INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

HYPOXIC PULMONARY VASOCONSTRICTION (HPV) has been established as an important regulatory mechanism both for reducing arterial hypoxemia in the presence of abnormal ventilation-perfusion ratios and as a cause of pulmonary hypertension in many chronic and acute cardiopulmonary disease states (3, 19). When a local region of the lung becomes hypoxic in vivo, the HPV response is characterized by a rapid and persistent vasoconstriction of the pulmonary arteries (PAs) that remains essentially constant for at least 4 h (5). When isolated rat lungs are ventilated and perfused in vitro and exposed to hypoxia, a similarly rapid constrictor response is observed, with a maximum constriction at 2-5 min and a small 10-20% decline over the next 10 min, and thereafter constriction is sustained for at least 40 min (16, 24). Although there is little disagreement about the form of these responses, those reported for isolated rat PAs are quite disparate. The only consistent observation has been an initial rapid constriction, but thereafter, investigators (1, 10, 12, 31, 32) reported a variable vasodilation that often more than abolished the initial constriction and may or may not be succeeded by a slow constriction developing over many minutes.

Isolated PAs are convenient for the study of the physiological properties, pharmacological influences, and nature of the mechanism of HPV. However, the form of these responses to hypoxia in isolated rat PAs is so variable and differs so much from the whole lung responses that the applicability of the results to HPV in general remains uncertain. In considering the experimental designs of many investigators, differences in three principal variables became apparent. The first was the size of the PA, the second was the choice and dose of the vasoconstrictor used to generate an active tension before eliciting HPV, and the third was the passive or mechanical stretch tension maintained throughout the study. Although there have been reports on the influence of arterial size (12, 13) and vasoconstrictor choice (14), the role of active and passive tensions on the responses to hypoxia has not been investigated systematically.

Transient initial hypoxic constrictor responses have been observed in all rat PAs including the main extrapulmonary and large and small intrapulmonary arteries, and this has led some to the belief that size is not important. However, differences in the form of the response observed in small and large arteries (13) and the fact that HPV in vivo results from constriction of small PAs of <600 µm in diameter (6) undermine this conclusion.

The choice of mechanical stretch tension has generally been derived directly from the techniques used for systemic arteries. Some investigators observed the responses to repeated doses of KCl as the arterial ring (or strip) was increasingly stretched. The smallest passive tension that elicited a maximum response to KCl is then used for the study of HPV. However, the passive tensions used have varied widely (0.5-3 g, even in the same sized arteries), and the assumption that this tension is entirely passive with no active myogenic component and is also optimum for HPV has not been formally tested.

Finally, in lungs in vitro, it has been shown that HPV responses are more readily and consistently observed when the PAs are preconstricted. This practice was therefore introduced to the study of HPV in isolated PAs, but despite studies demonstrating that hypoxic responses are influenced by the choice of vasoconstrictor (14, 25), the dose and choice of specific vasoconstrictor in most publications are probably not regarded as a critical variable. For systemic arteries, Mulvany and Halpern (20) drew attention to the importance of standardizing wall tension, and these principles were applied to PAs by Leach and colleagues (12-14) in a series of elegant studies that form the basis for the present work. Those studies established not only that the form of the responses could be altered by changing stretch tension but also demonstrated that out of the many vasoconstrictor agents that have been used, only prostaglandin F₂ (PGF₂) had comparable dose-response curves in both small and large PAs, presumably because the receptor density, affinity, and endothelium-dependent actions of PGF₂ were less variable with artery size than those of other agents (14).

