O2 Sensing in the Human Ductus Arteriosus
Regulation of Voltage-Gated K+ Channels in Smooth Muscle Cells by a Mitochondrial Redox Sensor
Functional closure of the human ductus arteriosus (DA) is initiated within minutes of birth by O2 constriction. It occurs by an incompletely understood mechanism that is intrinsic to the DA smooth muscle cell (DASMC). We hypothesized that O2 alters the function of an O2 sensor (the mitochondrial electron transport chain, ETC) thereby increasing production of a diffusible redox-mediator (H2O2), thus triggering an effector mechanism (inhibition of DASMC voltage-gated K+ channels, Kv). O2 constriction was evaluated in 26 human DAs (12 female, aged 9±2 days) studied in their normal hypoxic state or after normoxic tissue culture. In fresh, hypoxic DAs, 4-aminopyridine (4-AP), a Kv inhibitor, and O2 cause similar constriction and K+ current inhibition (IK). Tissue culture for 72 hours, particularly in normoxia, causes ionic remodeling, characterized by decreased O2 and 4-AP constriction in DA rings and reduced O2- and 4-AP–sensitive IK in DASMCs. Remodeled DAMSCs are depolarized and express less O2-sensitive channels (including Kv2.1, Kv1.5, Kv9.3, Kv4.3, and BKCa). Kv2.1 adenoviral gene-transfer significantly reverses ionic remodeling, partially restoring both the electrophysiological and tone responses to 4-AP and O2. In fresh DASMCs, ETC inhibitors (rotenone and antimycin) mimic hypoxia, increasing IK and reversing constriction to O2, but not phenylephrine. O2 increases, whereas hypoxia and ETC inhibitors decrease H2O2 production by altering mitochondrial membrane potential (ΔΨm). H2O2, like O2, inhibits IK and depolarizes DASMCs. We conclude that O2 controls human DA tone by modulating the function of the mitochondrial ETC thereby varying ΔΨm and the production of H2O2, which regulates DASMC Kv channel activity and DA tone.
In the fetus, the ductus arteriosus (DA) is tonically relaxed by its hypoxic environment. This allows venous blood to bypass the nonventilated lungs. At birth, simultaneous pulmonary vasodilatation and DA vasoconstriction direct right ventricular blood flow into the pulmonary circulation. Functional closure of the DA (due to vasoconstriction) precedes anatomical closure (due to cell proliferation) by days and is crucial to the newborn’s transition to an air-breathing organism. The DA constrictor response to O2, although modulated by the endothelium (reinforced by endothelin [ET] and inhibited by vasodilatory prostanoids and nitric oxide [NO]), is intrinsic to the DA smooth muscle cell (DASMC).1 O2 constriction persists ex vivo, after endothelial denudation, and despite inhibition of prostaglandin H synthase (PGHS), nitric oxide synthase (NOS, and ET-A receptors.2 We have recently provided pharmacological evidence that O2-induced constriction of the human DA involves inhibition of voltage-gated potassium channels (Kv) in DASMCs.2 As in most arteries, this K+ channel inhibition leads to SMC depolarization, opening of the voltage-gated L-type Ca2+ channels, influx of Ca2+, and vasoconstriction. However, the molecular identity of the relevant DASMC Kv channels, as well as the mechanism of O2 sensing in this important human vessel, remain unknown.
