6 M) and to 80 mmol/L KCl were compared between control and O2-induced CLI PAs.
Quantitative real-time polymerase chain reaction (qRT-PCR) was performed on isolated dPAs from newborn rats as previously described (29, 30), using purchased primers for human Kv 1.5, rat Kv 1.5, and Kv 2.1, BKCa (Applied Biosystems, Foster City, CA). Levels of mRNA were normalized to a stable control gene (β2 microglobulin [β2MG]) and expressed as 2ΔΔCt (Ct is cycle time) (29, 30).
A 2.1-kb cDNA fragment of Kv1.5 was obtained by reverse transcription of mRNA derived from the proximal PA of a cardiac transplant donor (29). The virus was given intratracheally to CLI rats at P10 (26).
PASMC were freshly isolated from dPAs of control, O2-induced CLI, Ad5-Kv1.5–treated, and Ad5-green fluorescent protein (GFP)–treated rats. Their response to hypoxia was studied using a whole cell patch clamping technique. PASMC were voltage-clamped at −70 mV, and currents were evoked from −70 to +50 mV by steps of 200 milliseconds, as previously described (29).
Data are expressed as mean ± SE, except where stated otherwise. Statistical analysis was performed using unpaired Student's t test or ANOVA, post hoc test (least significant difference) and Mann-Whitney as appropriate. Values were considered significant with P < 0.05.
Chronically mechanically ventilated preterm lambs had persistent baseline elevation of PVR for the entire 3 weeks. PVR increased significantly during hypoxia at the end of Week 1, but HPV was blunted during Weeks 2 and 3 (Figure 1).
In vivo echo-Doppler studies showed that PAAT was significantly decreased in newborn rats with O2-induced CLI as compared with normoxic controls (Figure 2A). During the hypoxic challenge, PAAT decreased in control pups, but not in pups with O2-induced CLI (Figure 2A).
In vitro, dPAs constricted in response to hypoxia while dPAs from newborn rats with O2-induced CLI failed to constrict in response to hypoxia (Figure 2B).
The contractile response to the α-agonist phenylephrine (10 mmol/L) was decreased significantly in O2-induced CLI as compared with controls (Figure 3A). KCl constriction (Figure 3B) and the response to the BKCa and KATP channel blockers iberiotoxin and glyburide were similar between groups (Figure 3C). There was no difference in constriction in response to the Kv channel inhibitor 4-AP (10 mmol/L) (Figure 3D). In contrast, the vascular reactivity in O2-induced CLI arteries to the specific Kv1.x inhibitor correolide was significantly decreased as compared with normoxic control arteries (Figure 3E).
Blunted HPV in newborn rats exposed to O2 was associated with a decrease in mRNA expression of Kv1.5 and Kv2.1 in dPAs as compared with controls (Figure 4A). There was no difference in BKCa mRNA expression between groups (Figure 4B).
In vivo, echo-Doppler showed that PAAT was significantly decreased in newborn rats with O2-induced CLI, Ad5-Kv O2-induced CLI and Ad5-GFP O2-induced CLI as compared with normoxic controls (Figure 5). During the hypoxic challenge, PAAT decreased only in Ad5-Kv O2-induced CLI and control pups, but not in pups with O2-induced CLI or Ad5-GFP O2-induced CLI (Figure 5).
Intratracheal administration of Ad5-Kv1.5 increased Kv1.5 expression in PAs of O2-induced CLI as shown by immunofluorescence imaging (Figure 6A) and qRT-PCR. qRT-PCR analysis using species-specific Kv primer revealed that human Kv1.5 was only expressed in Ad5-Kv1.5 transfected PAs (Figure 6B). In dPAs rings, Ad5-Kv1.5 gene transfer restored HPV as compared with Ad5-GFP dPAs and nontransfected O2-induced CLI dPAs (Figure 6C).
