INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

RECENT MEASUREMENTS of regional pulmonary perfusion in the resting horse indicate that gravity is a minor determinant of regional pulmonary perfusion (r) (1, 7), despite the large hydrostatic gradient imposed by the height of the lung (lung height ~60 cm at total lung capacity). Furthermore, exercise causes a further, although slight, redistribution of r to dorsal lung regions (1). The observed distribution of r and the change from rest to exercise have been hypothesized to be related to the effect of the structure of the pulmonary vascular tree and to other undefined factors (1).

Endothelial cells (EC) modulate vascular tone and blood pressure in vivo (19), modify responses to many vasoactive agents (18), and mediate flow-dependent vasodilation (10, 18) through the release of nitric oxide (NO) or vasodilator prostanoids (11-13). One possible mechanism for the gravity-independent distribution of pulmonary blood flow and the preferential perfusion of caudodorsal lung regions is regional difference in regulation of pulmonary vascular smooth muscle tone by EC. We previously reported the results of a preliminary study of the effects of the endothelium-dependent vasodilator methacholine (MCh) on pulmonary arteries from caudodorsal (top) and ventral (bottom) regions of the horse lung (17). MCh caused very different responses in top and bottom arteries. In top pulmonary arteries with intact endothelium, MCh (10⁶ M) consistently caused a profound and sustained relaxation, whereas in bottom arteries MCh only caused a small and transient relaxation followed by a sustained and profound contraction. Removal of endothelium converted the MCh-induced relaxation of top vessels to a contraction and, in bottom vessels, abolished the relaxation and significantly enhanced the contraction.

The present study expands on this original study and investigates whether these regional differences in endothelium-mediated responses occur over a range of MCh concentrations and in response to other endothelium-dependent agonists. We used endothelium-dependent agonists because they cause vasodilation by mechanisms similar to those responsible for flow-dependent vasodilation.

We tested the following hypotheses: 1) in the horse, arteries from the top (i.e., caudodorsal regions) of the lung relax more than those from the bottom (i.e., cranioventral regions) in response to the endothelium-dependent vasodilators MCh, bradykinin (BK), and calcium ionophore A-23187 (A-23187); 2) these regional differences in vasodilation are mediated by the endothelium; and 3) NO is involved in the regional differences in relaxation.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Animals. Twenty-four healthy horses [5.3 ± 4.4 (SD) yr, 827 ± 83 pounds] were used in this study, which was approved by the All-University Committee on Animal Use and Care of Michigan State University. Horses were purchased specifically to obtain tissue samples for this and other studies. The animals were euthanized by an overdose of pentobarbital sodium (100 mg/kg body wt). The heart and lungs were removed quickly and examined to ensure the absence of gross lesions. Pulmonary arteries from the top and bottom of the lungs were dissected (Fig. 1). "Top vessels" were taken from the caudodorsal aspect of the caudal lobe, i.e., the region where gravity-independent preferential perfusion occurs in several animal species (1, 2, 5, 7, 8). "Bottom vessels" were obtained from the ventral aspect of the cranial part of the caudal lobe. Arteries were dissected out of the lungs and placed in cold physiological salt solution (PSS; in mM concentration: 127.0 NaCl, 4.7 KCl, 1.1 MgSO₄, 1.2 KH₂PO₄, 25.0 NaHCO₃, 2.5 CaCl₂, 0.02 EDTA, 5.5 glucose, and 2.0 Na pyruvate) that had been aerated with 95% O₂-5% CO₂. They were then transported a short distance to the laboratory.

View larger version (27K): [in this window] [in a new window]   Fig. 1.   Schematic representation of sampling sites of pulmonary arteries. Top pulmonary arteries were taken from caudodorsal diaphragmatic lobe, and bottom pulmonary arteries were taken from cranioventral aspect of caudal lobe.

Preparation of tissues. In cold PSS oxygenated with 95% O₂-5% CO₂, the arteries were cleaned of surrounding pulmonary parenchyma and of excessive connective tissue. Several arterial rings of similar size (6 mm OD, 4 mm wide) were prepared from each artery. In some rings, EC were removed by rubbing the arterial inner surface with a moistened cotton swab. Endothelium removal was verified by examining serial sections of rings stained with either hematoxylin and eosin or an immunostain to the factor VIII-related antigen.

Rings were suspended horizontally between a stationary hook and a movable stainless steel stirrup in 15-ml organ baths filled with warm (38°C) oxygenated PSS. The movable stirrup was attached by 3-0 silk suture to a force transducer (model FT-03 Force Transducer, Grass Instruments, Quincy, MA), mounted on a micromanipulator to allow adjustment of tissues to optimal length. The isometric tension generated was continuously recorded on a multichannel polygraph (model 7E Physiograph, Grass Instruments). Rings were equilibrated for 60 min at a passive tension of 4-5 g. To determine optimal passive tension, a length-tension curve was constructed for each ring by stretching the rings in a stepwise manner. At each length, the rings were contracted with a submaximal concentration of norepinephrine (NE) (10⁶ M; concentration that elicited 60% of the maximal NE contraction). Optimal passive tension was the tension at which further stretching of the ring no longer increased the magnitude of the contraction elicited by NE. The optimal passive tension was maintained during the remainder of the experiment.

