Pulmonary arteries from the Madison (M) strain relax more in response to acetylcholine (ACh) than those from the Hilltop (H) strain of Sprague-Dawley rats. We hypothesized that differences in endothelial nitric oxide (NO) synthase (eNOS) expression and function, metabolism of ACh by cholinesterases, release of prostacyclin, or endothelium-derived hyperpolarizing factor(s) (EDHF) from the endothelium would explain the differences in the relaxation response to ACh in isolated pulmonary arteries. eNOS mRNA and protein levels as well as the NO-dependent relaxation responses to thapsigargin in phenylephrine (10−6 M)-precontracted pulmonary arteries from the M and H strains were identical. The greater relaxation response to ACh in M compared with H rats was also observed with carbachol, a cholinesterase-resistant analog of ACh, a response that was not modified by pretreatment with meclofenamate (10−5M). N ω-nitro-l-arginine (10−4 M) completely abolished carbachol-induced relaxation in H rat pulmonary arteries but not in M rat pulmonary arteries. Precontraction with KCl (20 mM) blunted the relaxation response to carbachol in M rat pulmonary arteries and eliminated differences between the M and H rat pulmonary arteries. NO-independent relaxation present in the M rat pulmonary arteries was significantly reduced by 17-octadecynoic acid (2 μM) and was completely abolished by charybdotoxin plus apamin (100 nM each). These findings suggest that EDHF, but not NO, contributes to the strain-related differences in pulmonary artery reactivity. Also, EDHF may be a metabolite of cytochrome P-450 that activates Ca2+-dependent K+ channels.
- nitric oxide
- cytochrome P-450
- calcium-dependent potassium channels
- potassium chloride
- endothelium-derived hyperpolarizing factor
pulmonary vascular responsiveness may differ substantially between species, strains, or even individuals. Understanding the mechanism(s) behind these differences may provide insight into the susceptibility of some individuals to develop pulmonary hypertension. Our previous studies have shown that the Madison (M) strain of Sprague-Dawley rat has greater pulmonary vasoresponsiveness to acute hypoxia than the Hilltop (H) strain (13, 24) but then has an attenuated pulmonary hypertensive response to chronic hypoxia (23, 25). Using isolated pulmonary arteries, we demonstrated that the pulmonary vascular endothelium contributes to these differences (27). We found that the endothelium-dependent vasodilator acetylcholine (ACh) relaxes phenylephrine-precontracted pulmonary arteries from M rats more than those from H rats, perhaps offering insight into the mechanism for the blunted pulmonary hypertensive response of the M rats to chronic hypoxia. When it is considered that nitric oxide (NO) release from the endothelium appears to be the primary mechanism by which ACh relaxes rat pulmonary arteries (33), the greater relaxation to ACh in M rats may be due to greater endothelial NO synthase (eNOS) expression and activity. This hypothesis is supported by recent studies showing that alterations in endogenous NOS activity can contribute to strain-related differences in various experimental models (1, 14, 19, 29).
Besides NO, the pulmonary vascular endothelium also releases endothelium-derived hyperpolarizing factor(s) (EDHF) in response to ACh (3). Because EDHF appears to mediate 20–25% of the total relaxation induced by ACh in pulmonary arteries (3), it is possible that the greater relaxation to ACh in M rats is due to differences in EDHF release. Other possible mechanisms for the augmented vasodilator response in M rats include diminished hydrolysis of ACh by intramural pulmonary arterial cholinesterases or concomitant release of other endothelium-derived vasodilator agents such as prostacyclin (PGI2) or some combination of these mechanisms.
In the present study, we hypothesized that either eNOS mRNA and protein levels or EDHF release is greater in M rat compared with H rat pulmonary arteries and that the relaxation response to non-receptor-mediated endothelium-derived NO-dependent vasodilators such as thapsigargin would be enhanced in M rat pulmonary arteries. We also examined the possibility that other potential mechanisms such as reduced cholinesterase activity or enhanced release of PGI2might be playing a role.
MATERIALS AND METHODS
Animals and materials.
Adult male Sprague-Dawley rats (4–6 wk old, 250–275 g) were obtained from Harlan Sprague Dawley Laboratories, Madison, WI (M rats) and Hilltop Laboratories, Scottsdale, PA (H rats). All reagents were purchased from Sigma (St. Louis, MO) except where otherwise stated. The following vasoactive agents were used: l-phenylephrine, carbamylcholine chloride (carbachol), meclofenamic acid,N ω-nitro-l-arginine (l-NNA), potassium chloride (Fisher Scientific, Fair Lawn, NJ), thapsigargin (Calbiochem, La Jolla, CA), and charybdotoxin and apamin (Alomone Labs, Jerusalem, Israel). All drug concentrations are expressed as the final molar concentration in the organ bath. Unless stated otherwise, drugs were dissolved in deionized water. Thapsigargin was first dissolved in DMSO, followed by dilution in deionized water. 17-Octadecynoic acid (17-ODYA) was dissolved in ethanol.