Vascular O2 sensing systems consist of a sensor, the function of which is altered by changes in Po2, a mediator, produced by this sensor, and an effector, which alters vascular tone in response to the mediator.3 Many aspects of the O2 sensor-effector pathway are conserved among O2-sensitive mammalian tissues, the pulmonary artery (PA), the DA, the adrenomedullary cells, the neuroepithelial body, and the carotid body. In each tissue, K+ channels have been implicated in the effector mechanism (see review3), whereas the O2 sensor has been proposed to involve a change in redox state, as determined by mitochondria or NADPH oxidase. Attention has focused on the mitochondria because electron transport chain (ETC) inhibitors mimic hypoxia, constricting the PA and activating the carotid body. Mitochondria respond to changes in Po2 by altering their respiration and production of reactive O2 species (ROS).4,5⇓ This changes cellular redox potential and alters the function of many redox-sensitive genes, second messenger systems, and O2-sensitive K+ channels in the membrane, before any depletion of ATP. Because K+ channels control SMC membrane potential (EM), and thus tone, in most vascular beds, O2-sensitive K+ channels are attractive candidate effectors in vascular redox-based O2-sensing systems. O2-sensitive K+ channels include homo- and heterotetramers composed of the several Kv α-subunits (Kv1.2,6 Kv1.5,7 Kv2.1,6–8⇓⇓ Kv3.1b, Kv4.3,9 and Kv9.3) and the calcium-sensitive K+ channels, BKCa.10,11⇓
In rabbit DAs, O2, and oxidants such as H2O2,12,13⇓ inhibit DASMC Kv current and depolarize EM, leading to vasoconstriction. We now directly assess the hypothesis that O2 constricts the human DA via inhibition of DASMC Kv channels. We also hypothesize that the channels are under the control of an O2 sensor within the mitochondrial ETC and speculate that the sensor and effector are linked by mitochondrial-derived ROS, specifically H2O2.
Materials and Methods
The human studies committee of the University of Alberta approved the use of discarded DAs, excised from 26 neonates with hypoplastic left heart syndrome during the Norwood procedure (aged 9±2 days, 12 female, arterial Po2 ≈40 mm Hg). DAs were transported in iced saline, maintained in hypoxia, and used within 1 hour. Ring tension was recorded as previously described2 (see the expanded Materials and Methods section, which can be found in the online data supplement available at http://www.circresaha.org).
Whole-Cell Patch Clamping Technique
Current and voltage clamp technique, SMC isolation, and pipette solutions were performed as previously described (see online data supplement).14 The response to O2, ETC inhibitors, and the membrane permeable H2O2 analog, t-butyl hydrogen peroxide, were compared between DASMCs enzymatically dispersed from fresh DAs versus those maintained in tissue culture for 72 hours (either under normoxic or hypoxic conditions).
DASMC EM was measured both using patch clamp technique in current clamp mode and noninvasively with DiBAC43 (20 μmol/L, bis-barbituric acid oxonol), a potentiometric dye that increases green fluorescence on depolarization (see online data supplement).
Laser capture microdissection and immunofluorescence were performed as described in the online data supplement.
Quantitative real-time polymerase chain reaction (qRT-PCR) was used to quantify human Kv channel mRNA, relative to GAPDH. The qRT-PCR methodology and equations for calculating the relative copy number of Kv channel mRNA, normalized to GAPDH, (2ΔΔCt) are described in the online data supplement.
Measurement of ROS by Chemiluminescence
H2O2, was measured using a specific, fluorometric AmplexRed assay, as previously described16 (see online data supplement and a calibration curve in online Figure 1).
Mitochondrial membrane potential (ΔΨm) was measured using two independent and well-validated probes JC-1 and TMRM (tetramethyl rhodamine methyl ester), as previously described.16 These well-characterized, cationic fluorescent dyes exhibit potential-dependent mitochondrial accumulation. JC-1 dimerizes and fluoresces red in hyperpolarized mitochondria versus green in depolarized mitochondria (high versus low ΔΨm, respectively), whereas TMRM accumulates and fluoresces red in proportion to ΔΨm16 (see online supplement).
Ex Vivo Gene Transfer
Adenoviruses (serotype 5, Ad5) carrying genes for rat Kv2.1 and green fluorescent protein (GFP) (GFP, Ad5-GFP-Kv2.1) or GFP alone (Ad5-GFP), each under a CMV promoter, were constructed using the Adeasy-1 system as previously described.17 DAs were incubated for 12 hours with either vehicle or viruses. The rings were then incubated in tissue culture media under normoxic (Po2 ≈120 mm Hg) or hypoxic (Po2 ≈45 mm Hg) conditions for ≈60 hours. The effects of gene transfer on vasoconstriction to O2 and 4-AP, whole-cell electrophysiology (IK and EM), and gene expression were studied in fresh DAs and DAs exposed to normoxic or hypoxic culture.