The present work has systematically examined the influence of stretch (imposed by stretching the artery mechanically) and active (imposed by PGF₂) wall tensions on large (1-mm-diameter) and small (400-µm-diameter) intrapulmonary arteries exposed to 40 min of hypoxia. The results demonstrate that, independent of size, there is a systematic and dramatic change in the form of the hypoxic response that is determined primarily by the stretch tension, whereas the active tension has a more subtle influence.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

PA isolation. The protocols for the preparation and care of the rats used in these studies were approved by the University of Pennsylvania Animal Care and Use Committee. Male Wistar rats were anesthetized with 10 mg of Ketalar (ketamine hydrochloride) intramuscularly and 50 mg/kg of pentobarbital sodium intraperitoneally. The chest was opened in the midline, and heparin sulfate (200 units) was injected into the right ventricle to anticoagulate the blood. The rats were exsanguinated, and the heart and lungs were removed en bloc and immersed in ice-cold Hanks' balanced salt solution containing (in mM) 1.3 CaCl₂, 5.0 KCl, 0.5 MgCl₂ · 6H₂O, 0.3 KH₂PO₄, 0.4 MgSO₄ · 7H₂O, 138 NaCl, 4.0 NaHCO₃, 0.3 Na₂HPO₄, 5.6 D-glucose, and 0.03 phenol red. With the aid of a dissecting microscope the first (large)- and second (small)-generation PAs were immediately isolated. The arteries were cut into ring segments ~1 mm in length and threaded with two wires (25-µm-diameter tungsten steel). At the end of each study, the wet weight of the arterial segment was recorded.

Experimental circuit. The perfusate circuit and small-vessel myograph are similar to those described by others (14, 20). The arteries were mounted on a small-vessel myograph, bathed in circulating Earle's balanced salt solution containing (in mM) 1.8 CaCl₂, 5.3 KCl, 0.8 MgSO₄ (anhydrous), 117 NaCl, 26.0 NaHCO₃, 1.0 NaH₂PO₄ · H₂O, 5.6 D-glucose, and 0.03 phenol red. The perfusate was circulated through one of two water-jacketed reservoirs at 19 ml/min by a Harvard Apparatus model 1203 peristaltic pump. In the reservoirs, the Earle's balanced salt solution was gassed by either the normoxic (21% oxygen-5% carbon dioxide-balance nitrogen) or hypoxic (0% oxygen-5% carbon dioxide-balance nitrogen) gas mixture, and the temperature of the water jacket was regulated by a Haake model FE2 thermostat/pump to maintain the perfusate temperature at 37°C.

The arterial segment was suspended by the two wires in a small-vessel myograph based on that described by Mulvany and Halpern (20). Each wire was attached horizontally to stainless steel "jaws," with one end fixed to a calibrated force transducer (Kulite BG-10GM, Kulite Semiconductor Products, Ridgefield, NJ) and the other to a calibrated micrometer.

Stretch tension determination. For every arterial segment, the circumference-wall tension relationship was initially determined so that stretch tensions could be accurately selected. At the start of each study, the micrometer was adjusted so that the two wires overlapped as viewed from the monocular scope situated perpendicularly above. After a 30-min equilibration period, the length and diameter of the mounted strip were measured with a monocular scope grid when no tension was imposed. The micrometer was then adjusted to stretch the PA circumferentially until the recorder detected force development. The micrometer was thereafter adjusted to stretch the PA in a stepwise manner while the distance between the wires (f; in µm) and the force developed were recorded until the wall tension was ~2 mN/mm. The circumference of the PA (C_in; in µm) was calculated from

(1)

where d is the diameter of the wire (25 µm). The relationship between the circumference and the wall tension was fit with a simple exponential function (13, 19) of the form

(2)

where T (in mN/mm) is the wall tension calculated by dividing the recorded force (in mN) by twice the measured length of the PA segment (in mm), a is the intercept, and b is the slope.

With the assumption that the arterial wall thickness was much less than the diameter and that the form of the wall curvature was not critical, the Laplace equation relates wall tension, circumference, and the effective transmural pressure (P; in mmHg; 1 mN  7.5 mmHg)

(3)

For example, for an effective transmural pressure of 30 mmHg, T = 0.64C_in. The wall tension from Eq. 3 was calculated for each C_in recorded for an artery, and the line obtained was superimposed on the curve generated from Eq. 2 for that artery. The line intersected the curve at the point where the observed wall tension corresponded to an effective intravascular pressure of 30 mmHg. This procedure was repeated for different arteries for effective transmural pressures of 5, 15, 30, 50, and 100 mmHg so that for each specific artery, the individual circumference, and therefore stretch wall tension corresponding to a particular effective transmural pressure, could be selected for subsequent studies.

