Abstract
Hypoxic pulmonary vasoconstriction (HPV) is critically important in regionally heterogeneous lung diseases by directing blood toward better-oxygenated lung units, yet the molecular mechanism of HPV remains unknown. Transient receptor potential (TRP) channels are a large cation channel family that has been implicated in HPV, specifically in the pulmonary artery smooth muscle cell (PASMC) Ca2+ and contractile response to hypoxia. In this study, the authors probed the role of the TRP family member, TRPV4, in HPV.
HPV was assessed by using isolated perfused mouse lungs or by intravital microscopy to directly visualize pulmonary arterioles in mice. In vitro experiments were performed in primary human PASMC.
The hypoxia-induced pulmonary artery pressure increase seen in wild-type mice (5.6 ± 0.6 mmHg; mean ± SEM) was attenuated both by inhibition of TRPV4 (2.8 ± 0.5 mmHg), or in lungs from TRPV4-deficient mice (Trpv4−/−) (3.4 ± 0.5 mmHg; n = 7 each). Functionally, Trpv4−/− mice displayed an exaggerated hypoxemia after regional airway occlusion (pao2 71% of baseline ± 2 vs. 85 ± 2%; n = 5). Direct visualization of pulmonary arterioles by intravital microscopy revealed a 66% reduction in HPV in Trpv4−/− mice. In human PASMC, inhibition of TRPV4 blocked the hypoxia-induced Ca2+ influx and myosin light chain phosphorylation. TRPV4 may form a heteromeric channel with TRPC6 as the two channels coimmunoprecipitate from PASMC and as there is no additive effect of TRPC and TRPV4 inhibition on Ca2+ influx in response to the agonist, 11,12-epoxyeicosatrienoic acid.
TRPV4 plays a critical role in HPV, potentially via cooperation with TRPC6.
Abstract
A role for the calcium-permeant transient receptor potential channel TRPV4 in hypoxic pulmonary vasoconstriction was demonstrated in mouse models in vivo and in vitro both pharmacologically and in Trpv4−/− mice. TRPV4 is critical for hypoxia-induced pulmonary vasoconstriction, in contrast to its vasodilatory roles in renal and mesentery arteries.
Hypoxic pulmonary vasoconstriction is an intrinsic response of the pulmonary vasculature that is critical to perfusion–ventilation matching
Vasoconstriction effected by pulmonary artery smooth muscle cells is mediated by calcium-activated contraction, but the route of calcium entry is unclear
A role for the calcium-permeant transient receptor potential channel TRPV4 in hypoxic pulmonary vasoconstriction was demonstrated in mouse models in vivo and in vitro both pharmacologically and in Trpv4−/− mice
TRPV4 is critical for hypoxia-induced pulmonary vasoconstriction in contrast to its vasodilatory roles in renal and mesenteric arteries
HYPOXIC pulmonary vasoconstriction (HPV) optimizes gas exchange by redirecting blood flow from poorly aerated to well-ventilated areas of the lung.1 Many molecular aspects of this process have been elucidated to date, including the recent identification that the hypoxic signal is initially sensed at the level of the capillary endothelium and propagated upstream to resistance arterioles as a conducted response.2 The ultimate effector of HPV is the pulmonary artery smooth muscle cell (PASMC). Hypoxia-induced PASMC contraction—and therefore vasoconstriction—stems from membrane depolarization and Ca2+ influx from several sources including voltage-operated Ca2+ channels,3 store-operated Ca2+ channels,4 and sarcoplasmic reticulum Ca2+ release channels.5
An important class of Ca2+ channel in lung vascular biology is the transient receptor potential (TRP) channel family.1,6 TRP channels are polymodal cation channels that are expressed in cellular membranes as homotetramers or heterotetramers and have been implicated as sensors of various cellular functions including mechanosensation and thermosensation.6,7 Specifically, the classical channel, TRPC6, has been implicated in the intact HPV response: TRPC6 is expressed in human PASMC, and inhibition of TRPC6 diminished the PASMC Ca2+ response to hypoxia.8–10 TRPC6 is activated by key mediators of HPV, such as epoxyeicosatrienoic acids (EETs),11,12 and lungs isolated from Trpc6−/− mice have an attenuated HPV response.13