Statistics and Drugs
For detailed information regarding statistics and drugs, please see the online data supplement.
Kv Channels Control Tone
Fresh human DA rings were studied in hypoxia to mimic conditions in utero, except when otherwise specified. The Kv channel inhibitor 4-AP (1 to 10 mmol/L) significantly constricts the human DA in a dose-dependent manner (Figure 1A); iberiotoxin (IBTx, 200 nmol/L, a BKCa blocker) does not. The magnitude of O2 and 4-AP constriction are strongly correlated (Figure 1B). Although both 4-AP and IBTx significantly inhibit whole-cell K+ current (IK; Figures 1C and 1D), IBTx causes minimal inhibition of IK at potentials close to the resting EM (Figure 1E). In contrast, O2 and 4-AP cause similar, significant decreases in IK at −30 mV (Figure 1E).
mRNA for Kv1.5 and Kv2.1 is expressed in the media and SMCs of DA, selectively extracted with laser capture microdissection (LCM) (Figure 2A). This technique is important because the DA is composed of many cell types and LCM allows specific selection of the media and thus preferential measurement of channel expression in SMCs. Furthermore, immunofluorescent colocalization shows that Kv2.1 (Figure 2B) and Kv1.5 (not shown) protein is expressed in SM α-actin positive DASMC. This is also confirmed by conventional immunohistochemistry shown in online Figure 2.
Proximal ETC Is the O2 Sensor
Both rotenone, a complex I ETC inhibitor, and antimycin, a complex III inhibitor, relax the O2-preconstricted DA, mimicking hypoxia (Figures 3A and 3B). In contrast, cyanide does not dilate the DA (Figures 3A and 3B). Whereas either rotenone or antimycin alone only partially relax the normoxic DA, pretreatment with both inhibitors completely eliminates O2 constriction (Figure 4A). ETC inhibitor relaxation is not due to DA damage or nonspecific suppression of tone, because phenylephrine constriction is unaltered by rotenone and antimycin A (Figure 4A). Further support for the contention that O2 and rotenone target the same mechanism is the strong, direct correlation between the magnitude of O2 constriction and rotenone-relaxation (Figure 3C).
In fresh human DASMCs, O2 rapidly and reversibly inhibits IK (Figures 3D through 3F). Rotenone mimics hypoxia, precisely restoring the current that was rapidly suppressed by O2 (within 5 minutes). In contrast, cyanide (a complex IV blocker, 10 μmol/L) does not alter IK (data not shown). These data provide strong evidence that the mechanism by which both proximal ETC inhibitors and hypoxia relax the DA is DASMC Kv activation.
A New Model of Ionic Remodeling: DAs in Tissue Culture
The effects of O2 and rotenone on tone (Figure 4A), K+ channel function (Figures 4B and 5A through 5⇓C), and expression (Figure 5E) were studied in DAs kept in normoxic versus hypoxic tissue culture for 72 hours. The chronically normoxic DA rings retain the ability to constrict to phenylephrine, whereas they loose both the O2 constriction and the rotenone-relaxation responses (Figure 4A). SMCs from chronically normoxic DAs have a significantly decreased current density, compared with the freshly isolated cells studied under identical conditions (Figure 4B). In addition, when these DASMC are returned to hypoxic conditions, their IK is unresponsive to abrupt increase in Po2, 4-AP, or rotenone (Figure 4B). The loss of O2 and ETC sensitivity in chronically normoxic DA is associated with basal, hypoxic EM depolarization and loss of the ability to depolarize in response to rapid increases in Po2 or 4-AP, whereas the response to KCl is preserved (Figures 5A through 5D). The concordant findings of impaired membrane responses to O2 using both whole-cell current clamp and potentiometric dyes in both DASMCs and DA rings excludes the theoretical possibility that this could be an artifact, related to enzymatic dispersion of cells or loss of cell-cell connection.
