Introduction

Top Abstract Introduction Methods Results Discussion References

Chronic hypoxic lung disease generally is associated with PHT and subsequent remodeling and altered vasoreactivity of the pulmonary vasculature. The precise nature of the changes in vasoreactivity, in particular those involving mechanisms that could potentially be targets for vasodilator therapy (Karamsetty et al., 1995), is still controversial. Two such systems that have a major influence on vasomotor tone are the activities of voltage-dependent ionic currents in the membrane and the intracellular nucleotides cAMP and cGMP.

Acute hypoxia inhibits a delayed rectifier K⁺ current and causes depolarization in the smooth muscle cells from the pulmonary artery (Post et al., 1992), and we have shown that chronic hypoxia results in a maintained reduction in delayed rectifier K⁺ current in cells from small pulmonary arteries of rats, even in normoxic conditions (Smirnov et al., 1994). This theoretically should result in depolarization of the intact artery, but Suzuki and Twarog (1982) have reported that although large pulmonary artery segments from chronic hypoxic rats were depolarized compared with controls, small pulmonary arteries were hyperpolarized (Suzuki and Twarog, 1982). In agonist-constricted large pulmonary arteries both blockers of voltage-dependent Ca⁺⁺ channels or K⁺ channel openers have been more effective in hypoxia- and monocrotaline-induced PHT (Rodman, 1992; Wanstall and O'Donnell, 1992; Wanstall et al., 1994), and a study on small pulmonary arteries after monocrotaline-induced PHT also showed an enhanced response to pinacidil, a K⁺ channel opening agent (Wanstall et al., 1993). These reports would be consistent with an enhanced depolarization during agonist stimulation, and a subsequent increase in Ca⁺⁺ entry via voltage-dependent Ca⁺⁺ channels. However, they also could reflect a decrease in Ca⁺⁺ release from intracellular stores, such that the proportion of the rise in cytosolic Ca⁺⁺ caused by voltage-dependent Ca⁺⁺ entry mechanisms was increased, as conjectured by Wanstall et al. (1994). Whereas enhanced depolarization alone might be expected to increase the rise in intracellular Ca⁺⁺ and thus tension, a decreased release from Ca⁺⁺ stores would do the reverse and reduce tension. It is interesting, therefore, that some reports have shown either no change or a reduction in agonist-induced tone of pulmonary arteries from animals with experimental PHT (Rodman, 1992; Rogers et al., 1992; Wanstall et al., 1995).

cAMP and cGMP influence vasomotor tone primarily via mechanisms involving intracellular Ca⁺⁺ handling (Ushio-Fukai et al., 1993), although cGMP also has been reported to activate K⁺ channels (Hampl et al., 1995). The roles of NO and cGMP in PHT have been studied extensively, and several studies have shown a decreased potency of NO donors such as SNP (Crawley et al., 1992; Rodman, 1992; Wanstall et al., 1992). Conversely, few reports have been concerned with cAMP-dependent mechanisms during PHT. Isoproterenol-induced relaxation of the pulmonary vasculature may be either unchanged (Rodman, 1992; Russ and Walker, 1993) or reduced (Altiere et al., 1986; Shaul et al., 1990). It has been proposed that the latter is caused by a decrease in beta adrenoceptor density, because there was no change in adenylate cyclase activity (Shaul et al., 1990), and Rodman (1992) reported that dibutyryl-cAMP-induced relaxation was unchanged in large pulmonary arteries after chronic hypoxia.

Several aspects remain to be clarified. It is unclear whether the reduced outward current described for isolated cells from small pulmonary arteries of chronically hypoxic rats is of physiological significance in the intact artery, nor whether any depolarization is sufficient to explain the enhanced effects of voltage-dependent vasodilators such as verapamil and K⁺ channel openers, if the latter occurs in small arteries during chronic hypoxia. The role of cAMP, which may affect Ca⁺⁺ handling, also has been largely overlooked. Changes in pulmonary vascular resistance primarily depend on the activity of small muscular pulmonary arteries, and we have shown that these small arteries differ from conduit arteries in their response to a variety of agents (Leach et al., 1992). However, the great majority of previous studies has been performed on large arteries. We therefore have examined whether chronic hypoxia results in a physiologically significant depolarization in small (~300 µm internal diameter) pulmonary arteries of the rat, and whether this is reflected in the actions of the voltage-dependent vasodilator verapamil and the K⁺ channel opener levcromakalim. We have studied further the response to agents that affect intracellular cAMP, including isoproterenol, forskolin and papaverine, and for comparison, we have re-examined the effects of the NO-donor SNP.