Consistent with our in vivo and in vitro studies, whole-cell K+ current (IK) was significantly decreased in PASMC from experimental O2-induced CLI compared with controls (Figures 7A and 7B). There was a loss of the outward rectifying portion of IK (evident in a comparison of Figures 7A versus 7D as less total current and less upward deflection from the linear, ohmic current between voltage steps +10 to +70 mV). In control PASMC, IK was inhibited by hypoxia (Figure 7A), whereas IK in O2-induced CLI remained unaffected (Figure 7B). CLI Kv1.5 gene therapy restored not only the net current, but specifically the outward rectifying portion of the current, so that Figures 7A and 7C look similar. This (together with the qRT data in Figure 4) indicates the loss of K+ current with CLI is largely due to loss of Kv current. Conversely, IK was low in PASMC from Ad5-GFP–transfected animals (Figure 7D).
Many studies of CLI have focused on changes in lung mechanics and airway function (31–35) to explain hypoxic episodes in premature infants. However, little attention has been paid to the possible mismatch of ventilation and perfusion that may develop in CLD. In this study, we show for the first time in vivo and in vitro evidence that HPV is impaired in two experimental models, created in different species and by different approaches, that mimic human CLD. In addition, impaired HPV in newborn rats was associated with decreased expression of the O2-sensitive Kv1.5 channel. And finally, HPV was restored after adenovirus-associated gene transfer of Kv1.5. These findings support the notion that blunted HPV may contribute to the impaired gas exchange seen in CLD, and that therapeutic targeting of O2-sensitive Kv channels in the pulmonary circulation might restore HPV and thereby improve arterial oxygenation in lung vascular diseases associated with impaired O2 responsiveness.
HPV is a unique and important physiologic response that facilitates matching of ventilation and perfusion in the lung. In segmental hypoxia the blood is diverted from poorly ventilated areas to well-ventilated segments of the lung, optimizing Po2 without increasing systemic vascular resistance (14). Hypoxia depolarizes the PASMC membrane and causes an increase in intracellular Ca2+. This is related to inhibition of O2-sensitive Kv channels, particularly Kv1.5 (15, 16). Decreased HPV in chronic hypoxia results from loss of Kv1.5 and 2.1 (29, 36), and enhanced expression of Kv1.5 by adenovirus-mediated gene transfer restores HPV (29). Patients with pulmonary hypertension have decreased mRNA levels of the O2-sensitive Kv1.5 channel (37), suggesting an etiologic role for O2-sensitive Kv channels in the development of this disease.
During development, maturational changes in K+ channel expression account for differences in O2 constriction. In the fetus, the resting membrane potential in PASMC is controlled by large-conductance BKCa, the predominant O2-sensitive channel (38). After birth, there is a shift to O2-sensitive Kv channels, presumably enhancing the capacity for HPV, without increasing pulmonary vascular resistance (38). Whether there is a delay in the maturational shift from BKCa to O2-sensitive Kv channels in experimentally induced BPD is unknown. O2 toxicity and/or chronic ventilation, the two main causes contributing to lung injury in premature infants, may impede Kv channel expression/function in the pulmonary vasculature. More recently, the deleterious effect of even brief (30 min) hyperoxia (decreased vasodilation to nitric oxide and acetylcholine) on the pulmonary circulation was reported (39, 40). Interestingly, the authors also show that lambs previously exposed for a brief period (30 min) to 100% O2 did not have enhanced pulmonary vasoconstriction in response to subsequent hypoxic ventilation. Likewise, pulmonary vascular dysfunction, including pulmonary hypertension and altered response to inhaled nitric oxide in CLD, has been reported previously in preterm lambs (18–20, 41). Conversely, there is no information about the effects of hyperoxia on the regulation of Kv channels.
It is interesting that the PA and the ductus arteriosus (DA), although adjacent, behave in such opposite ways in response to O2 and yet share a common pathway for O2 sensing (10). dPAs exhibit HPV for optimizing gas exchange in the lung. In contrast, the DA is largely open in the hypoxic environment in utero and constricts when O2 levels rise at birth, allowing the lung to take over its postnatal role of gas exchange. A patent DA is a common complication in premature infants. Decreased O2-induced constriction in the preterm rabbit DA is associated with decreased O2-sensitive IK and O2-sensitive Kv1.5 expression, and overexpression of Kv1.5 in preterm rabbit DA restores O2-induced constriction (30). In this study we show similar results in dPAs, suggesting the importance for Kv channels in O2 sensing.