The following two preliminary studies were performed: the first to establish the stretch wall tension dependence of the response to KCl and the second to determine the dose-response relationship for the active wall tension generated by PGF₂.

Stretch tension with 75 mM KCl challenge. Large and small PAs were prepared as in Stretch tension determination to establish the circumference-wall tension relationship, and then, after a 30-min resting period, the perfusate was replaced with one containing 75 mM KCl (NaCl was reduced by 75 mM to preserve osmolality, and the temperature and normoxic gas tensions were unchanged). The contractile response was observed for 2 min; then the KCl was washed out, and the PAs returned to baseline tension. This maneuver was repeated as the circumference of the PA was increased in a stepwise manner until the active tension developed in response to KCl did not increase further. The active wall tension was calculated by subtracting the stretch wall tension from the total wall tension recorded at each C_in.

PGF₂ dose-response curves. PGF₂ was selected as the source of the imposed active wall tension for this work because, in contrast to many other vasoconstrictors, the dose-response curves for small and large rat PAs were reported to be similar (14). Small and large arteries were prepared as described in Stretch tension determination, after which a stretch tension corresponding to an effective transmural pressure of either 5, 15, or 30 mmHg was imposed. The active wall tension change was recorded with the addition of PGF₂ to achieve 1, 5, 10, 50, and 100 µM concentrations. The dose-response curves were fit with an equation of the form

(4)

where R_obs and R_max are the observed and maximum responses respectively; D is the PGF₂ concentration (in mol/l), pD is the negative logarithm of EC₅₀, and S is the slope at the EC₅₀.

2

2

2

2

2

6

2

6

5

2

2

6

2

6

5

2

2

2

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Mechanical stretch wall tension with 75 mM KCl challenge. In the first preliminary study, for both large and small arteries, the response to stimulation with 75 mM KCl revealed an increase in developed active tension with each stepwise increment of the mechanical stretch wall tension until a maximum response was achieved, and further increments of stretch tension were accompanied by a decreased response. For large and small arteries, the maximum active responses of 1.67 ± 0.13 and 1.74 ± 0.41 mN/mm, respectively, were observed at stretch tensions of 470 ± 50 and 356 ± 83 mg, respectively. To allow for comparisons, these data were normalized by expressing the responses of each artery as a percentage of the maximum active tension developed [active tension (%Max)] and the circumferences as the ratio (C_in/C_max) of the C_in to the circumference at the maximum active tension (C_max). These normalized data are shown in Fig. 1. The maximum stretch force used was 1,579 ± 54 and 1,131 ± 156 mg for the large and small arteries, respectively.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Influence of mechanical stretch wall tension on response to 75 mM KCl. Results of this preliminary study have been normalized to allow comparison of large and small pulmonary arteries (PAs). Active tension is observed active tension developed and expressed as percent maximum active tension developed (%Max). Cin/Cmax, ratio of circumference observed at a particular stretch tension to circumference when KCl response was maximal. Maximal active tensions were developed at Cin/Cmax = 1, when stretch forces were 470 ± 50 and 356 ± 83 mg for large and small PAs, respectively. Active tension development was reduced at greater or lesser stretch forces.

PGF₂ dose-response curves. In the second preliminary study, small and large arteries were prepared with imposed stretch tensions equivalent to transmural pressures of 5, 15, or 30 mmHg and n = 9 arteries for each of these six groups. Each artery was exposed to 1, 5, 10, 50, and 100 µmol/l of PGF₂, and the slope and negative logarithm of the EC₅₀ coefficient (pD) for the sigmoid dose-response relationship were derived. There were no significant differences between the dose-response curve coefficients for the three stretch tensions in either the large or small arteries (data not shown), and, therefore, the results were combined in Table 1. For the subsequent studies, at all stretch tensions (including 50 and 100 mmHg), the following concentrations of PGF₂ were derived from these equations. For large arteries, concentrations of PGF₂ of 0.50, 1.65, and 8.54 µmol/l were used to achieve active tensions corresponding to EC₂₅, EC₅₀, and EC₇₅, respectively, and for the small arteries these concentrations were 0.27, 1.01, and 8.06 µmol/l, respectively.