	Methods

Top Abstract Introduction Methods Results Discussion References

Animals. Adult male Wistar rats (230-250 g initial weight) were maintained in a normobaric environmental chamber breathing either air (control group) or 10% O₂ (balance N₂) (chronic hypoxic group, CH). O₂ was monitored continuously (Servomex OA 580 analyzer, Servomex Ltd., Sussex, UK), and CO₂ was maintained at less than 0.2% by use of soda-lime scrubbers. A 12-hr light-dark cycle was imposed and the temperature controlled at 22 ± 1°C. After 28 days animals were removed from the chamber, weighed, anesthetized with ether and sacrificed by cervical dislocation. The lungs and a section of small bowel were excised quickly and placed in a PSS containing (in mM): NaCl, 118; NaHCO₃, 24; MgSO₄, 1; NaH₂PO₄, 0.435; glucose, 5.56; Na-pyruvate, 5; CaCl₂, 1.8; and KCl, 4, equilibrated with 5% CO₂ in O₂ (pH 7.35-7.40). One group of 16 control and 16 CH rats were used for estimation of hematocrit, right ventricular and heart weight and arterial blood PO₂. Arterial blood was taken from anesthetized rats breathing room air or 10% O₂ as appropriate. All animals received humane care in compliance with the Home Office (UK) "Guidance on the operation of the Animals (Scientific Procedures) Act" published by Her Majesty's Stationery Office.

Tissue preparation. Small intrapulmonary arteries (control: 331 ± 15 µm, n = 66; CH: 311 ± 18 µm, n = 57) or small mesenteric arteries (control: 276 ± 19 µm, n = 29; CH: 238 ± 11 µm, n = 25) were dissected free of connective tissue and mounted in a small vessel myograph as described previously (Leach et al., 1992) and equilibrated with 5% CO₂ in O₂ (pH 7.35-7.40, 37°C). Both control and CH pulmonary arteries were stretched to give an equivalent transmural pressure of 30 mm Hg, at which both were at the peak of the length-tension curve (Leach et al., 1991, 1992). The presence of a functioning endothelium was determined by application of ACh (10 µM) after agonist-induced contraction. After 60 min equilibration the arteries were subjected to a standard run-up procedure of three 4-min exposures to PSS containing high K⁺ (KPSS, 75 mM [K⁺], equimolar substitution for NaCl) (Leach et al., 1992).

Resting membrane potential measurement. RMP was determined in small pulmonary arteries mounted on a myograph and stretched and equilibrated as above. Conventional KCl-filled borosilicate glass microelectrodes of ~60 megohm resistance were advanced into the vessel wall with a micromanipulator. RMP was measured with an opto-isolated high-impedance headstage coupled to a standard microelectrode amplifier (Neurolog, Digitimer Ltd, Crawley, UK). The results obtained from between 5 and 17 impalements were averaged for each preparation. Each electrode rarely was used for more than one impalement, and the voltage was corrected for tip potential for each electrode and impalement.

Experimental protocols. Chronic hypoxia has been associated with a decrease in voltage-gated K⁺ current density in single smooth muscle cells from pulmonary artery, and as a result a more positive RMP estimated by current-clamp (Smirnov et al., 1994). To ascertain whether the data from single cells were reflected by functional changes in the intact artery, the effects of chronic hypoxia on RMP, K⁺ depolarization and the response to K⁺ channel blockers were examined in small pulmonary arteries. RMP was estimated in arteries from control and CH rats as described above. However, it was impossible to obtain good impalements and meaningful estimations of membrane potential during agonist stimulation in these small arteries. The concentration-tension relationship for K⁺-depolarization was examined in control and CH arteries by replacing the bath solution at 5-min intervals with PSS containing increasing concentrations (20-90 mM) of KCl, iso-osmolar substitution for NaCl. Concentration-response relationships were constructed for TEA (large conductance Ca⁺⁺-sensitive K⁺ channels), 4-AP (delayed rectifier) and apamin (small conductance Ca⁺⁺-sensitive K⁺ channels) in arteries from control and CH rats.