Ionic Remodeling Results From Downregulation of O2-Sensitive Kv Channels
DAs were divided in thirds and both the function and expression of K+ channels and ETC complexes were compared in fresh hypoxic DAs versus DAs cultured in chronic normoxia or chronic hypoxia. In chronic normoxia, mRNA for Kv1.5 and Kv2.1 (measured using qRT-PCR) and Kv4.3, Kv9.3 and BKCa (measured using conventional RT-PCR; online Figure 3) is downregulated, relative to the housekeeping gene GAPDH (Figure 5E). There is no decrease in Kv1.1, Kir2.1, or TASK expression, measured using conventional RT-PCR, suggesting that the downregulation of these O2-sensitive channels is somewhat specific (see online Figure 3). Although there was also a trend to lower Kv1.5 and Kv2.1 mRNA in chronic normoxia versus chronic hypoxia (Figure 5E), this was not statistically significant. Expression of selected subunits from ETC complexes I-IV is unaltered by chronic normoxia (online Figure 4).
Kv2.1 Gene Restoration in the Ionically Remodeled DA
If the depressed Kv expression in this model is both real and important, it would follow that restoring Kv expression would be sufficient to restore the missing O2 responsiveness of chronically normoxic DASMCs. Indeed, Kv2.1 gene therapy does partially restore O2 responsiveness (Figure 6). DAs were divided into 4 pieces. One piece was studied immediately upon arrival in the laboratory under hypoxia, whereas the others were exposed to normoxia for 12 hours in the presence of vehicle, Ad5-GFP, or Ad5-GFP-Kv2.1. This was followed by 60 hours of normoxic incubation to allow gene and protein expression. Of the 6 human DAs infected, successful gene transfer, as measured by GFP fluorescence, was confirmed in 4 (Figures 6A and 6B). The Kv2.1 transgene was derived from rat and, using the species specificity of the qRT-PCR probe, we were able to show that expression of rat Kv2.1 mRNA occurred exclusively in Ad5-GFP-Kv2.1-infected DAs (Figure 6C). The transgene yielded functional channels, indicated by the fact that DASMCs isolated from Ad5-GFP-Kv2.1 rings had a larger Kv current compared with the DASMCs from the noninfected DAs (Figure 6B). Kv2.1 gene transfer, significantly restores the ability of the normoxic DA to respond to O2 and 4-AP (Figures 6D and 6E).
Mitochondria-Derived ROS Are the Redox Mediators of Normoxic DA Constriction
Inhibitors of the proximal ETC and authentic hypoxia rapidly depolarize ΔΨm in DASMCs in primary, hypoxic culture (Figures 7A through 7D). Conversely, cyanide 10 μmol/L does not rapidly alter ΔΨm (Figure 7C). To address the concern that cyanide should (at some dose) be effective in collapsing ΔΨm, we also assessed its effects on a cardiac HL-1 cell line.18 These experiments showed that 10 μmol/L CN readily depolarizes ΔΨm in HL-1 cells, suggesting diversity in mitochondria between vascular versus cardiac cells (online Figure 5). Thus, ΔΨm in DASMCs is much more sensitive to rotenone than to cyanide, consistent with its lack of electrophysiological and hemodynamic effects.
Lucigenin-enhanced chemiluminescence (n=2) and H2O2 production (n=5) are increased within minutes by raising Po2 from 40 to 100 mm Hg in freshly isolated, human DA rings (Figures 8A and 8B). Rotenone decreases H2O2 production (Figures 8A and 8B). t-butyl-H2O2 mimics the effects of O2 on DASMC electrophysiology. It inhibits IK, depolarizes fresh hypoxic DASMCs, and both these effect are lost after exposure to chronic normoxia (Figures 8C and 8D).