View this table: [in this window] [in a new window]   Table 1.   Combined coefficients for PGF2 dose-response curves

View larger version (24K): [in this window] [in a new window]   Fig. 2.   Relationship between mechanical stretch wall force and equivalent transmural vascular pressure of large and small PAs. P5, P10, P30, P50, and P100, transmural pressure of 5, 15, 30, 50, and 100 mmHg, respectively. Values are means ± SE. It is noted that forces < 500 mg for large and 300 mg for small PAs correspond to transmural pressures in physiological range, whereas forces > 700 mg for large and 500 mg for small PAs correspond to pulmonary hypertension.

View larger version (45K): [in this window] [in a new window]   Fig. 3.   Responses of large PAs to 40 min of hypoxia. Each point is mean ± SE of change in (delta) wall tension developed to 5 different stretch tensions during course of exposure; n = 12 arteries. A: no added PGF2. B: 25% effective concentration (EC25) of PGF2. C: 50% effective concentration (EC50) of PGF2. D: 75% effective concentration (EC75) of PGF2. Phase 1 constriction is observed at 1 min, but subsequent response varies with imposed stretch and active wall tensions. See text for further discussion.

View larger version (41K): [in this window] [in a new window]   Fig. 4.   Responses of small PAs to 40 min of hypoxia. Each point is mean ± SE of delta wall tension developed to 5 different stretch tensions during course of exposure; n = 12 arteries. A: no added PGF2. B: EC25 of PGF2. C: EC50 of PGF2. D: EC75 of PGF2. Phase 1 constriction is observed at 1 min, but subsequent response varies with imposed mechanical stretch and active wall tensions. See text for further discussion.

2

2

2

2

2

2

2

2

2.

2

View larger version (30K): [in this window] [in a new window]   Fig. 5.   Contour plot summarizing influence of mechanical stretch and active tensions on overall response to hypoxia of all arteries tested. Stretch wall tension is expressed as equivalent transmural pressure of imposed mechanical stretch tension. Active wall tension is ECx of PGF2 used to exert active preconstriction. Contours indicate area (in mN · min · mm1) under curves calculated as in Table 2 but combined for large and small PAs. There is a sharp transition from constrictor (Cons) to dilator (Dilation) response as mechanical stretch tension equivalent is changed from 30 to 50 mmHg. The more subtle influence of preconstrictor PGF2 is also revealed by peak and trough corresponding to constrictor and dilator responses at EC25 to EC50 of PGF2.

View larger version (44K): [in this window] [in a new window]   Fig. 6.   Summary of results of studies of myogenic tone with varying stretch wall tension. A and B: small arteries (n = 6) with stretch tensions of 30 and 50 mmHg, respectively. C and D: large arteries (n = 6) with stretch tensions of 30 and 50 mmHg, respectively. Horizontal dashed line, starting baseline. Values are means ± SE. For both small and large PAs, when stretch tension was 30 mmHg (A and C), removal () of Ca2+ and addition (+) of papaverine (Pap) did not significantly alter wall tension, and, therefore, imposed tension was entirely passive and myogenic tone was absent. When mechanical stretch tension was increased to 50 mmHg (B and D), addition of relaxing solution significantly reduced wall tension, and this reduction was the same under normoxic and hypoxic (Hypox) conditions. Therefore, stretch wall tensions of 50 mmHg or more are associated with generation of myogenic tone, and final wall tension has a passive and a small active component.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The present studies have shown that variations in mechanical stretch and active wall tensions applied to isolated large and small rat PAs determine the overall form of the response to 40 min of hypoxia. Stretch wall tension was identified as the principal variable; when the applied tension was equivalent to a transmural pressure of 30 mmHg or less, the response was predominantly constrictor, and when the equivalent transmural pressure was 50 mmHg or greater, the response was predominantly dilator. Active preconstriction by PGF₂ had a more subtle influence such that constrictor or dilator responses were enhanced when PGF₂ was present in small (EC₂₅) or moderate (EC₅₀) concentrations. For both the large and small PAs, the most active and sustained constrictor responses were obtained when the stretch applied was equivalent to 15 or 30 mmHg combined with preconstriction with EC₂₅ or EC₅₀ of PGF₂. These observations therefore confirm the hypothesis that HPV is a property of all pulmonary arteries irrespective of size, a conclusion consistent with the observations that hypoxic constriction is observed in vascular smooth muscle cells isolated from large or small PAs (15, 21).