Vanilloid subfamily (TRPV) channels are highly abundant in endothelial and vascular smooth muscle cells and in perivascular neural tissue.14 TRPV channels are differentially expressed in different vascular beds, indicating their specific roles in different organs. For example, nanomolar concentrations of the TRPV1 agonist capsaicin cause endothelial-dependent relaxation of mouse mesenteric, yet not renal arteries, whereas TRPV4 stimulation causes endothelial-dependent vasodilation in both vascular beds.15 Stimulation of TRPV4 channels in arterial smooth muscle cells of the systemic circulation likewise promotes vasodilation via activation of large-conductance Ca2+-activated K+ channels, resulting in membrane hyperpolarization.16,17 Conversely, activation of TRPV4 in PASMC by serotonin results in vasoconstriction.18 Similar to TRPC6, TRPV4 is highly expressed in PASMC19 and activated by EETs.20 Our group and others have demonstrated a role for endothelial TRPV4 in the disruption of vascular permeability and edema formation due to increased hydrostatic pressure or ventilator pressure.21–23 Based on these findings, an oral TRPV4 inhibitor has been developed that decreases cardiogenic lung edema in mice and is the current focus of further clinical testing.24 There is similar interest in pharmacological targeting of TRPV channels for therapeutic benefit in asthma, acute lung injury, and other relevant lung diseases.25,26
In addition to its general vasoactive role, TRPV4 has been shown to be up-regulated in chronic hypoxia in rat pulmonary arteries and was responsible for the increased myogenic tone in pulmonary vessels of hypoxic rats.27 Although these data demonstrate a clear link between TRPV4 and the PASMC response to agonists or chronic hypoxia, a potential role for TRPV4 in acute HPV has not been addressed so far. In this study, we hypothesized that TRPV4 would be required for an intact HPV response and that loss of TRPV4 function would result in a blunted PASMC Ca2+ response to hypoxia.
Materials and Methods
Reagents
HC-067047 and SKF-96365 were purchased from Sigma (Canada). Hank’s buffered salt solution (HBSS), 15-μm diameter yellow-green fluorescent microspheres, and Fura-2-acetoxymethyl ester (AM) were purchased from Invitrogen (USA). Anti-caveolin-1 antibody was purchased from BD Biosciences (Canada), and rabbit anti-TRPV4 antiserum was a generous gift from Dr. Stefan Heller (Stanford University, Stanford, California).28 Antimyosin light chain (MLC) and anti-phospho-MLC were purchased from Cell Signaling Technology (USA).
Animals
All mice were housed in accordance with national guidelines, and all animal experiments were approved by the Animal Care Committee at St. Michael’s Hospital (Toronto, Ontario, Canada). Male C57/Bl6 mice were obtained from Charles River Laboratories (Canada). Male mice deficient in TRPV4 (Trpv4−/−) bred on a background of C57/Bl6 were from Dr. W. Liedtke29 and were maintained on site. For experiments, mice were anesthetized with a single intraperitoneal injection of ketamine and xylazine, unless otherwise indicated.
Isolated Perfused Lung
Mice were anesthetized as described in Intravital Microscopy section and ventilated via a tracheostomy with a tidal volume of 10 ml/kg and at a rate of 90 breaths/min. The mice were then sacrificed by exsanguination. The pulmonary artery and left atrium were cannulated, and the lungs were perfused with 20% fetal bovine serum in HBSS containing Ca2+ and Mg2+ at 37°C, with a flow rate of 50 ml min−1 kg−1 as previously reported.2 For normoxic treatment, the ventilating gas contained 21% O2, 5% CO2, and 74% N2. Hypoxic gas was 1% O2 in 5% CO2, and 94% N2. After stabilization of the pulmonary artery pressure (PAP) on normoxic gas, ventilation was switched to hypoxia for 5 min. The peak PAP during this 5-min period was recorded. Ventilation was then returned to normoxia, and the PAP was allowed to return to baseline.