This is the first comprehensive study of the mechanism of O2 sensing in the human DA. There are 3 major findings of this study. First, Kv channels are the effectors of O2 constriction (Figures 1 and 2⇑). Inhibiting these channels, whether by 4-AP or O2 leads to vasoconstriction. Second, the proximal mitochondrial ETC serves as the O2 sensor (Figures 3, 4A, and 7⇑⇑). Inhibition of the complex I or III inhibits O2 constriction. Third, consistent with data in rabbit DA,13 the mediator linking the sensor and effector in the human DA appears to be a ROS (H2O2) (Figure 8). The ability of hypoxia and proximal ETC inhibitors to depolarize ΔΨm (Figure 7) provides a probable mechanism by which hypoxia alters ROS production. In addition, a new model of ionic remodeling is introduced that is useful in understanding the relative contributions of Kv channels and mitochondria to O2 sensing in the human DA (Figures 4 through 6⇑⇑). Together, these findings suggest that ROS produced in the proximal ETC in response to O2 could be redox mediators that link the mitochondrial sensor to the Kv effector and thus control tone, as illustrated schematically in Figure 8D.
Roulet and Coburn first demonstrated that O2-induced DA constriction is associated with membrane depolarization.19 Although the K+ channel effector mechanism is widely conserved among O2 sensitive tissues, the type of K+ channel and downstream response to channel inhibition may vary among species, between tissues, and with maturation.3 Therefore, we have focused our studies on the mechanism of O2 constriction in term human DA. In addition, special effort was made to ensure that all DAs were maintained in their normal hypoxic environment, except when they were intentionally exposed to acute or chronic normoxia. Despite the congenital heart disease, these DAs have normal responses to O2 with similar thresholds for onset of O2 constriction and maximal O2 responses,2 as in the literature.1
In the human DA, O2 and 4-AP cause nonadditive vasoconstriction.2 Nifedipine, an inhibitor of L-type Ca2+ channels, blocks both 4-AP and O2 constriction,1,2⇓ consistent with an obligatory and coordinated role for Kv and Ca2+ channels in O2 constriction. The present study confirms the ability of Kv, but not BKCa, channel blockers to constrict the human DA and demonstrates a strong correlation between the magnitude of 4-AP constriction and O2 constriction (Figure 1A and 1B). The basis for these physiological observations is now directly identified. The DASMCs express several O2- and 4-AP–sensitive K+ channels, which are involved in hypoxic pulmonary vasoconstriction (Kv1.5 and Kv2.1)7 and O2 responses in the carotid body9 (Figure 2 and online supplement Figure 3).
A new model was developed to further understand the role of Kv channels in O2 constriction. In this model, designed to mimic the conditions in the first days after birth, exposure to tissue culture conditions ex vivo (particularly at normoxic Po2), inhibited DASMC IK (Figure 4B) and downregulated several O2-sensitive Kv channels, including Kv1.5 and Kv2.1 (Figure 5E). The results shown in the online data supplement demonstrate that expression of other O2-sensitive channels, Kv4.3,9 Kv9.3,8and BKCa channels,10,11,20⇓⇓ also decrease in this model. However, the lack of effect of iberiotoxin on tension suggests that the role for BKCa channels may be less important in O2 response in the DA than in the carotid body. The role of Kv4.3 is somewhat less likely as these channels generate a rapidly inactivating current, unlike the slow non-inactivating O2-sensitive current in DASMCs.21
The loss of these Kv channels is associated with membrane depolarization (Figures 5A through 5C) and impaired ability of the membrane to further depolarize to either O2 or 4-AP (Figure 5D), even hours after the DASMCs are returned to a hypoxic environment. These changes in the electrical properties of the DA are termed ionic remodeling. Similar ionic remodeling and loss of O2 responsiveness has also been reported in other O2-sensitive tissues. Loss of acute hypoxic pulmonary vasoconstriction, which occurs with chronic hypoxia, is also associated with downregulation of Kv2.1 and Kv1.5 expression and loss of the O2-sensitive portion of IK.15,17⇓
The ionic remodeling model further highlights the importance of the Kv channels to O2-induced constriction. The reduction in K+ current density is associated with loss of sensitivity of current and tone to O2 and 4-AP (Figure 4). Furthermore, the ability of rotenone to increase IK is lost in this model (Figure 4B). Coupled with the qRT-PCR evidence for loss of O2-sensitive channels, but not ETC complexes (online Figure 4), these data suggest that the loss of O2 constriction is primarily due to modifications of the Kv effector, rather than the mitochondrial sensor. Consistent with this interpretation, restoration of Kv2.1 expression by ex vivo gene transfer significantly restores the constriction to both 4-AP and O2 (Figure 6). The fact that restoration of O2 and 4-AP constriction is incomplete after Kv2.1 restoration may relate to the fact that other downregulated O2- and 4-AP–sensitive channels (eg, Kv1.5, Kv9.3, and Kv4.3) were not replaced. Alternatively, alterations in function of other components of the contractile apparatus may be abnormal.