2

2

Chemicals and solutions. All drugs were obtained from Sigma (Poole, UK), with the exception of PGF₂ (Upjohn Pharmaceuticals Ltd., Crawley, UK) and L-NMMA (Novabiocem, Nottinghamshire, UK). Other chemicals were of analytical quality (BDH, Southampton, UK). All drugs were prepared as stock solutions with use of PSS, and for 4-AP the pH was corrected to 7.4 by addition of HCl acid. PSS was made up for each experiment with water freshly drawn from a reverse osmosis-deionization plant with UV irradiation (Elgastat, Elga Ltd, Welwyn Garden City, UK).

Data and statistical analysis. Absolute developed tension is given as mN mm¹ artery length, or as % of tension developed to KPSS (Leach et al., 1992). For vasorelaxation experiments tensions are expressed as percent of the initial tension. The EC₅₀ and maximum response were estimated for individual concentration-response curves by nonlinear least-squares regression (SigmaStat, SPSS Scientific GmbH, Erkrath, Germany) where appropriate. EC₅₀ values were converted to negative logarithmic values for all statistical analysis, although for ease of comprehension EC₅₀ values [±95% confidence limits] are given in the text. All other values are given as mean ± S.E. Data were compared using an unpaired Student's t-test or ANOVA as appropriate (SigmaStat, SPSS Scientific GmbH). Differences were considered significant at P < .05.

	Results

Top Abstract Introduction Methods Results Discussion References

Body weight, RV/LV ratio, hematocrit and PaO₂. Both body weight and PaO₂ were reduced significantly in the CH rats compared with the controls, whereas the RV/LV ratio, the RV body weight ratio and hematocrit were increased significantly, consistent with the development of chronic hypoxemia and concomitant pulmonary hypertension (table 1).

Pulmonary artery function. The tension developed in small pulmonary arteries from control and CH rats in response to KPSS or PGF₂ (50 µM) is shown in figure 1. Arteries from CH rats did not develop a significantly greater tension than control arteries for either agent. Preincubation with L-NMMA (100 µM) did not significantly increase PGF₂-induced tension in arteries from either the control or CH group. L-NMMA, however, did cause a small but significant increase in basal (unstimulated) tone in arteries from both control (0.03 ± 0.01 mN·mm¹, n = 11, P < .05) and CH rats (0.07 ± 0.03 mN·mm¹, n = 10, P < .05), which suggests that basal release of NO was unimpaired.

View larger version (15K): [in this window] [in a new window]   Fig. 1.   KPSS and PGF2 (50 µM) induced tension development in small pulmonary arteries from control () and CH () rats. There was no significant difference between control and CH groups (KPSS: control n = 41, CH n = 36; PGF2: control n = 30, CH n = 33). L-NMMA (100 µM) did not cause a significant increase in PGF2-induced tension in either group (Control n = 10; CH n = 8).

RMP, K⁺ depolarization and K⁺ channel blockade. The RMP in small pulmonary endothelium-intact arteries from CH rats was stable, and significantly more positive than in those from control rats (Control: 60.2 ± 0.5 mV, n = 5; CH: 54.4 ± 1.1 mV, n = 5; P < .01). This was reflected by a significant shift to the left in the K⁺ depolarization concentration-response curve (fig. 2), such that the EC₅₀ for K⁺ was reduced from 41.9 [1.11,+1.14] mM (n = 7) in controls to 31.4 [2.61,+2.85] mM in CH rats (n = 8, P < .01). The maximum tension extrapolated from the least-squares fit was increased in arteries from CH rats compared with those from control rats, but this did not reach significance (Control: 2.36 ± 0.13 mN·mm¹, n = 7; CH: 2.58 ± 0.19 mN·mm¹, n = 8; NS). TEA (to 30 mM) and apamin (to 10 µM) had no effect on pulmonary artery tension, suggesting that neither large nor small conductance Ca⁺⁺-sensitive K⁺ channels play a significant role in the maintenance of RMP in this tissue. 4-AP however caused a concentration-dependent increase in tension greater than 1 mM in arteries from controls (fig. 3), reaching a tension of 10.5 ± 1.2% KPSS at 10 mM (n = 6). This would be consistent with a delayed rectifier K⁺ current having a major role in the modulation of RMP. In pulmonary arteries from CH rats the 4-AP concentration-response curve was shifted substantially to the left, such that significant tension was developed at 100 µM 4-AP, and at 3 mM 4-AP tension had increased to 41.4 ± 5.8% KPSS (n = 7) (fig. 3). At 10 mM the preparations from CH rats became oscillatory, and measurements could not be made.