Intravital Microscopy
Mice were anesthetized with an intraperitoneal injection of fentanyl, midazolam, and meditomidine. After tracheostomy, they were ventilated with room air at 100 breaths/min. A window into the right chest was surgically opened as described earlier.30 An intrapleural catheter was inserted, the chest was resealed with a polyvinylidine membrane, and negative intrapleural pressure was reestablished at −3 mmHg. Mice were placed on a custom-made heated, motorized stage under an upright microscope (Axiotechvario 100HD; Zeiss, Germany). After baseline image acquisition (TillVision; Till Photonics, Germany) of subpleural pulmonary arterioles (PAs), ventilation was shifted to a fraction of inspired oxygen (Fio2) of 0.11. Vessels were imaged again after 10 min of hypoxic ventilation, and changes in vessel diameter in response to hypoxia were calculated from the acquired images.
Intratracheal Saline Instillation
The effect of ventilation/perfusion (V/Q) mismatch in vivo was assessed in C57/Bl6 or Trpv4−/− mice as previously described.2 In brief, the left carotid artery was cannulated with a heparinized catheter. An arterial blood gas sample was measured to obtain a baseline reading of the arterial partial pressure of oxygen (paO2). A small amount of normal saline (1 μl/g) was instilled intratracheally using a 25-gauge needle inserted into the tracheal tube (time 0). After a further 2 min of ventilation, a second arterial blood gas sample was analyzed.
One-lung Ventilation
As described earlier,2 anesthetized mice were ventilated, and their right internal jugular vein was cannulated. One-lung ventilation was established by advancing the tracheal tube into the left mainstem bronchus. After 5 min, a 150 μl bolus of 15-μm fluorescent microspheres was delivered to the right atrium via the internal jugular vein. Mice were then sacrificed by exsanguination. Each lung was harvested and weighed and then dissolved overnight in 4 N KCl. Microspheres were recovered by filtration and then were dissolved in Cellsolve acetate (Sigma). Fluorescence emission in each lung sample was assessed using a standard plate reader, and fluorescence was normalized per milligram of lung weight. A small section of kidney tissue was used as a negative control to ensure complete microsphere capture in the pulmonary circulation during the first passage (data not shown).
Intracellular Ca2+ Measurements
Human primary PASMC were purchased (Lonza, Switzerland) and cultured in normoxic conditions according to the manufacturer’s instructions. Cells were routinely used between passages 4 and 9. For real-time fluorescence imaging, cells were plated on 25-mm square coverglasses in six well plates. Cells were loaded with Fura 2-AM as per the manufacturer’s directions. Coverglasses were mounted inside a perfusion chamber heated to 37°C and were perfused with HBSS containing Ca2+ and Mg2+. Hypoxia was induced by bubbling HBSS with 1% O2 in a hypoxic chamber for 45 min before experiments, and oxygen content in the perfusate was verified using a clinical blood gas analyzer. Hypoxic buffer routinely had a pO2 less than 35 mmHg. Fura-2 imaging was performed as described previously,2 and 340/380 Fura-2 ratios were converted to Ca2+ concentration based on appropriate calibration parameters.31 Confocal experiments using Fluo-4 (Invitrogen) were performed similarly. Data were collected using a Leica GET SCOPE INFO and MetaMorph (Molecular Devices, USA) and are expressed as the peak change in Fluo-4 intensity after treatment, relative to baseline.
Fractionation of Caveolae
Primary human PASMC were maintained as mentioned in the section Intracellular Ca2+ Measurements. PASMC were exposed to hypoxia for 15 min in a hypoxic chamber, or left in room air, before lysis. Lysates were homogenized and then applied to sucrose gradients (40% to 5%) and centrifuged for 18 h at 60,000g. Seven equal fractions were collected, and equal amounts of protein were resolved on a 10% polyacrylamide gel. Membranes were immunoblotted for TRPV4 and caveolin-1 to identify caveolae. Data were quantified and expressed as the percent of TRPV4 signal in cav-1-containing fractions relative to total TRPV4.
Coimmunoprecipitation
Mice were anesthetized and ventilated as in the isolated perfused lung experiments. After 3 min of normoxic or hypoxic ventilation (the time at which the maximal PA pressure was recorded in other experiments), samples of whole lung were extracted and snap frozen in liquid nitrogen. Whole-lung lysates were collected by homogenization of these samples in ice-cold lysis buffer. Equal amounts of precleared whole-lung lysate were immunoprecipitated with anti-TRPV4 antibody, which was captured on Protein A/G beads (GE Healthcare, Canada). After washing the beads, captured protein was run on sodium dodecyl sulfate polyacrylamide gel electrophoresis and transferred to a nitrocellulose membrane. Membranes were probed with anti-TRPV4 or anti-TRPC6 (Alomone Labs, Israel). Supernatants from beads as well as samples incubated with beads alone (i.e., in the absence of immunoprecipitating TRPV4 antibody) also were run as negative controls.