Although the pulmonary artery and the DA are contiguous, their response to O2 is reversed. Hypoxia causes pulmonary vasoconstriction versus DA relaxation. It is intriguing that similar Kv channels are present in both arteries.2,7⇓ The fact that the Kv inhibitor 4-AP constricts both arteries suggests that the Kv channel setting EM in the 2 tissues may be similar.1,7⇓ By extension, this implies that their differential O2 response may relate to differences in either the O2 sensor or the response of the K+ channels to the redox messenger produced by a shared sensor. In human DAs, it appears that O2-sensitive Kv channels are the effectors and the mitochondrial ETC is the O2 sensor, very similar to the pulmonary artery.5 What differs, as is discussed subsequently, is the response to ROS.
What is the evidence that Kv function and vascular tone is regulated by the mitochondria? First, rotenone and antimycin mimic hypoxia better than any other class of drugs, suggesting a role for mitochondria in vascular O2 sensing.5,13,15,22,23⇓⇓⇓⇓ Proximal ETC inhibition causes pulmonary vasoconstriction,5 systemic dilatation,24 carotid body activation,25 and DA vasodilatation (Figure 3). Rotenone and antimycin, but not cyanide, selectively inhibit O2 constriction in the human DA and reverse O2-induced inhibition of IK in DASMCs (Figures 3D through 3F). Indeed, the effects of rotenone and antimycin are additive, and the combination completely prevents O2 constriction, without preventing phenylephrine constriction (Figure 4A). In addition, the more a DA constricts to O2 the more relaxation occurs in response to rotenone, but not to cyanide (Figure 3C). These findings are a mirror image of those in the pulmonary circulation where ETC inhibitors, also mimicking hypoxia, cause pulmonary vasoconstriction and inhibition of IK.5 However, in both the pulmonary circulation and DA, hypoxia, and ETC inhibitors decrease ROS production (Figure 8),26 suggesting the difference between the vessels relates to the response of their K+ channels to H2O2 and ROS. Indeed, DASMCs depolarize in response to H2O2 (Figure 8), the opposite of the response seen in pulmonary artery SMCs.12
The link between the mitochondria, Kv channels, and tone is through a redox mediator, rather than ATP levels (see Figure 8D). Although mitochondria have been dismissed as poor candidates for O2 sensing because the Km of their cytochromes may be too low for modulation by physiological levels of hypoxia, they exhibit a reversible inhibition of respiration during prolonged hypoxia due to inhibition of cytochrome c oxidase.27 Mitochondria are increasingly recognized to be involved in intracellular Ca2+ control and in redox signaling, in part, because they are important sources of ROS.5,23⇓ ΔΨm, a major determinant of cellular redox potential, changes rapidly over a physiological range of Po2 values in type I carotid body cells, depolarizing in response to hypoxia and metabolic inhibitors.25 Similarly, hypoxia, rotenone, and antimycin depolarize ΔΨm in human DASMCs within 1 to 2 minutes (Figure 7) and this is associated with impaired ROS production (Figures 8A and 8B). Our confocal images also show that the mitochondria in DASMCs, far from being remote from the plasma membrane, form a ubiquitous, filamentous network that permeates the cytosol and is thus positioned to signal changes in Po2 to all cellular compartments, including the plasma membrane (Figure 7).