View larger version (12K): [in this window] [in a new window]   Fig. 2.   Tension development in small pulmonary arteries from control (, n = 7) and CH (, n = 8) rats to increasing depolarization with K+ (iso-osmolar substitution for Na+). Symbols are mean ± S.E. * denotes P < .05 at that concentration between control and CH arteries.

View larger version (12K): [in this window] [in a new window]   Fig. 3.   Effect of increasing concentrations of 4-AP on tension in small pulmonary arteries from control (, n = 6) and CH (, n = 7) rats, and in arteries from control rats in the presence of 30 mM K+ (, n = 7). Tension is expressed in terms of that to KPSS, and symbols are mean ± S.E. Where symbols have no error bar the S.E. is smaller than the symbol. * denotes P < .05 at that concentration compared with control for each group.

Effect of chronic hypoxia on the actions of vasodilators. Verapamil caused a concentration-dependent relaxation of arteries from control rats with a significantly smaller maximum relaxation in small pulmonary arteries (46.3 ± 6.1% initial tone, n = 6) than in small mesenteric arteries (93.9 ± 2.1%, n = 6, P < .001), and a significantly greater EC₅₀ (Pulmonary: 80 [22,+30] nM; mesenteric: EC₅₀: 26 [7,+10] nM, P < .001) (fig. 4). In pulmonary arteries from CH rats the maximum relaxation was significantly greater than in controls (66.8 ± 4.9%, n = 5, P < .05), and the curve was shifted to the left (EC₅₀: 30 [11,+18] nM, P < .01). There was no significant difference in either maximum relaxation or EC₅₀ in mesenteric arteries from CH rats (87.9 ± 1.9%, EC₅₀: 19 [6,+8] nM, n = 5) (fig. 4).

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Verapamil concentration-response curves for small pulmonary arteries from control (, n = 6) and CH (, n = 5) rats, and small mesenteric arteries from control (, n = 6) and CH (, n = 5) rats. Symbols are mean ± S.E. Where symbols have no error bar the error is smaller than the symbol. * denotes P < .05 at that concentration between control and CH pulmonary arteries.

View larger version (19K): [in this window] [in a new window]   Fig. 5.   Levcromakalim concentration-response curves for small pulmonary arteries from control (, n = 7) and CH (, n = 6) rats, and small mesenteric arteries from control (, n = 8) and CH (, n = 6) rats. Symbols are mean ± S.E. Where symbols have no error bar the error is smaller than the symbol. * denotes P < .05 at that concentration between control and CH pulmonary arteries.

2

View larger version (19K): [in this window] [in a new window]   Fig. 6.   SNP concentration-response curves for small pulmonary arteries from control (, n = 10) and CH (, n = 8) rats, and small mesenteric arteries from control (, n = 9) and CH (, n = 8) rats. Symbols are mean ± S.E. Although there was no significant difference between individual means, the EC50 was increased significantly in CH rats (see text).

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Forskolin concentration-response curves for small pulmonary arteries from control (, n = 5) and CH (, n = 6) rats, and in the presence of 100 µM L-NMMA (control: , n = 5; CH: , n = 4). Symbols are mean ± S.E. Where symbols have no error bar the error is smaller than the symbol. * denotes P < .05 at that concentration between control and CH arteries in the absence of L-NMMA;  denotes P < .05 between control and CH arteries in the presence of L-NMMA.

View larger version (17K): [in this window] [in a new window]   Fig. 8.   Isoproterenol concentration-response curves for small pulmonary arteries from control (, n = 6) and CH (, n = 6) rats, and in the presence of 100 µM L-NMMA (control: , n = 5; CH: , n = 4). Symbols are mean ± S.E. Where symbols have no error bar the error is smaller than the symbol. * denotes P < .05 at that concentration between control and CH arteries in the absence of L-NMMA;  denotes P < .05 between control and CH arteries in the presence of L-NMMA.

2

View larger version (15K): [in this window] [in a new window]   Fig. 9.   Papaverine concentration-response curves for small pulmonary arteries from control (, n = 6) and CH (, n = 7) rats. * denotes P < .05 at that concentration between control and CH arteries. Symbols are mean ± S.E.

	Discussion

Top Abstract Introduction Methods Results Discussion References

PHT induced by chronic hypoxia is associated with vascular remodeling and muscularization of small pulmonary arteries. Although there was significant polycythemia and right ventricular hypertrophy in the CH group (table 1), which suggests the presence of PHT, there was no significant increase in maximum force development to either PGF₂ or KPSS (fig. 1). A similar lack of increase in agonist-induced tone was reported in small pulmonary arteries from CH rats (Rogers et al., 1992), whereas in large pulmonary arteries at least two reports showed a reduction in force development, even though muscle mass was increased (Rodman, 1992; Wanstall et al., 1995). It is unclear whether this reflects a change in the ability of the smooth muscle to generate force, although it is unlikely to be related to an increase in basal tone as suggested by Wanstall et al. (1995), because maximum relaxation to papaverine did not differ between the two groups (fig. 9).

The RMP in small pulmonary arteries from CH rats was significantly more positive than in controls, consistent with our previous finding in isolated cells (Smirnov et al., 1994). This is in direct contrast to the results of Suzuki and Twarog (1982), who showed an hyperpolarization in small pulmonary arteries after chronic hypoxia. A depolarization from 60 to 54 mV as reported here for arteries from CH rats would be unlikely to have a very significant effect on voltage-activated Ca⁺⁺ channels, and it is notable that there was no increase in basal tone (see above). It was therefore important to establish whether this relatively small depolarization was of any functional significance.

We examined the response to depolarization with K⁺ and found that the curve was shifted significantly to the left, as would be expected if there was a pre-existing and functionally significant depolarization (fig. 2), although maximum force was not altered. This is consistent with a previous report on small arteries from CH rats (Rogers et al., 1992). Because neither TEA nor apamin had any effect on resting tone, it can be assumed that neither large nor small conductance Ca⁺⁺-activated K⁺ channels play a role in maintenance of the RMP, which is consistent with previous patch-clamp studies (Smirnov et al., 1994; Turner and Kozlowski, 1997). However, 4-AP, a relatively selective blocker of the delayed rectifier, did cause vasoconstriction, which suggests that this current is the major determinant of the RMP in this tissue, as proposed previously (Smirnov et al., 1994; Turner and Kozlowski, 1997). In arteries from CH rats the 4-AP response curve was shifted to the left (fig. 3). This is also consistent with a pre-existing depolarization, as the delayed rectifier is voltage activated.

Many agonists cause depolarization in addition to release of intracellular Ca⁺⁺ stores and/or Ca⁺⁺ entry via receptor-operated channels. If the cell was already slightly depolarized the total depolarization in the presence of such an agonist might be expected to be enhanced, leading to increased activity of voltage-activated Ca⁺⁺ channels. Agonist-induced tension therefore might be more sensitive to blockers of these channels, or agents that cause hyperpolarization (Rodman, 1992). As previously shown for nifedipine in rat large pulmonary arteries (Rodman, 1992), verapamil caused a greater maximum relaxation, and the EC₅₀ was smaller in small pulmonary arteries from CH compared with control rats (fig. 4). A similar response was observed for the K⁺ channel opener levcromakalim, again consistent with previous reports in large arteries for levcromakalim (Rodman, 1992) and pinacidil, although for the latter no increase in potency was observed in PGF₂-constricted arteries (Wanstall and O'Donnell, 1992; Wanstall et al., 1994). The possibility that PGF₂ itself causes a greater depolarization in arteries from CH rats cannot be ruled out, however.

The maximum relaxation to verapamil and levcromakalim in small arteries from both groups was substantially less than that previously reported for similar agents in large arteries (Rodman, 1992; Wanstall and O'Donnell, 1992; Wanstall et al., 1994), and even in CH rats it only approached 60 to 70% of initial tone (figs. 4 and 5). In comparison there was almost complete relaxation to both verapamil and levcromakalim in mesenteric arteries, which were not affected by chronic hypoxia. This implies that unlike mesenteric and large pulmonary arteries, PGF₂-induced constriction in small pulmonary arteries depends only partially on Ca⁺⁺ entry via voltage-activated Ca⁺⁺ channels, although the fact that Ni abolished the remaining tone suggests that Ca⁺⁺ entry via other routes is required. Because total PGF₂-induced tension was not increased in arteries from CH rats compared with controls, our results imply that other mechanisms may be decreased proportionately after chronic hypoxia, which suggests an alteration in Ca⁺⁺ handling. A similar proposal was made by Wanstall et al. (1994) for constriction of large pulmonary arteries from CH rats by ET-1 and noradrenaline, although not PGF₂.

As previously reported for large pulmonary arteries from CH rats (Crawley et al., 1992; Rodman, 1992; Wanstall et al., 1992), and small arteries from monocrotaline-treated rats (Wanstall et al., 1993), the potency of SNP was reduced in small arteries from CH rats compared with controls, and there was no change in maximum relaxation. However, SNP only induced a maximum relaxation of ~64% of initial tone in our small pulmonary arteries compared with nearly complete inhibition of tone at maximally effective concentrations in large pulmonary arteries (Rodman, 1992; Wanstall et al., 1992), small pulmonary arteries from monocrotaline-treated rats (Wanstall et al., 1993) and the mesenteric arteries in this study (fig. 6). The relatively small (~4-fold) decrease in potency coupled with the lack of change in maximum relaxation in these small arteries, which contribute the major component of the increased pulmonary vascular resistance in PHT, may explain why the response to SNP in perfused lungs has been reported to be unchanged in CH rats (Adnot et al., 1991; Russ and Walker, 1993).

The greater maximum relaxation, and similar or greater potency for verapamil, levcromakalim and SNP in mesenteric arteries compared with small pulmonary arteries suggests that all these agents are essentially selective for the systemic circulation, even after chronic hypoxia, and that they would therefore be poor drugs for treating PHT. The greater maximum relaxation to SNP and pinacidil, and the significant increase in basal tone in small pulmonary arteries reported for monocrotaline-treated rats also suggests that the monocrotaline model of PHT may cause a more profound lesion in these small arteries than chronic hypoxia.

The experiments with forskolin, isoproterenol and papaverine were designed to investigate cAMP-mediated relaxation after chronic hypoxia. Isoproterenol is described classically as a cAMP-dependent vasorelaxant, acting via beta adrenoceptors to stimulate adenylate cyclase, and the response to isoproterenol has been reported to be depressed (Altiere et al., 1986; Shaul et al., 1990; Shaul et al., 1991) or unaffected (Rodman, 1992; Russ and Walker, 1993) in experimental PHT. Our results suggest that isoproterenol-induced relaxation is impaired substantially in small pulmonary arteries from CH rats (fig. 8), although there was also a significant proportion of the response that was sensitive to L-NMMA, and presumably therefore involved NO. The NO-dependent part of the response to isoproterenol was reduced only partly in arteries from CH rats, whereas in the presence of L-NMMA relaxation to isoproterenol in CH arteries, it was effectively abolished.