Besides maturational changes in Kv channels, other factors may contribute to the altered response to hypoxia in experimental BPD. The mechanism by which hypoxia is sensed remains unknown. It has been proposed that HPV results from a redox sensor system present in dPAs, but also found in other O2-sensing organs (10). A sensor, the proximal portion of the mitochondria transport chain, produces a mediator (reactive oxygen species). This mediator would alter the effector (K+ channels) though a process of reduction/oxidation. When the mitochondria detects a hypoxic environment, reactive oxygen species decrease thereby inhibiting redox-sensitive K+ channels. Inhibitors of complex I and III of the electron transport chain in the mitochondria inhibit the formation of reactive oxygen species and mimic HPV in the isolated perfused lung (11, 42). In experimental BPD, a loss of O2 sensing ability caused by chronic exposure to hyperoxia could alter mitochondrial function–reactive oxygen species production, making the pulmonary circulation unable to detect hypoxia and thus failing to constrict.
Abnormal vascular reactivity in older children with a history of severe BPD and pulmonary hypertension has been reported (43). Cardiac catheterization of these patients showed that the pulmonary vasculature remained responsive to O2, but PVR was enhanced in response to hypoxia, suggesting increased HPV in this subpopulation of patients. The discrepancy with our data could be explained by differences in species, but is most likely due to a different timing in the assessment. HPV was assessed at a median of 5 years of age in the clinical study. Our study assessed animals during the neonatal period. As an adaptative and maturational response, the pulmonary vasculature with an initial blunted HPV could become hyperreactive to hypoxia later in life. Indications for such long-term consequences of injury during the perinatal period have been reported in the fawn-hooded rat (44) and more recently in rats exposed during fetal life to hypoxia showing that perinatal hypoxia alters the maturational shift in K+ channels and influences pulmonary vascular tone in adulthood (45). Remarkably, even ambient air may represent relative hyperoxia for infants born too early and impair lung development and function (46).
In conclusion, HPV is blunted in experimental CLD due, at least in part, to down-regulation of O2-sensitive Kv channels. Hypoxic episodes in premature infants with CLD may result from impaired HPV. We speculate that restoration of O2-sensitive Kv channels in the pulmonary circulation may decrease episodes of hypoxemia, facilitate management, and ultimately improve outcome of premature infants with CLD.
The authors thank Donna Beker and Paul Waszak for technical assistance with echographic studies. The authors acknowledge the technical assistance of L. Kullama and P. Davis for their participation in the newborn lamb studies.
Supported by Canadian Institutes of Health Research (CIHR), Alberta Heritage Foundation for Medical Research (AHFMR), Heart and Stroke Foundation Canada, Canada Foundation for Innovation (CFI), and a Canada Research Chair in Translation Lung and Vascular Developmental Biology (BT). S.L.A. is supported by NIH grant HL071115, CFI, and CIHR. G.J.R.-P. is supported by a stipend from the Maternal Fetal Neonatal Health Training Program sponsored by CIHR-IHDCYH. The sheep component of this work was supported by March of Dimes Birth Defects Foundation Grant 6-FY97–0138 (to R.D.B.) and NHLBI Grants HL-62512 (R.D.B.) and HL-62875 (K.H.A.).
Originally Published in Press as DOI: 10.1164/rccm.200711-1631OC on May 29, 2008
Conflict of Interest Statement: G.J.R.-P. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. S.L.A. has a patent pending for the use of pyruvate dehydrogenase kinase inhibitors to treat human cancer. R.D.B. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. K.H.A. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. D.P.C. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. S.-C.C. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. B.K. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. A.H. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. F.E. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. X.W. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. B.T. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript.