In recent years, transgenic mouse models have been developed to examine the underlying cellular and molecular mechanisms of lung disease and pulmonary vascular disease, such as asthma, pulmonary thromboembolic disease, and pulmonary hypertension. However, there has not been systematic characterization of the basic physiological pulmonary vascular reactivity in normal and transgenic mice. This represents an intellectual “gap”, since it is important to characterize basic murine pulmonary vascular reactivity in response to various contractile and relaxant factors to which the pulmonary vasculature is exposed under physiological conditions. The present study evaluates excitation- and pharmacomechanical-contraction coupling in pulmonary arteries (PA) isolated from wild-type BALB/c mice. We demonstrate that both pharmaco- and electromechanical coupling mechanisms exist in mice PA. These arteries are also reactive to stimulation by α1-adrenergic agonists, serotonin, endothelin-1, vasopressin, and U-46619 (a thromboxane A2 analog). We conclude that the basic vascular responsiveness of mouse PA is similar to those observed in PA of other species, including rat, pig, and human, albeit on a different scale and to varying amplitudes.
- store depletion
- excitation-contraction coupling
- G protein-coupled receptors
mice, especially transgenic or knockout variants, are increasingly being used to study disease mechanisms. With respect to pulmonary hypertension, transgenic or knockout mouse models have been developed to evaluate the modulation of pulmonary vasoconstriction and remodeling due to 1) endothelial nitric oxide synthase disruption (18, 47, 57); 2) vasoactive and mitogenic agonists, such as hypoxia-inducible factors-1α (63) and -2α (2), calcitonin gene-related peptide (5), serotonin (5-HT) (9, 32, 34, 45), matrix metalloproteinases (67), transforming growth factor-β (35), and vasoactive intestinal peptide (51); 3) bone morphogenetic protein receptor type II gene mutations and altered bone morphogenetic protein signaling (19, 28, 40, 62); 4) altered function of ion channels (12) and transporters (49); and 5) altered superoxide production (36, 38). Despite these studies, there is still relatively little information regarding the basic characterization of the pulmonary vascular reactivity in normal mice. Nor have the excitation-contraction coupling mechanisms, pharmacological properties of vasoactive response to agonists, been evaluated completely in wild-type mice. This represents an intellectual gap, since it is important to characterize basic murine pulmonary vascular reactivity in response to various contractile and relaxant factors to which the pulmonary vasculature is exposed under physiological conditions. The aim of this study was to characterize pulmonary arterial (PA) excitation-contraction coupling and pharmacological properties of PA using isolated PA rings from mice.
METHODS AND MATERIALS
Isolation of PA rings.
Male 5- to 8-wk-old BALB/c mice were used in this study. Use of mice for the experiments presented in this study was approved by the Institutional Animal Care and Use Committee (IACUC) of the University of California, San Diego. After decapitation using a procedure approved by the IACUC, the lungs were quickly removed from the mouse, washed with cold saline (to removed blood from the lung tissue), and placed in a dissection plate. Under a stereomicroscope, the right and left branches of the intrapulmonary arteries (third and fourth divisions) were isolated from the mice within 45 min from the time when the lungs were removed from the decapitated mice. Adipose and connective tissues were carefully removed with fine forceps and ophthalmological scissors, and the remaining arteries were cut into 1.5-mm-long rings. The arterial rings in which the vascular wall was damaged or contained small branches were discarded; only rings with an intact wall were used for our experiments.
Two tungsten hooks (75-μm diameter) were inserted through the lumen of the PA rings. One hook was mounted on the bottom of a perfusion chamber, and the other was connected to an isometric force transducer (Harvard Apparatus). Isometric tension was continuously monitored and acquired using DATAQ software (DATAQ Instruments). Unless indicated otherwise, resting passive tension was set at 300 mg, and the rings were equilibrated for 1 h at resting tension and then challenged three times with 40 mM K+ (40K) perfusate to obtain a stable contractile response.
Isolated PA rings were superfused with modified Krebs solution (MKS; at 37°C) consisting of the following (in mM): 138 NaCl, 1.8 CaCl2, 4.7 KCl, 1.2 MgSO4, 1.2 NaH2PO4, 5 HEPES, and 10 glucose (pH 7.4). For Ca2+-free solution, CaCl2 was replaced by equimolar MgCl2, and 1 mM EGTA was added to chelate residual Ca2+. In the high-K+ solution, NaCl was replaced by equimolar KCl to maintain osmolarity. The active tension induced by agonists was normalized by the basal tension and expressed as net increase in tension (mg).
Chemicals and reagents.
4-Aminopyridine (4-AP), sodium nitroprusside (SNP; Fluka), 5-HT, acetylcholine (ACh), and bradykinin (BK) were dissolved directly into MKS on the day of use. Nifedipine (Nif) and cyclopiazonic acid were prepared as 1 and 30 mM stocks, respectively, in dimethyl sulfoxide. U-46619 was prepared as a stock in ethanol. Phenylephrine (PE), endothelin-1 (ET-1), and vasopressin (VP) were prepared as concentrated stock solutions in distilled water. All stock solutions were aliquoted and kept frozen at −20°C until use. On the day of experiments, aliquots of the stock solutions were dissolved in MKS to the proper concentrations. The pH values of all solutions were measured after addition of the drugs and readjusted to 7.4. All drugs were from Sigma Chemical, unless otherwise indicated.
After contraction experiments, the PA rings were fixed overnight in 10% neutral buffered formalin. The formalin-fixed rings were serially cut for microscopic examination. The vessel tissues were processed and embedded in paraffin blocks in an automatic tissue processor (Sakura Tissue-Tek VIP; Sakura Finetek). The paraffin-embedded tissues were cut in 5-μm-thick sections for staining with either lectin (endothelium specific) or α-actin (smooth muscle specific).
The composite data are expressed as means ± SE. Statistical analysis was performed using paired or unpaired Student's t-test or ANOVA and post hoc tests (Student-Newman-Keuls) where appropriate. Differences were considered to be significant at P < 0.05.
Morphological feature of PA isolated for contraction experiments.
The PA rings used in this study were isolated from third- to fourth-order intrapulmonary arterial branches (Fig. 1A). Internal diameter of the arteries ranged from 200 to 500 μm. Bright-field and fluorescent images (Fig. 1B) show that the vessel wall was intact, with a distinct intraluminal wall composed of an endothelial cell monolayer, which can be stained with lectin and multiple layers of smooth muscle cells identified by α-actin staining (Fig. 1B, right). These structural characteristics are similar to those our laboratory previously identified in rat PA rings (31).
Zero-stress state of left and right branches of PA.
If a blood vessel is cut open or all loads are removed, the circular shape opens up, and the wall becomes a sector; this is the blood vessel in its zero-stress state (20, 59). This illustrates that, compared with the in vivo or no-load (i.e., no transmural pressure or longitudinal stress) states, there are residual stresses and strains in blood vessels.
Figure 2A depicts images of left and right PA in their native state, as well as after being cut longitudinally to relieve stresses. The zero-stress state after cutting was calculated as the angle (θ) of opening of the resting artery. Despite evidence that the opening angles at the zero-stress states vary in third- to ninth-order human PA (27), we found that the opening angles at the zero-stress states were comparable between left and right branches of intrapulmonary arteries (94.2 ± 9.10 vs. 88.2 ± 7.6°; P = 0.6218).
Optimal basal tension for high K+ and agonist-mediated PA contraction.
An optimal basal tension is required for blood vessels to obtain maximal active contraction. In isolated mouse PA rings, increasing the basal tension from 100 to 300 mg significantly augmented the 40K-induced active tension (Fig. 3). Increasing the basal tension from 300 to 600 mg did not significantly change the active tension, whereas basal tension beyond 600 mg actually reduced the 40K-mediated active tension (Fig. 3B). These data suggest an optimal 300-mg basal tension for the mouse intrapulmonary arteries with diameters of 200–500 μm. Therefore, we used 300-mg basal tension for all of the following experiments.
Dose-response curves of high K+- and PE-induced pulmonary vasoconstriction.
Two basic excitation-coupling mechanisms control pulmonary vascular tone: 1) electromechanical coupling, the mechanism that causes contraction or relaxation through changes in membrane potential; and 2) pharmacomechanical coupling, the complex of mechanisms that can cause contraction or relaxation by mechanisms not mediated by changes in membrane potential (56). We verified that pathways leading to both contractile mechanisms were intact in isolated PA rings and that they contribute to the regulation of vascular tone in mouse PA rings.