We propose that the increase in the H2O2 levels that occurs with normoxia at the time of birth inhibits DASMC Kv current, causing depolarization and vasoconstriction. Indeed, in human DAs, H2O2 production increases as Po2 rises (Figures 8A and 8B) and H2O2 inhibits IK, causing membrane depolarization (Figures 8C and 8D). Likewise, in rabbit DASMCs, intracellular administration of physiological doses of H2O2 decreases IK, and these effects are inhibited by catalase.13 The molecular basis for the differential electrophysiological response to H2O2 in the pulmonary artery SMC (activation) versus the DASMC (inhibition) is unknown. Possible explanations include tissue-specific differences in K+ channel heterotetramer composition, α-subunit splice variants, sulfhydryl redox state of key channel amino acids, or β-subunit expression.
Several questions remain to be answered. First, why is the distal ETC not as involved in O2 sensing as the proximal ETC? In the pulmonary circulation, rotenone and antimycin mimic hypoxia (decrease IK, cause constriction, and inhibit additional hypoxic constriction); cyanide does not.5 Cyanide’s lack of effect is not due to inadequate dosage. This dose of cyanide, administered under similar circumstances, depolarizes ΔΨm in HL-1 cells (online Figure 5) and large doses of cyanide depolarize ΔΨm in DASMCs (Figure 7C). Rather, we believe that because the majority of ROS production occurs at complex I and III,4 inhibitors of this portion of the pathway are most effective in interrupting the production of ROS signaling molecules and thus impair O2 sensing (Figure 8D).
Second, although there is growing agreement that mitochondria or a vascular NADPH oxidase are vascular redox O2 sensors, there is debate as to whether hypoxia and ETC inhibitors decrease26,28,29⇓⇓ or increase23 AOS production. We recently have found concordant depression of vascular generation of ROS and H2O2 by hypoxia using three independent techniques: dichlorofluorescein, lucigenin, and Amplex Red.16
The initial constriction of the DA is required for the remodeling and ultimately closure of the DA. Thus, understanding the mechanism of O2 constriction has implications for the common clinical problem of patent DA in premature infants. Perhaps DAs in premature infants fail to constrict to normoxia because their DASMCs are “deficient” in Kv channels. Restoration of Kv2.1 or Kv1.5 expression, whether accomplished by gene transfer, as in this study, or by other means,17 might have therapeutic potential. Augmenting Kv2.1 or Kv1.5 expression in premature DAs might restore its ability to constrict to O2 and thus enter a normal remodeling phase, leading to closure.
Drs Michelakis and Archer are both supported by Alberta Heritage Foundation for Medical Research (AHFMR), Canadian Institutes of Health Research (CIHR), the Heart and Stroke Foundation of Canada, and the Canadian Foundation for Innovation. Rat Kv2.1 cDNA kindly provided by Dr K. Takimoto, University of Pittsburgh, Pa.
Original received November 15, 2001; revision received August 13, 2002; accepted August 15, 2002.
- ↵Archer SL, Huang J, Henry T, Peterson D, Weir EK. A redox based oxygen sensor in rat pulmonary vasculature. Circ Res. 1993; 73: 1100–1112.
- ↵Hulme JT, Coppock EA, Felipe A, Martens JR, Tamkun MM. Oxygen sensitivity of cloned voltage-gated K+ channels expressed in the pulmonary vasculature. Circ Res. 1999; 85: 489–497.
- ↵Archer SL, Souil E, Dinh-Xuan AT, Schremmer B, Mercier JC, El Yaagoubi A, Nguyen-Huu L, Reeve HL, Hampl V. Molecular identification of the role of voltage-gated K+ channels, Kv1.5 and Kv2.1, in hypoxic pulmonary vasoconstriction and control of resting membrane potential in rat pulmonary artery myocytes. J Clin Invest. 1998; 101: 2319–2330.
- ↵Patel AJ, Lazdunski M, Honore E. Kv2.1/Kv9.3, a novel ATP-dependent delayed-rectifier K+ channel in oxygen-sensitive pulmonary artery myocytes. EMBO J. 1997; 16: 6615–6625.