Although the general constrictor or dilator nature of the overall hypoxic response, represented by the areas under the curves (Table 2, Fig. 5), is the most obvious result, the details of the responses throughout the 40 min of hypoxia (Figs. 3 and 4) reveal that the initial, transient (first 1 min of hypoxia), phase 1 constriction is present in all preparations and is generally followed by a phase 2 dilation (at 5 min of hypoxia). The principal differences between the responses are therefore due to the phase 3 constrictor response. The dilator phase 2 at lower stretch tensions is seen only as a dip in an otherwise continuous constrictor response. But at the higher stretch tensions, the phase 3 constriction is totally abolished so that the dilator response more than reverses the entire initial constriction. These data therefore confirm the suggestions of others (1, 12) that two constrictor phases separated by a dilator phase can be observed in the hypoxic response of isolated rat PAs. But different investigators have reported inconsistencies in these phases.

The apparent contradictions are of two sorts. The first is that many early investigators (8, 18, 25, 32) tested the hypoxic responses in rat PAs with only 5- to 10-min exposures to hypoxia, probably because the response was poorly sustained. Those studies have therefore investigated only the phase 1 constrictor responses, and although these responses are less sensitive to the stretch forces applied to the arteries and the pharmacological properties reported may play a role in the initial response to HPV, there is little evidence that the results are useful for the interpretation of the prolonged HPV responses of whole lungs. For example, several studies (1, 8, 10, 20, 22, 23, 25, 31) have demonstrated that removal of the endothelium or inhibition of nitric oxide synthase abolishes or attenuates the phase 1 constriction in isolated arteries and has little or no effect on phase 3 constriction. But nitric oxide synthase inhibition enhances the HPV responses of intact lungs in vivo (7, 27) or in vitro (9) without changing the character of the response, and therefore the phase 3 constrictor responses are probably more representative of the physiologically (and clinically) relevant HPV response. This outcome remains controversial because others (11, 30) have concluded that phase 3 is endothelium dependent in isolated PAs, which Ward and Robertson (30) attribute to an endothelium-derived contracting factor. Although endothelial products are clearly important modulators of the final response to hypoxia, there is uncertainty about whether any of the phases are specifically endothelium dependent (17).

The second type of apparent contradiction concerns the phase 3 response that different authors (1, 10, 12) have reported to be absent (Fig. 7A), slow and delayed (Fig. 7B), or partially merged (Fig. 7C) with the phase 1 response. Analysis of these and other reports reveals that the weak, delayed, or absent phase 3 responses were observed when stretch tensions ranged from 0.7 to 3.0 g, whereas the strongest responses, those reported by Leach et al. (14), were observed at the lowest stretch tension employed by these authors, which was equivalent to a transmural pressure of 30 mmHg. The data from the present studies (Fig. 2) demonstrate that forces of <500 mg for large arteries and 300 mg for small arteries correspond to transmural pressures in the physiological range, whereas forces > 700 mg for large arteries and 500 mg for small arteries correspond to pulmonary hypertension. Thus we hypothesize that the characteristics of each of the published reports can now be understood primarily in terms of the stretch tensions applied in preparing the arteries. For most reported studies, the stretch wall tensions were equivalent to severe pulmonary hypertension, often >100 mmHg, and these are conditions known to inhibit HPV in lungs (2, 4).