The results for isoproterenol contrast strongly to those for forskolin, a direct activator of adenylate cyclase, where the response was enhanced greatly in small arteries from CH rats (fig. 7). As we described previously (Priest et al., 1997), there was a component of forskolin-induced relaxation that was NO-dependent, but this was relatively unchanged in arteries from CH rats. The increased response to forskolin would be consistent with either an up-regulation of the adenylate cyclase, a down-regulation of phosphodiesterase activity or an increased sensitivity to intracellular cAMP. The increase in response to forskolin coupled with the decrease in the response to isoproterenol also suggests a defect in receptor-coupling, and indeed beta adrenoceptor density was reported to be decreased in large pulmonary arteries after chronic hypoxia (Shaul et al., 1990). However, the same report showed no change in the response to forskolin or in basal activity of adenylate cyclase in pulmonary arteries (Shaul et al., 1990). Rodman (1992) also reported that direct relaxation with dibutyryl-cAMP was unchanged in large pulmonary arteries from CH rats. Because this could represent a real difference between large and small pulmonary arteries, we re-examined the effects of forskolin in large arteries. In contrast to its effect on small arteries, forskolin could cause complete inhibition of induced tone in large arteries from controls. In large arteries from CH rats the maximum relaxation was unchanged, although there was a small (~2.5-fold) reduction in potency. Therefore, after chronic hypoxia small pulmonary arteries apparently behave differently to large pulmonary arteries in terms of adenylate cyclase and/or cAMP, which may have implications for the control of pulmonary vascular resistance in PHT.

Papaverine is a phosphodiesterase inhibitor that primarily causes relaxation by reducing the breakdown of cAMP (Holzmann et al., 1977). It was the only vasorelaxant used in this study that caused complete relaxation of PGF₂-induced tension in small pulmonary arteries, and in fact elicited more than 100% relaxation in both control and CH preparations, which indicates that small pulmonary arteries have some degree of intrinsic basal tone, although this was not increased in arteries from CH rats. There was a small increase in potency between control and CH arteries. Papaverine also may cause nonselective inhibition of voltage-activated Ca⁺⁺ channels (Iguchi et al., 1992), and the increased potency in CH preparations therefore could reflect the more positive RMP. In the presence of verapamil, papaverine no longer showed any difference in potency between arteries from control and CH rats, which is entirely consistent with the latter hypothesis. The lack of effect of chronic hypoxia on the voltage-independent component of relaxation to papaverine might argue against any alterations in either phosphodiesterase activity or sensitivity to cAMP after chronic hypoxia, and in combination with the forskolin data provides further evidence for up-regulation of the adenylate cyclase.

It has been suggested that acute HPV is caused by inhibition of a K⁺ current, which results in depolarization and increased Ca⁺⁺ entry (Post et al., 1992). A pre-existing depolarization therefore might be expected to enhance HPV, and in tissue from normal rats this is indeed the case, both in terms of the contractile response and the hypoxia-induced depolarization (Leach et al., 1994; Turner and Kozlowski, 1997). However, although we demonstrate here that pulmonary arteries from CH rats are depolarized, it has long been established that chronic hypoxia decreases the pressor response to acute hypoxia (McMurtry et al., 1978). We previously proposed that HPV depends on mechanisms in addition to depolarization alone, including Ca⁺⁺ release from stores (Ward and Robertson, 1995). The decrease in HPV after chronic hypoxia may therefore be related partly to the putative changes in Ca⁺⁺ handling that are discussed above.

In conclusion, we have demonstrated a small but physiologically significant depolarization in small pulmonary arteries from CH rats, in direct contrast to the report of Suzuki and Twarog (1982); and although this depolarization (or the underlying decrease in outward current) potentiated the effects of further depolarization by K⁺ or channel blockade with 4-AP, it was insufficient to cause any noticeable increase in basal tone. The increased efficacy of the voltage-dependent vasodilators verapamil and levcromakalim was similar to that previously reported for large arteries and would be consistent with an enhanced depolarization during agonist stimulation, as suggested by Rodman (1992). However, in comparison with other studies our results imply that the increase in cytosolic Ca⁺⁺ in small pulmonary arteries depends less on voltage-dependent Ca⁺⁺ entry than in large pulmonary arteries or systemic arteries, and there is circumstantial evidence that the voltage-independent component, that is release from stores or entry through voltage-independent channels, is reduced after chronic hypoxia. It is currently unclear whether this could be related to the apparent up-regulation of the adenylate cyclase in small arteries, and further studies on the function of intracellular Ca⁺⁺ stores, direct measurement of adenylate cyclase activity and intracellular cAMP during chronic hypoxia are required.