Statistical Analysis
Data are shown as mean ± SEM and as scatter plots for four or less samples. Statistical analyses were performed using GraphPad Prism (GraphPad Software, USA). Pairwise comparisons were carried out using the Mann–Whitney U Test. Multiple comparisons used Kruskall–Wallace testing and Dunn multiple comparison test. A P value less than 0.05 was considered significant. Animals for experimentation were randomly selected from a given cage by the experimenter. As most data were collected in real time, the investigator was not blinded to the experimental conditions. Sample sizes were based on previous experience, and no formal power calculation was carried out before the study. There were no missing data in any experiments.
Results
To probe for a potential role for TRPV4 in HPV, we first used the isolated perfused mouse lung model. In lungs isolated from wild-type C57/Bl6 mice, upon shifting ventilating gas from a normoxic (21% O2) to a hypoxic (1% O2) mixture, we observed a rapid increase in PAP (fig. 1). The maximal increase in PAP during a 5-min period of hypoxic ventilation was 5.6 ± 0.6 mmHg. When the specific TRPV4 inhibitor, HC-06704732 , was added to the perfusate at a concentration of 20 μM, the hypoxic response was severely attenuated, by approximately 50%. This result indicates that TRPV4 is involved in the HPV response in isolated mouse lungs. To further assess this effect, we tested lungs isolated from mice lacking functional TRPV4 expression (Trpv4−/−).29 Lungs from Trpv4−/− mice also exhibited a diminished HPV response, with a similar magnitude as that seen with pharmacological inhibition of TRPV4. These results indicate that TRPV4 is a required component of the normal HPV response. Basal perfusion pressures during normoxia did not differ between groups (10 ± 0.6 mmHg), indicating that the different reponses to hypoxia were not the result of different baseline offsets. Because TRPC6 has previously been shown to be involved in HPV,13 we tested for a possible compensatory change in TRPC6 expression in our Trpv4−/− mice. Western blots of whole-lung lysates from Trpv4+/+ and Trpv4−/− mice revealed no significant difference in TRPC6 protein expression in mice lacking TRPV4 compared with controls (fig. 1C). Similarly, in a functional assay in the isolated perfused mouse lung, we observed no difference in pulmonary vasoconstriction in response to the TRPC6 channel agonist, hyperforin,33 in Trpv4+/+ and Trpv4−/− mice, with an increase in PAP of 4.0 ± 0.3 mmHg in wild-type versus 3.8 ± 0.6 mmHg in Trpv4−/− mice (P = 0.9) (fig. 1E). In both cases, this increase could be largely attenuated by treatment with the TRPC inhibitor, SKF-96365 (SKF). Together, these data demonstrate that the loss of TRPV4 function is not compensated for by modulating TRPC6 expression or function in vivo.
Although these whole-organ studies demonstrated a functional role for TRPV4 in HPV, we next aimed to verify this effect in vivo. To this end, we used intravital microscopy to visualize the HPV response in real time in mice. Anesthetized mice were ventilated through a tracheostomy, and a thoracic window was prepared as described in Materials and Methods section. Imaging of medium-sized PAs of 20 to 50 μm in diameter, which we have shown previously to yield a robust diameter response to hypoxia, allowed us to visualize HPV directly (fig. 2). In control C57Bl/6 mice (Trpv4+/+), hypoxic ventilation resulted in an approximately 10% decrease in arteriolar diameter. In contrast, Trpv4−/− mice had two thirds less vasoconstriction in response to hypoxia than wild-type controls. These data indicate that TRPV4 plays a critical role in HPV not only in isolated lungs but also in vivo.