Raising extracellular K+ concentration ([K+]o) (e.g., from 4.7 to 40 mM) shifts the K+ equilibrium potential and causes membrane depolarization. Subsequent opening of voltage-dependent Ca2+ channels promotes Ca2+ influx, increases cytosolic Ca2+ concentration, and causes vasoconstriction. In PA rings with a 300-mg basal tension, increasing [K+]o from 4.7 (in MKS) to 10, 20, 40, 60, 80, and 120 mM significantly increased the amplitude of active tension. The dose-response curve was almost linear at [K+]o between 20 and 40 mM and gradually stabilized (or approached to the plateau phase) when [K+]o reached 60–80 mM. The EC50 for high-K+-mediated active tension in mouse PA rings is ∼35 mM (Fig. 4A).
The α-adrenergic receptor agonist, PE, also caused vasoconstriction in mouse PA rings. The minimal dose of PE to induce PA contraction was 2 nM. The steepest increase of active tension occurred at doses between 5 and 20 nM; the EC50 for PE-induced pulmonary vasoconstriction was ∼10 nM (Fig. 4B). The data obtained from these experiments indicate that electromechanical and pharmacomechanical coupling mechanisms are both involved in the regulation of mouse pulmonary vascular tone.
Ca2+ dependence of high-K+- and agonist-mediated PA contraction.
Extracellular Ca2+ influx is required for both electro- and pharmacomechanical coupling to occur. Removal of extracellular Ca2+ almost abolished 25 mM K+-mediated PA contraction, whereas extracellular application of the voltage-dependent Ca2+ channel blocker Nif (100 nM) also significantly inhibited 25 mM K+-induced active tension. As shown in Fig. 5, the 25 mM K+-induced active tension was 257.7 ± 4.2 mg in PA rings superfused with 1.8 mM Ca2+-containing MKS, 14.5 ± 6.8 mg (a 94% inhibition; P < 0.001) in PA rings superfused with Ca2+-free MKS, and 286.0 ± 44.2 mg when extracellular Ca2+ was restored (Fig. 5A). Furthermore, the 25 mM K+-induced active tensions were 325.0 ± 29.3, 22.0 ± 14.4 (a 93% inhibition), and 269.0 ± 35.5 mg before, during, and after application of Nif (Fig. 5B). These data indicate that 25 mM K+-induced PA contraction is mainly due to membrane depolarization-mediated Ca2+ influx through Nif-sensitive, voltage-dependent Ca2+ channels in PA smooth muscle cells (PASMCs).
In PA rings precontracted with PE (200 nM), removal of extracellular Ca2+ reversibly reduced the active tension from 484.0 ± 49.8 to 31.0 ± 24.5 mg (a 94% inhibition; P < 0.001) (Fig. 5C). In contrast to our observation with 25 mM K+, extracellular application of Nif (100 nM) reduced the PE-induced active tension only by 29% (from 523.5 ± 47.7 to 369.4 ± 54.1 mg; P < 0.001) (Fig. 5D). From these data, we conclude that PE-induced PA contraction not only involves Ca2+ influx through Nif-sensitive, voltage-dependent Ca2+ channels (e.g., L- and T-type channels), but also involves Ca2+ mobilization from intracellular stores (e.g., sarcoplasmic or endoplasmic reticulum) and Ca2+ influx through Nif-insensitive Ca2+ channels.
Activation of various receptors causes comparable PA contraction.
In addition to high K+ and PE, extracellular application of 5-HT (5 μM), ET-1 (0.1 μM), VP (10 μM), U-46619 (10 μM, a thromboxane A2 analog), and thrombin (10 U/ml) also reversibly constricted murine PA rings (Fig. 6, A–E). The level of maximal active tension and the kinetics of tension rise both vary dramatically among different agonists (Fig. 6, F and G). For example, 5-HT- and U-46619-mediated contraction are composed of a rapid initial increase in tension followed by a slow tension increase up to a plateau; in comparison, the thrombin-mediated initial increase in tension seems to be much slower than 5-HT- and U-46619-mediated tension increase (Fig. 6G). The amplitude of active tension induced by ET-1 and VP was much smaller than that induced by other agonists tested, but the kinetics of tension rise (determined by the time required for tension to reach the maximal level) was much faster than thrombin-mediated contraction (Fig. 6G, left). Furthermore, 5-HT- and U-46619-mediated PA contraction seemed to be comparable in terms of the level of maximal active tension and the kinetics of tension development (Fig. 6G, right). Moreover, 5-HT- and U-46619-mediated active tension was also similar to thrombin-mediated PA contraction, but the rise in tension in 5-HT- and U-46619-mediated contraction was much faster than that induced by thrombin (Fig. 6G).