Isolated artery preparation.
Rats were anesthetized with pentobarbital sodium (100 mg/kg ip) and exsanguinated by cutting of the abdominal aorta. Heart and lungs were removed en bloc and placed in oxygenated Earle's balanced salt solution (EBSS) containing (in mM) 116.3 NaCl, 5.4 KCl, 0.83 MgSO4, 19.0 NaHCO3, 1.04 NaH2PO4, 1.8 CaCl2 · 2H2O, and 5.5d-glucose. The main intralobar pulmonary artery from the left lung and the middle lobe of the right lung (∼1.5–2.0 mm ID) were isolated, cut into rings (2- to 3-mm long), and mounted in 10-ml organ chambers filled with EBSS and bubbled with 95% O2-5% CO2, as described previously (8). Force was measured in milligrams by use of isometric force transducers (Grass FT03) and recorded on a Grass polygraph (model 790).
Organ bath experiments.
Arterial rings were allowed to stabilize for 60 min at 1 g of resting tension. The viability of the smooth muscle and endothelium was checked by obtaining a contractile response to phenylephrine (10−6 M) and a subsequent relaxation response to carbachol (10−6 M). After washout of the drugs from the organ chamber and return to baseline tone (20 min), the rings were contracted again with a concentration of phenylephrine (10−6 M) that was previously shown to produce equal responses in M and H rat pulmonary arteries (27). When the contractile response plateaued, concentration-response curves to the eNOS activators thapsigargin and carbachol were obtained. In parallel experiments, which served as a time control, equal volumes of vehicle were added after phenylephrine-induced contraction. In some experiments, arterial rings were preincubated for 30 min with the cyclooxygenase inhibitor meclofenamate (10−5 M) (27), the NOS inhibitor l-NNA (100 μM), l-NNA plus the Ca2+-dependent K+ channel blockers charybdotoxin (100 nM) and apamin (100 nM), or l-NNA plus the cytochrome P-450 inhibitor 17-ODYA (2 μM) (4,12). Artery rings were then precontracted with phenylephrine (10−7 to 10−6 M), and concentration-response curves were obtained for thapsigargin or carbachol.
In some experiments, a second concentration-response curve to carbachol was constructed after precontraction of the pulmonary artery rings with KCl (20 mM). After completion of the protocol, vessel preparations were again tested for viability by washing out drugs, allowing the vessel to return to baseline tone, and once again precontracting with phenylephrine (10−7 M) and obtaining a relaxation response to carbachol (10−4 M).
RT-PCR for eNOS.
The pulmonary artery was dissected as described above and was immediately transferred to a sterile tube containing TRI Reagent (Sigma). Total RNA was extracted from the artery by following manufacturer's protocols, dissolved in diethyl pyrocarbonate water, and quantified by measuring absorbance at 260 nm. By use of Omniscript reverse transcriptase (QIAGEN, Valencia, CA) and random hexamers, 2 μg of total RNA were reverse transcribed to cDNA. PCR was then performed to amplify the eNOS gene by use of eNOS-specific primers 5′-TCCAGAGCATACCCGCACTTC-3′ (sense) and 5′-GTCCAGACGCACCAGGATTG-3′ (antisense). The eNOS-specific primers were designed on the basis of published rat cDNA (GenBank accession no.U02534). Tubulin was amplified in parallel experiments as previously described (16) and was used as a housekeeping gene. Primer sequences for tubulin were 5′-CTCCATCCTCACCACCCACAC-3′ (sense) and 5′-CAGGGTCACATTTCACCATCT-3′ (antisense). eNOS and tubulin were detected in parallel PCR analyses of samples from the same RT reaction. The linear range for PCR amplification was determined in pilot experiments by use of various amounts of RT products from 10 to 500 ng of RNA and from 15 to 40 cycles of PCR amplification. On the basis of these experiments, 100 ng of RNA and 35 cycles of amplification were found to be optimal for eNOS and 30 cycles for tubulin. RNA not reverse transcribed to cDNA was used as a template in some PCR and showed no evidence of genomic DNA contamination. The amplification for eNOS was carried out as follows. The initial cycle used 3 min at 94°C for denaturation; subsequent cycles of PCR were performed under the following conditions: denaturation, 20 s at 94°C; annealing, 30 s at 60°C; and extension, 40 s at 72°C. Final extension was carried out at 72°C for 10 min. For tubulin, 68°C was used for annealing and extension.
Western analysis of eNOS.
Proteins were isolated from the phenol-ethanol supernatant that remained after RNA extraction by means of the TRI Reagent protocols. Protein (50 μg) was separated electrophoretically on 7.5% SDS-PAGE gels and then blotted to a nitrocellulose membrane. Nonspecific binding of the primary and secondary antibodies was blocked by incubating blots in 5% nonfat dry milk made up in TTBS (50 mM Tris · HCl, pH 7.4, 0.15 M NaCl, and 0.1% Tween 20) for 1 h and in 2% BSA in TTBS for 10 min at room temperature. Subsequently, blots were incubated overnight at 4°C with 1:2,000 diluted anti-eNOS monoclonal antibody (Transduction Laboratories, Lexington, KY). Blots were then washed three times (15 min/wash) in TTBS and incubated for 1 h at room temperature with 1:10,000 diluted goat horseradish peroxidase-conjugated antimouse IgG (Transduction Laboratories). Specifically bound eNOS antibody was detected with ECL chemiluminescence substrate (Amersham) and exposure to double-emulsion X-ray film. The intensity of bands was quantified by densitometry.
Results are expressed as means ± SE. Responses to the vasodilator agents are expressed as a percentage of phenylephrine or KCl precontraction. The potency was calculated as the negative logarithm of the concentration causing a 50% relaxation response. Differences between mean values were evaluated using Student's t-test. When more than two means were compared, one-way analysis of variance followed by Fisher's exact test was used to determine statistically significant differences. A P value <0.05 was considered statistically significant.
eNOS expression in pulmonary arteries.
To determine whether the greater relaxation responses to ACh observed in M rat pulmonary arteries compared with those in H rats were associated with increased eNOS expression, eNOS mRNA and protein levels were assayed in pulmonary arteries from M and H rats. As predicted from the published sequences, the eNOS-specific primers amplified a 393-bp product (Fig. 1 A), and tubulin primers amplified a 365-bp product (Fig. 1 B) in cDNA synthesized from pulmonary artery segments of M and H rats. The mean eNOS-to-tubulin ratios in pulmonary arteries from M rats were similar to those in H rats (Fig. 1 C). Western blot analysis using a mouse monoclonal antibody to eNOS detected a band at ∼135 kDa in pulmonary arteries from M and H rats (Fig.2 A). Similar to the RNA levels, no differences in mean eNOS protein levels were seen between pulmonary arteries from M and H rats (Fig. 2 B).
Vasorelaxation response to thapsigargin.
Thapsigargin, a receptor-independent stimulator of eNOS, was used to determine whether enzyme function of eNOS in response to direct stimulation would differ between M and H rats. Thapsigargin at concentrations from 10−7 to 10−5 M elicited concentration-dependent reversal of phenylephrine-induced contraction in pulmonary arteries from M and H rats. The magnitude of relaxation at any given concentration was comparable between the rat strains (Fig.3). As anticipated, treatment of pulmonary artery segments with l-NNA (100 μM) completely abrogated the relaxation response to thapsigargin (10−5 M) in both strains (from 45.4 ± 6.6 to 7.0 ± 3.4% in M rats and from 43.7 ± 14.6 to −8.42 ± 3.3% in H rats,P < 0.01).
Vasorelaxation response to carbachol in phenylephrine-contracted pulmonary arteries.
To determine whether the greater relaxation response to ACh might be a consequence of diminished metabolism of ACh by cholinesterases within the M rat pulmonary arterial wall, we obtained concentration-response curves to carbachol, a cholinesterase-resistant analog of ACh. A small increase in phenylephrine-induced tone was noted during treatment with lower concentrations (from 10−10 to 3 × 10−8 M) of carbachol (Fig.4 A). This was most likely due to time-dependent changes in phenylephrine-induced contraction because time controls manifested similar increases in tension over time (Fig.4 B). Maximal relaxation responses to carbachol (from 10−7 to 10−4 M) were significantly greater in pulmonary arteries from M rats than from H rats. However, the concentration of carbachol that produced a 50% relaxation in pulmonary arteries did not differ significantly between M and H rats (6.67 ± 0.07 vs. 6.48 ± 0.07, P > 0.05). Notably, pretreatment of pulmonary artery segments with l-NNA (100 μM) completely abolished the relaxation response to carbachol in H rat pulmonary arteries, whereas a small but significant relaxation persisted in M rat pulmonary arteries (Fig.5). Pretreatment with meclofenamate had no effect on the carbachol-induced relaxation response in either strain (data not shown).
Vasorelaxation to carbachol in KCl-contracted pulmonary arteries.
To determine whether EDHF contributes to the differing relaxation responses to carbachol, concentration-response curves to carbachol were obtained after inhibition of the EDHF-mediated component of the endothelium-dependent relaxation by depolarizion of pulmonary arteries with 20 mM KCl. Force generated by 20 mM KCl was comparable to that of phenylephrine (10−6 M) in M (0.33 ± 0.07 vs. 0.36 ± 0.04) and H (0.29 ± 0.02 vs. 0.25 ± 0.02) rat pulmonary arteries. Notably, inhibition of KCl-induced contraction by carbachol in M rat pulmonary arteries did not differ from the response in H rat pulmonary arteries (Fig. 6). Carbachol-induced relaxation was significantly blunted in KCl-contracted pulmonary arteries compared with phenylephrine-contracted pulmonary arteries from M rats (Fig.7 A). In contrast, the response to carbachol in KCl-contracted rings from H rat pulmonary arteries did not differ from that in phenylephrine-contracted rings (Fig.7 B). The reduced relaxation response to carbachol in KCl-contracted pulmonary arteries from M rats was not due to a loss of endothelial function because the relaxation response to carbachol remained intact in these same rings precontracted with phenylephrine (10−6 M) after completion of the KCl protocol (data not shown). To further substantiate the presence of EDHF, we obtained concentration-response curves to carbachol in phenylephrine-contracted,l-NNA-treated M rat pulmonary artery rings in the presence of the Ca2+-dependent K+ channel blockers charybdotoxin and apamin (100 nM each) or the cytochromeP-450 inhibitor 17-ODYA (2 μM). As shown in Fig.8, the residual relaxation resistant to eNOS inhibition was significantly reduced by 17-ODYA and was completely abolished by charybdotoxin and apamin.
We previously observed enhanced ACh-induced relaxation in pulmonary arteries of M compared with H rats and localized this difference to the endothelium. In the present study, we demonstrate that other muscarinic agonists such as carbachol also elicit enhanced relaxation, and we provide evidence to suggest that the enhancement is related to release of EDHF. Contrary to our original hypothesis, however, we found no evidence for increased expression or activity of eNOS in M rat pulmonary arteries.
We found that pulmonary artery eNOS mRNA and protein levels are similar in the two rat strains, demonstrating that the strain differences in the muscarinic agonist-induced pulmonary artery relaxation are not associated with differences in eNOS expression. Nonetheless, eNOS enzyme function could still differ depending on the availability of the substrate l-arginine and cofactors tetrahydrobiopterin, NADPH, flavin adenine dinucleotide, flavin mononucleotide, and Ca2+. For this reason, we used the relaxation response to thapsigargin, a receptor-independent stimulator of eNOS (17, 20, 21), as a functional assay of eNOS enzyme activity. We found that vasodilator responses to thapsigargin were virtually identical in pulmonary arteries of the two strains. Consistent with previous observations in pulmonary arteries (21) and aortas (20), we also demonstrated that l-NNA completely abolished the relaxation response to thapsigargin equally in both strains, indicating that NO release was entirely responsible for the thapsigargin-induced relaxation. Taken together, the similar eNOS mRNA and protein levels and responses to thapsigargin support the idea that the greater carbachol-induced relaxation in M rat pulmonary arteries is not due to greater expression or function of eNOS.
We also examined the possibility that cholinesterases in the pulmonary arteries of the two rat strains might differ. Cholinesterases, including acetylcholinesterase and butyrylcholinesterases, are known to be present in the pulmonary vessels of a variety of species, including rats, rabbits, and humans (11, 31). Several studies have shown that these enzymes regulate and limit the action of ACh by rapidly hydrolyzing it (2, 32). Moreover, strain-related differences in the function of cholinesterases have been reported (6, 11, 26). However, the likelihood that the greater relaxation of M than H rat pulmonary arteries was related to decreased cholinesterase activity seems remote because the relaxation response to carbachol, a cholinesterase-resistant analog of ACh (15), that we observed remained significantly greater in pulmonary arteries from M rats. Likewise, the lack of any effect of the cyclooxygenase inhibitor meclofenamate on the carbachol-induced relaxation response indicates that differences in release of the vasodilator prostaglandin PGI2 were not responsible for the differences between the strains.
Our results suggest that EDHF(s) plays a role in the strain differences. This factor is thought to be responsible for the component of the relaxation response to ACh that is resistant to pharmacological inhibitors of NOS and is attributable to membrane hyperpolarization (3). EDHF has been shown to contribute to the variability in responsiveness to ACh of rabbit arteries isolated from different vascular beds (30). Our observation that l-NNA completely abolished carbachol-induced relaxation of pulmonary arteries in H rats suggests that NO is the main factor responsible for relaxation in this strain. In stark contrast, thel-NNA-insensitive component of vasorelaxation detected in pulmonary arteries from M rats demonstrates that, in this rat strain, other factors such as EDHF contribute to the net vasodilator response to carbachol.
EDHF has been shown to induce relaxation by hyperpolarizing the cell membrane by opening potassium channels (9). Depolarizing artery rings in vitro by increasing extracellular potassium prevents the relaxation response to EDHF (7, 10, 18). In line with these observations, we found that increasing extracellular potassium by adding 20 mM KCl to pulmonary artery rings reduced the relaxation response to carbachol in M rat pulmonary artery rings but had no effect on carbachol-induced relaxation in H rats. Furthermore, the difference in the relaxation response to carbachol between the strains was eliminated. These observations provide more evidence to suggest that EDHF is a contributor to the variability in pulmonary vasoresponsiveness between the M and H strains. Alternatively, it is possible that potassium channels on the vascular smooth muscle cells that are involved in EDHF-induced relaxation may differ between the strains.
The chemical identity of EDHF(s) has not been precisely determined, although studies on the systemic vascular bed suggest that the chemical nature of EDHF may vary depending on the species and tissue involved. For example, a metabolite of cytochrome P-450 acts as an EDHF in rabbit carotid arteries, whereas NO and PGI2 appear to act as hyperpolarizing factors in rabbit middle cerebral arteries (30). The lack of effect of meclofenamate on carbachol-induced relaxation and the persistence of carbachol-induced relaxation after l-NNA treatment suggest that PGI2 and NO are not likely candidates for EDHF in M rat pulmonary arteries. The reduction in the relaxation that persisted in the presence of l-NNA by the cytochrome P-450 inhibitor 17-ODYA suggests that a metabolite of cytochromeP-450 acts as an EDHF in M rat pulmonary arteries. The potassium channels that are thought to mediate EDHF responses also differ depending on the vascular bed and species studied. In the canine pulmonary artery preparation, ATP-dependent K+ channels have been shown to mediate EDHF responses (8), whereas in rat mesenteric arteries, Ca2+-dependent K+channels appear to be responsible (5). Our observation that charybdotoxin and apamin abolished l-NNA-resistant relaxation implies that Ca2+-dependent K+channels mediate EDHF responses in M rat pulmonary arteries.
Differences in EDHF release between rat strains have previously been shown to have possible pathophysiological significance. Compared with the normotensive Wistar-Kyoto rat, stroke-prone hypertensive rats have diminished release of EDHF from isolated mesenteric arteries (22,28). In this context, it is interesting to note that the M rat strain, whose pulmonary arteries show evidence of EDHF release in response to carbachol, develops milder chronic hypoxic pulmonary hypertension than the less carbachol-responsive H strain. However, further study will be necessary to determine whether EDHF plays a role in these strain-related differences in responsiveness to hypoxia.
In conclusion, our findings indicate that the differing pulmonary vasoresponsiveness to muscarinic agonists observed in our two rat strains cannot be explained by differences in the synthesis and/or release of NO. Rather, an additional mediator (or mediators) that has the pharmacological properties of EDHF appears to be contributing to the differences. Furthermore, our findings suggest that EDHF is a metabolite of cytochrome P-450 and causes hyperpolarization by opening Ca2+-dependent K+ channels. Although earlier studies have identified a role for EDHF in differences in systemic vascular reactivity between strains (22, 28), ours is the first to suggest that EDHF contributes to strain-related variability in pulmonary vascular reactivity. Further studies will be necessary to definitively characterize EDHF and to determine whether it contributes to differences in pulmonary vasoresponsiveness of other species, including humans.
We thank Drs. C. L. Jackson and M. De Paepe for kindly making their PCR facilities available and P. Verdier, M. Benson, and A. Brooks for technical support.
This work was supported by grants from The Rhode Island Foundation (M. R. Karamsetty), a developmental grant from Rhode Island Hospital (M. R. Karamsetty), and National Heart, Lung, and Blood Institute Grant HL-40505 (N. S. Hill).
Address for reprint requests and other correspondence: N. S. Hill, Div. of Pulmonary and Critical Care Medicine, Rhode Island Hospital, 593 Eddy St., Providence, RI 02903 (E-mail:).
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 © 2001 the American Physiological Society