Although these data clearly indicate that TRPV4 is required for HPV in both the intact lung and the whole animal, the functional relevance of this channel had not yet been established. To assess the physiological effect of TRPV4 in HPV, we used an in vivo model of induced ventilation-perfusion (V/Q) mismatch. In this system, the carotid artery of anesthetized, ventilated mice is cannulated to facilitate arterial blood gas analysis. After a baseline measurement, a small volume (25 μl) of 0.9% saline solution is instilled intratracheally. These saline droplets then occlude small airways in both lungs, inducing V/Q mismatch.2,13 A further arterial blood gas sample is then analyzed 2 min later to assess the degree of hypoxemia in the animal. In a control animal, an appropriate HPV response should be invoked to minimize hypoxemia in response to the induced V/Q mismatch. However, if TRPV4 is of functional importance in HPV, a Trpv4−/− mouse should demonstrate worsened hypoxemia after saline instillation. This is, in fact, the case (fig. 3A). In control mice, 2 min after V/Q mismatch induction, paO2 decreased to 85% of baseline. In mice lacking functional TRPV4 expression, however, paO2 dropped to 71% of baseline, indicating a failure to compensate for V/Q mismatch via HPV. These data demonstrate that functional loss of TRPV4 attenuates an animal’s ability to invoke HPV in response to V/Q mismatch.
To directly assess the degree of V/Q mismatch, we next measured the distribution of blood flow between both lungs during one-lung ventilation. To this end, mice were ventilated via tracheostomy, and the endotracheal tube was advanced into the left mainstem bronchus to allow for ventilation of the left lung only. After 5 min of one-lung ventilation, fluorescent microspheres of 15 μm diameter were introduced into the pulmonary circulation via the right internal jugular vein. Microsphere fluorescence in each lung, therefore, yields a quantitative measure of blood flow to that organ. In a control animal, one would expect one-lung ventilation to result in the redistribution of blood flow to the ventilated side. However, if HPV is dysfunctional, this redistribution should be attenuated. Our results were consistent with this notion (fig. 3B); after one-lung ventilation, control animals had a 36% difference in pulmonary blood flow between the ventilated and nonventilated lungs. In contrast, Trpv4−/− mice demonstrated only a 22% blood flow discrepancy between lungs, indicating a failure to redistribute pulmonary blood flow appropriately in response to regional alveolar hypoxia during one-lung ventilation.
TRPV4 forms a channel for bivalent cations and is known to be highly permeable to Ca2+.34 Therefore, we next directly measured the role of TRPV4 for the PASMC Ca2+ response to hypoxia. Primary human PASMC were loaded with the ratiometric Ca2+ indicator, Fura 2-AM, and were imaged during perfusion with normoxic or hypoxic buffer solutions. As expected, perfusion with hypoxic buffer produced a significant increase in the intracellular Ca2+ concentration ([Ca2+]i) (fig. 4). To study the role of TRPV4 in the observed hypoxia-induced [Ca2+]i response, we perfused PASMC with 20 μM HC-067047 for 10 min before hypoxia and continued perfusion with the inhibitor throughout the hypoxic period. When TRPV4 activity was inhibited, we observed a significant attenuation of the maximal [Ca2+]i increase during hypoxia. These data suggest that the characteristic [Ca2+]i response to hypoxia is mediated, in large part, by TRPV4. Intracellular Ca2+ is a critical factor in several signaling events during HPV, most importantly, the activation of MLC kinase and the subsequent phosphorylation of MLC. This is a critical regulatory step in HPV because MLC phosphorylation is required for cross-bridge formation and contraction of the PASMC and, therefore, vasoconstriction.1 Acute exposure of PASMC to hypoxia resulted in robust phosphorylation of MLC. However, pretreatment with the TRPV4 inhibitor, HC-067047, attenuated this effect (fig. 4, D and E). This finding suggests that Ca2+ influx through TRPV4 is involved in MLC kinase activation and subsequent phosphorylation of MLC as a necessary precondition for PASMC contraction and vasoconstriction.
TRPV4 has several known biological agonists,26 including the arachidonic acid metabolites of the EET family.35 Interestingly, EET species, such as 11,12-EET, are produced in the lung during acute hypoxia and can activate TRPC6 in PASMC as part of the HPV response.36 Thus, we hypothesized that increased abundance of EET may similarly activate TRPV4 in PASMC. To test this hypothesis, we imaged PASMC loaded with the confocal Ca2+ indicator, Fluo-4. Stimulation with 3 nMol 11,12-EET caused a robust increase in [Ca2+]i in these cells, which could be partially blocked by TRPV4 inhibition (fig. 5A).
Epoxyeicosatrienoic acids can also activate TRPC6, which has previously been shown to be critically involved in HPV in that Trpc6−/− mice largely lack the lung vasoconstrictive response to hypoxia.13 We therefore tested next whether TRPV4 and TRPC6 contribute independently to the PASMC Ca2+ response to EETs or whether their roles may be functionally linked. Notably, TRPC6 and TRPV4 share a series of similarities that make a potential functional link particularly interesting. First, both are activated by EETs and inhibited by cyclic guanosine monophosphate.21,37 Second, in sensory neurons, TRPV4 and TRPC6 have been shown to have a cooperative function in nociception.38 Indeed, different TRP channel subunits can associate to form heteromeric mature channels,39 thus providing a structural correlate to potential functional interdependencies. To test the role of TRPC6 in EET-stimulated Ca2+ flux in our system, we treated PASMC with the TRPC inhibitor SKF (30 μM) before stimulating the cells with 11,12-EET. SKF attenuated the 11,12-EET-stimulated increase in [Ca2+]i to a similar degree as seen with TRPV4 inhibition. Yet, a combination of both TRPC and TRPV4 inhibitors showed no additive effect, suggesting functional cooperation between TRPC6 and TRPV4 in the PASMC Ca2+ response. These data are consistent with, but not limited to, a model whereby TRPC6 and TRPV4 form a heteromeric channel in the PASMC membrane.
To further probe this model, we performed coimmunoprecipitation experiments in whole-lung lysates collected during normoxic or hypoxic ventilation (fig. 5B). During both normoxia and hypoxia, TRPC6 protein could be detected after immunoprecipitation of TRPV4. Omission of TRPV4 antibody resulted in loss of the TRPC6 signal, demonstrating the specificity of TRPC6 binding to TRPV4. These data demonstrate that TRPV4 and TRPC6 form a structural complex in the mouse lung both at normoxia and hypoxia.
To gain insights into the mechanism by which TRPV4 may become activated during HPV, we probed for the subcellular localization of TRPV4 in PASMC during hypoxia. Previous work has demonstrated the importance of caveolae in lung vascular biology.40 Caveolae are plasma membrane microdomains that form nuclei for complex signaling networks between proteins and lipids. Key mediators of endothelial and vascular smooth muscle biology, for example, endothelial nitric oxide synthase, are found in caveolae and require this localization for normal function. Association with caveolae can also be dynamic and may be a consequence of stimulation. TRPV4 has previously been found within caveolae in human umbilical vein endothelial cells.41 In addition, TRPC6 has previously been shown to translocate to caveolae during hypoxia in PASMC.42 The proposed interaction and functional cooperation between TRPV4 and TRPC6 would therefore suggest that TRPV4 in PASMC may similarly translocate to caveolae in response to hypoxia. Using sucrose gradient ultracentrifugation to purify caveolae, we found a significant redistribution of TRPV4 to caveolin-1-containing cellular fractions in response to 15 min of hypoxia (fig. 6). There was an approximately two-fold increase in the amount of TRPV4 associated with caveolae during hypoxia compared with controls. This finding sets up an intriguing possible scenario, in which TRPV4 activation and signaling is modulated during hypoxia by cooperation with TRPC6 and affiliation with the unique protein and lipid environment of caveolae.
Discussion
Our current studies have established a critical role for TRPV4 in the pulmonary vascular response to hypoxia. Using in vitro, in vivo, and whole-organ approaches, we have demonstrated that TRPV4 is required for a normal HPV response and that the loss of TRPV4 function is associated with deficient oxygenation in animals challenged with V/Q mismatch. In addition, we have shown that during one-lung ventilation—a necessary step for a variety of cardiothoracic procedures—TRPV4 is required for maximal redistribution of pulmonary blood flow to the ventilated lung. Finally, we have determined that TRPV4 is required for the hypoxia-triggered [Ca2+]i increase and MLC phosphorylation in isolated human PASMC, that TRPV4 can form protein–protein complexes in these cells with TRPC6, and that hypoxia results in the redistribution of TRPV4 to plasma membrane microdomains.