Inhibition of membrane K+ channels induces membrane depolarization and causes PA contraction.
Membrane depolarization can be caused either by raising [K+]o or by blockade of K+ channels that are active under the resting conditions. In human and rat PASMC, inhibition of K+ channels, especially voltage-gated K+ channels, with 4-AP causes membrane depolarization, opens voltage-dependent Ca2+ channels, and increases cytosolic Ca2+ concentration (64, 65). As shown in Fig. 7, treatment of isolated mouse PA rings with 10 mM 4-AP significantly increased active tension. Removal of extracellular Ca2+ markedly abolished 4-AP-induced PA contraction, indicating the contraction was mainly induced by Ca2+ influx through voltage-dependent Ca2+ channels (data not shown). The 4-AP-induced active tension in mouse PA was associated with oscillatory contraction (Fig. 7B), which has never been demonstrated in rat PA rings. The oscillation is mainly due to a balance between Ca2+ influx and extrusion and/or Ca2+ release and sequestration. The 4-AP-induced oscillatory contraction thus suggests that inhibition of voltage-gated K+ channels and membrane depolarization may also indirectly activate mechanisms or pathways for Ca2+ extrusion (e.g., via the forward mode of Na+/Ca2+ exchange) and sequestration (e.g., via sarcoplasmic/endoplasmic reticulum Ca2+-Mg2+-ATPase) in mouse PASMCs.
Inability of ACh to cause vasodilation in isolated mouse PA rings.
In human and rat PA rings, ACh is a potent endothelium-dependent vasodilator that activates endothelial muscarinic receptors, triggers nitric oxide (NO) synthase, and increases NO production and release. In isolated mouse PA rings with intact endothelium, however, ACh (10 μM) further increased active tension in mouse PA rings precontracted with high K+ (25 and 40 mM K+) (Fig. 8, A and B) or 20 nM PE (Fig. 8C).
Histological data showed that the endothelial cells were present in the arterial rings used for the experiments (Fig. 9A). Extracellular application of BK (20 μM), an endothelium-dependent vasodilator, significantly reduced active tension in PA rings preconstricted with 40K (Fig. 9, B and C), indicating that the endothelium is functionally normal in the isolated PA rings used in this study.
Although ACh was unable to cause pulmonary vasodilation in rings with an intact endothelium, extracellular application of the NO donor SNP (100 nM) significantly reduced (67.4 ± 7.8% reduction) 25 mM K+- and PE (200 nM)-induced PA contraction (Fig. 10). The IC50 for SNP-mediated PA relaxation is ∼5 nM (Fig. 10B), indicating that mouse PA smooth muscle is indeed sensitive to NO. The inability of ACh to cause vasodilation in endothelium-intact PA rings suggests that 1) ACh-induced muscarinic receptor activation is not enough to activate endothelial NO synthase (eNOS) in mouse PA endothelium; 2) the expression level and basal activity of eNOS are very low in mouse PA endothelium; 3) NO production in mouse PA endothelium can be rapidly deactivated or scavenged by potentially high levels of reactive oxygen species in endothelial and smooth muscle cells; and 4) the level of muscarinic receptors in mouse PASMCs is much higher than in PA endothelial cells.
Potential role of pulmonary vascular reactivity in pathogenic and therapeutic mechanisms involved in pulmonary vascular disease using the mouse vasculature is a relatively novel field of study, most notably due to the development of transgenic animals. A significant number of studies have evaluated the vasoreactivity of aorta to K+ (39, 58), angiotensin II (58, 71), U-44619 (48, 58), sphingosine 1-phosphate (41), ATP (23), UTP (23), ACh (6, 39, 72), nitric oxide (26), and PE (39). In contrast, only a few studies have examined the mechanical properties of mouse PA in response to vasoactive stimuli, such as K+ (30), Nif (30), levcromakalim (30), 5-HT (37), ACh (57), and noradrenaline (70). In many cases, vasoreactivity was tested in transgenic animals, but not in their wild-type littermates. Our data present comprehensive information regarding basal electro- and pharmacomechanical coupling in isolated mouse PA rings.