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Reported hypoxic responses of isolated PAs record a consistent phase 1 constrictor response but a variable phase 3 response relative to baseline (horizontal dotted line). A: phase 1 is succeeded by full phase 2 dilation, and phase 3 response is absent (10). B: phase 1 is succeeded by phase 2 dilation, but a phase 3 constriction develops late (1). C: response is constrictor throughout, with phase 1 and 3 constrictor responses overlapping and phase 2 dilation appearing as a notch between them (12).

A remaining source of variability is associated with the practice of preconstriction before testing with hypoxia; a wide range of doses and different agents have been employed, and it is clear that not all agents are equivalent in arteries of different sizes. Our choice of PGF₂ was based on the studies of Leach et al. (14), who reported that, in Cummin Sprague Europe rats, the dose-response curves for small and large arteries were more similar with PGF₂ than with norepinephrine or serotonin. However, we observed in Wistar rats that although the maximum response to KCl was similar in both arteries, the maximum response to PGF₂ in the small arteries was about one-half of that in the large arteries. It should also be noted that the increasing doses of PGF₂ were not randomized in the present studies, and although the responses to incremental doses were in the ranges expected from the dose-response curves developed in the second preliminary study, nevertheless changes in responsivity to PGF₂ cannot be ruled out. The present observations with PGF₂ were consistent for both large and small arteries, and we suggest not only that constrictor and dilator responses to hypoxia were more active in the presence of small or moderate concentrations of a vasoconstrictor agent but also, under the right conditions of stretch tension, that a sustained constrictor response to hypoxia was observed with no added constrictor agent.

The phase 2 dilation may be interpreted in two ways. If phase 1 and 3 constrictions are the only active responses, then phase 2 will vary as the relative speed of onset and persistence of these constrictor responses alter. Conversely, if phases 1-3 are all active responses, then the strength of the dilation is an additional independent variable. Either view can satisfy the present observations, but recent investigations support active mechanisms for all three phases. Under this hypothesis, phase 1 constriction coincides with increased intracellular Ca²⁺ attributed to partial depolarization, permitting Ca²⁺ entry (28) and Ca²⁺-induced Ca²⁺-release from sarcoplasmic reticulum stores (26, 29). Phase 2 dilation corresponds to a reuptake of sarcoplasmic Ca²⁺ through the activity of sarcoplasmic reticulum pumps (29). Phase 3 is least understood but seems most consistent with increased force sensitization where contraction increases while intracellular Ca²⁺ remains constant (26, 33). Although it remains for future studies to clarify the precise nature of the mechanisms responsible for the phases, this synthesis provides a useful foundation. On this basis, the hypoxic dilation associated with excessive mechanical stretch wall tensions is attributable to a loss of force sensitization.

HPV in intact lungs, both in vivo and in vitro, is characterized by a sustained constrictor response, and it appears desirable to select conditions for the study of HPV in isolated PAs that are consistent with a similar form of response. The mechanical stretch forces imposed on the arterial segments are composed of both passive and myogenic active components and should therefore not exceed the equivalent of an intravascular pressure of 30 mmHg. For large arteries in the present work, that coincided with forces less than or equal to the lowest wall tension consistent with a maximal response to 75 mM KCl (475 mg), but in small arteries, the optimum force for the hypoxic response (250 mg) was significantly less than that observed for the maximum KCl (356 mg). To ensure that consistently constrictor responses to hypoxia are present, the safest course may be to select a stretch force that is substantially lower, perhaps 50-70%, of that at which the maximal response to KCl is observed.

In summary, the present work has systematically examined the influence of active and stretch (passive + myogenic) wall tensions on the responses to hypoxia of isolated large and small rat PAs. The studies demonstrated that although the transient phase 1 constrictor response was relatively insensitive to wall tension, the phase 3 constrictor response was so critically determined by it that a maximal constrictor response was observed at a force equivalent to a transmural pressure of 30 mmHg or less, and the constrictor response was replaced by a dilator response when the equivalent stretch transmural pressure was 50 mmHg or greater. The responses were most active when the arteries were preconstricted by EC₂₅ or EC₅₀ of PGF₂. The present observations not only define the conditions required for isolated PAs to reproduce more closely the HPV response observed in intact lungs but also serve to reconcile many previously apparently contradictory results.