The polymodal cation channel TRPV4 is well positioned to play a pivotal role in HPV. It is highly expressed in PASMC19 and in pulmonary endothelial cells,43 both of which are required for HPV.1,2 Furthermore, TRPV4 is activated by EETs,20 the levels of which are increased in the lung during acute hypoxia, whereas depletion of EETs has been shown to diminish the HPV response.36 In the systemic vasculature, TRPV4 has previously been shown to be involved in the regulation of vasodilatory responses: mesenteric vessels express TRPV4, and such vessels isolated from Trpv4−/− mice have a diminished vasodilatory response to acetylcholine.16 In vessel myographs, TRPV4 agonists relax large conducting renal arteries, mesenteric arteries, and vasa recta and vascular dilation in response to TRPV4 agonists in isolated perfused kidneys is abrogated in Trpv4−/− mice.15 In contrast to this documented role of TRPV4 in vascular dilation, we demonstrate here a key role of TRPV4 as an important prerequisite in pulmonary vasoconstriction. Importantly, the fact that given stimuli such as hypoxia, serotonin, or EETs cause opposite effects with respect to vasomotor tone between the pulmonary and systemic circulations is well documented,44 and the divergent effects of TRPV4 on lung and systemic vascular tone may provide a possible mechanistic explanation for some of these divergences. Given the broad expression pattern of TRPV4, a tissue-specific knockout mouse would be an ideal animal model for studying the role of TRPV4 in individual tissues. However, to our knowledge, such an animal model has not been constructed at this time.
In line with the demonstrated role of TRPV4 as a bivalent cation channel, our findings suggest that the functional role of TRPV4 in HPV relates to its ability to mediate the characteristic [Ca2+]i increase in PASMC in response to hypoxia that has previously been shown to be critical for hypoxia-induced PASMC contraction.1 The mechanisms by which hypoxia causes activation of TRPV4 remain to be elucidated but may involve formation of EET, which have previously been shown to activate TRPV4,22 and to mediate HPV,36 and/or activation of src kinase which similarly has been shown to activate TRPV4 via phosphorylation of its tyrosine residues45 and also contributes to the lung vascular response to hypoxia.46 In addition, we show here that TRPV4 activation during hypoxia is associated with its translocation to caveolae. This finding complements a series of striking similarities between two different members of the TRP family in the regulation of HPV, namely TRPV4 and TRPC6. Similar to TRPV4, TRPC6 has previously been shown to contribute critically to HPV and to mediate the PASMC [Ca2+]i response to hypoxia.13 In addition, TRPC6 is—similar to TRPV4—activated by EETs47 and translocates to caveolae in response to hypoxia.36 This analogy between the function and regulation of TRPV4 and TRPC6 hence bears the question that whether both channels may in fact be functionally linked in the context of HPV. Indeed, functional cooperation between both channels was recently reported in the context of nociceptor sensitization, in that antisense to TRPC6 reversed the mechanical hyperalgesia induced by the TRPV4-selective agonist 4αPDD.38 Interestingly, TRP channels are composed of four subunits that commonly assemble to homomeric but occasionally heteromeric pore-forming channels, which may explain cooperativity between different TRP channels.6,48 TRPV4 has been shown to form a heteromer with TRPC1, and this heteromerization increases the surface expression and calcium conductance of TRPV449 and allows TRPV4 to function as a store-operated calcium channel, essentially bestowing a function associated with TRPC proteins upon TRPV4. By immunoprecipitation analyses in PASMCs, we show here that TRPV4 can similarly engage in protein–protein interactions with TRPC6, which may form the molecular basis for the observed parallel roles of both channels in HPV, and for their nonadditive effects in the PASMC [Ca2+]i response to EET. Our finding that the interaction between TRPV4 and TRPC6 in PASMC remained unaltered during hypoxia further suggests that both channels form heteromers already at baseline in PASMC and become recruited to caveolae (and presumably, activated) conjointly upon exposure to hypoxia.
Acknowledgments
Supported by an open operating grant from the Canadian Institutes of Health Research (Ottawa, Ontario, Canada; to Dr. Kuebler) and by Resident Research Grants from The PSI Foundation (Toronto, Ontario, Canada) and The Canadian Anesthesiologists’ Society (Toronto, Ontario, Canada; to Dr. Goldenberg).
Competing Interests
The authors declare no competing interests.