The optimal basal tension in our preparations (200–500 μm in diameter) was ∼300 mg. This value is within the 200–1,000 mg optimal tension range used in other mouse PA studies (30, 37, 57, 70); the wide range is likely due to variability in the size of the arteries and strain of mice used. That said, the crux of our study rests on the vasoreactivity of mouse PA rings to different stimuli. Our findings suggest that the mouse PA bears similar electropharmacological mechanical properties to PA rings from rats (7, 13, 22, 42, 44, 46, 54, 60, 61, 68) and humans (24, 55). For example, membrane depolarization by high K+ caused a slow and sustained contraction in mouse PA rings; the kinetics are very similar to that observed in rat PA (1, 54, 66, 68) and human (55) PA rings. Removal of extracellular Ca2+ and application of Nif almost abolished the high K+-mediated PA contraction, as in rat (66). However, while PE induced contraction similarly in mouse and rat PA (44, 52), removal of extracellular Ca2+ almost abolished the PE-induced contraction, but Nif treatment only reduced the active tension by 60%, indicating that Ca2+ resources other than influx via voltage-dependent Ca2+ channels are involved in agonist-mediated PA contraction.
Murine PA responses to other agonists were also similar to that observed in isolated PA rings from other species. For example, SNP produced endothelium-independent vasodilation of mouse PA rings precontracted with either 25 mM K+ or 200 nM PE, similar to what has been reported previously for rat PA (54, 66, 68). However, ACh failed to induce vasodilation in mouse PA rings with intact endothelium (while BK significantly inhibited the active tension in the PA rings), which is quite different from the results obtained from PA rings isolated from humans, rats, and dogs. This may suggest that the NO signaling system is intact in mouse PA, but that the activation of muscarinic receptors in endothelium by ACh is not sufficient (or masked by the contractile effect of ACh-mediated activation of muscarinic receptors in PASMCs) to activate this signaling cascade. In addition to ACh-mediated eNOS activation and NO production in endothelium, ACh-mediated production of endothelium-derived hyperpolarizing factors (EDHF) also contributes to ACh-induced vasodilation in many vessel types isolated from humans, rats, and dogs (3). However, it is unclear whether ACh-induced production or release of EDHF (e.g., K+ and epoxyeicosatrienoic acids) (3, 4, 14) in endothelial cells and/or EDHF-induced activation of K+ channels in smooth muscle cells is different between PA from mice and other species (e.g., rats, dogs, and humans).
The contractile response we observed in mouse PA rings to vasoactive substances like ET-1, 5-HT, and the thromboxane A2 analog U-44619 were characteristically similar to that observed in PA from other species (8, 10, 11, 24, 25, 33, 37, 42, 43, 52, 55, 60, 69). The contractile response to thrombin observed in mouse PA rings was also similar to that observed by others in porcine PA rings (21). However, canine PA do not respond to thrombin with a change in tension (29), while human PA dilate following exposure to thrombin (24). The thrombin response relies on the presence of endothelium and nature of G protein-coupled protease-activated receptors on the plasma membrane. Therefore, it is possible that the protease-activated receptors in mouse and canine PA differ from those in human and canine PA in terms of function, expression, and distribution (e.g., in PA smooth muscle and endothelial cells).
The response to VP represented the only deviation. In isolated mouse PA, we observed a slight, yet significant increase in tension (Fig. 6), which was completely reversible upon washout of VP. In contrast, VP has been reported to cause endothelium-dependent relaxation of PA isolated from rats (15, 16), humans (53), and dogs (17, 50), presumably due to VP inducing the release of endothelium-derived NO. In our experiments with mouse PA, the endothelium was intact; therefore, precisely why the tensile response to VP was opposite between PA isolated from mice and other species (e.g., rat, humans, and dogs) is unclear.
From our study, we can conclude that the physiological responses of mouse PA are highly similar to those previously reported in PA preparations from other species. The only exceptions are 1) the contractile response to VP and thrombin, and 2) the inability of ACh to cause vasodilation in endothelium-intact arteries. While no explanation is currently available to explain these latter differential responses, we have shown that the mouse represents an excellent study model for the evaluation of pulmonary vascular functions, such as excitation-contraction coupling, vasoreactivity, and drug response.
This work was supported in part by grants from the National Heart, Lung, and Blood Institute (HL064945, HL054043, and HL066012).
We thank A. Nicholson for technical assistance.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 by the American Physiological Society