- ↵Perez-Garcia MT, Lopez-Lppez JR, Riesco AM, Hoppe UC, Marban E, Gonzalez C, Johns DC. Viral gene transfer of dominant-negative Kv4 construct suppresses an O2-sensitive K+ current in chemoreceptor cells. J Neurosci. 2000; 20: 5689–5695.
- ↵Cornfield DN, Reeve HL, Tolarova S, Weir EK, Archer SL. Oxygen causes fetal pulmonary vasodilation through activation of a calcium-dependent potassium channel. Proc Natl Acad Sci U S A. 1996; 93: 8089–8094.
- ↵Riesco-Fagundo AM, Perez-Garcia MT, Gonzalez C, Lopez-Lopez JR. O2 modulates large-conductance Ca2+-dependent K+ channels of rat chemoreceptor cells by a membrane-restricted and CO-sensitive mechanism. Circ Res. 2001; 89: 430–436.
- ↵Michelakis ED, Hampl V, Nsair A, Wu X, Harry G, Haromy A, Gurtu R, Archer SL. Diversity in mitochondrial function explains differences in vascular oxygen sensing. Circ Res. 2002; 90: 1307–1315.
- ↵Archer SL, Nelson DP, Weir EK. Detection of activated oxygen species in vitro and in rat lungs using chemiluminescence. J Appl Physiol. 1989; 67: 1912–1921.
- ↵Reeve HL, Michelakis E, Nelson DP, Weir EK, Archer SL. Alterations in a redox oxygen sensing mechanism in chronic hypoxia. J Appl Physiol. 2001; 90: 2249–2256.
- ↵Michelakis ED, McMurtry MS, Wu XC, Dyck JR, Moudgil R, Hopkins TA, Lopaschuk GD, Puttagunta L, Waite R, Archer SL. Dichloroacetate, a metabolic modulator, prevents and reverses chronic hypoxic pulmonary hypertension in rats: role of increased expression and activity of voltage-gated potassium channels. Circulation. 2002; 105: 244–250.
- ↵Claycomb WC, Lanson NA Jr, Stallworth BS, Egeland DB, Delcarpio JB, Bahinski A, Izzo NJ Jr. HL-1 cells: a cardiac muscle cell line that contracts and retains phenotypic characteristics of the adult cardiomyocyte. Proc Natl Acad Sci U S A. 1998; 95: 2979–2984.
- ↵Roulet MJ, Coburn RF. Oxygen-induced contraction in guinea pig neonatal ductus arteriosus. Circ Res. 1981; 49: 997–1002.
- ↵Dixon JE, Shi W, Wang HS, McDonald C, Yu H, Wymore RS, Cohen IS, McKinnon D. Role of the Kv4.3 K+ channel in ventricular muscle: a molecular correlate for the transient outward current [published erratum appears in Circ Res. 1997;80:147]. Circ Res. 1996; 79: 659–668.
- ↵Waypa GB, Chandel NS, Schumacker PT. Model for hypoxic pulmonary vasoconstriction involving mitochondrial oxygen sensing. Circ Res. 2001; 88: 1259–1266.
- ↵Archer SL, Reeve HL, Michelakis ED, Puttagunta L, Waite R, Nelson DP, Dinauer MC, Weir EK. Oxygen sensing is preserved in mice lacking the gp91 fox subunit of NADPH oxidase. PNAS. 1999; 96: 7944–7949.
- ↵Chandel NS, Budinger GR, Schumacker PT. Molecular oxygen modulates cytochrome c oxidase function. J Biol Chem. 1996; 271: 18672–18677.
- ↵Freeman BA, Crapo JD. Hyperoxia increases oxygen radical production in rat lungs and lung mitochondria. J Biol Chem. 1981; 256: 10986–10992.
- ↵Archer S, Nelson D, Weir E. Simultaneous measurement of oxygen radicals and pulmonary vascular reactivity in the isolated rat lung. J Appl Physiol. 1989; 67: 1903–1911.