Response to endothelium-dependent agonists. Regional differences in endothelium-dependent vasodilation were studied by examining the responses of rings from top and bottom pulmonary arteries, with and without EC, to the endothelium-dependent vasodilators. In pilot studies, NE-induced contraction of top pulmonary arteries faded over time, making cumulative concentration-response curves impossible to interpret. For this reason, we performed noncumulative concentration-response curves to MCh (10⁸ M to 10⁴ M), BK (10¹¹ M to 3 × 10⁹ M), and A-23187 (10⁸ M to 3 × 10⁶ M). To ensure that any agonist-induced relaxation was not due to the time-dependent waning ("fade") of NE-induced contraction, paired time controls were done on top and bottom pulmonary arteries, with and without endothelium. The experimental ring was exposed to an endothelium-dependent agonist, whereas the other received PSS (vehicle) to serve as time control. Rings were contracted with NE (10⁶ M). When active tension reached a plateau, a concentration of agonist was added, and changes in active tension were recorded for 10 min. The vessels were then washed with warmed PSS until the active tension returned to baseline, and they were contracted with NE (10⁶ M) before the next higher concentration of agonist was added. This process was repeated until all concentrations of a given agonist had been administered. Each ring received only one agonist.

Endothelium-independent response. To determine whether the regional differences in the endothelium-mediated relaxation result from differences in the smooth muscle responsiveness to NO, we examined the response to the endothelium-independent agonist sodium nitroprusside (SNP) in endothelium-intact and -denuded top and bottom pulmonary arteries. Noncumulative response curves to SNP were carried out as for MCh, BK, and A-23187.

Mechanism(s) of the regional difference in endothelium-mediated relaxation. To assess the mechanisms of the regional difference in endothelium-mediated relaxation, we compared the response of endothelium-intact top and bottom arterial rings to MCh before and after incubation with either 1) PSS (time control), 2) the NO-synthase inhibitors N-nitro-L-arginine (L-NNA; 3 × 10⁵ M) or N^G-monomethyl-L-arginine (L-NMMA; 10⁴ M), 3) the guanylate cyclase inhibitor methylene blue (MB; 10⁵ M), 4) the cyclooxygenase inhibitor meclofenamate (Meclo; 10⁵ M), or 5) the combination of L-NNA and Meclo. Furthermore, to ascertain that the inhibition of MCh-induced relaxation caused by L-NNA was due to the specific inhibition of NO synthase, we examined the effect of incubating rings with either PSS, the NO-synthase substrate L-arginine (L-Arg), or its inactive isomer D-arginine (D-Arg).

Response to KCl. To determine whether regional differences in relaxation resulted from differences in the contractile properties of the smooth muscle, potassium-substituted PSS (KCl, 120 mM) was added to each bath at the end of each protocol. Increases in active tension were recorded until a peak in active tension was reached.

At the end of the experiment, each arterial ring was blotted, weighed, and fixed in Formalin. Rings were sectioned, stained with hematoxylin and eosin or an immunostain for the factor VIII-related antigen, and multiple sections of each ring were examined by light microscopy to assess the removal of endothelium.

Drugs. The following drugs were used: norepinephrine hydrochloride, methacholine hydrochloride, bradykinin acetate, calcium ionophore A-23187, L-NNA, L-arginine hydrochloride, D-arginine hydrochloride, sodium meclofenamate, methylene blue trihydrate (Sigma Chemical, St. Louis, MO) and N^G-monomethyl-L-arginine citrate (Cayman Chemicals, Ann Arbor, MI). All drug solutions were made fresh daily by dissolving drugs in either 0.1% ascorbic acid solution (NE and MCh) or deionized water. Stock solutions were diluted with either the ascorbic acid solution (NE and MCh) or PSS. Drug solutions were added in a volume of 0.1% of the liquid volume of the bath.

Data analysis. Data are expressed as means ± SE, and n denotes the number of animals used. Contraction responses to NE and potassium-substituted PSS were expressed as grams active tension per gram wet weight of tissue. All data included comparisons of treated and time-control rings. Changes in active tension were expressed as percent change from the NE-induced precontraction (change in active tension × 100/tension developed by NE 10⁶ M). To construct concentration-response curves, we calculated changes in active tension at the time at which the response reached a plateau: 2 min (BK and A-23187) and 5 min (MCh and SNP) after agonist administration. The effects of blood vessel location (top vs. bottom), endothelium removal, and inhibitors were assessed by two-way ANOVA for repeated measures. If one of these factors was significant (P < 0.05), Tukey's post hoc tests were performed.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Optimal passive tension did not differ between top (14.0 ± 0.7 g; n = 13) and bottom vessels (13.1 ± 0.8 g; n = 12). The contractions of top and bottom arteries to potassium-substituted PSS were not different (218.8 ± 13.3 g/g; n = 26, vs. 199.5 ± 11.5 g/g; n = 28). NE elicited contractions of similar magnitude in top (127.0 ± 9.4 g/g; n = 33) and bottom (121.0 ± 6.2 g/g; n = 28) rings. In time-control rings, tension decreased by 2.1 ± 1.1% (n = 61) by 10 min after peak contraction (data not shown).

Response to endothelium-dependent agonists. Figure 2 shows the time course of the responses of top and bottom pulmonary arteries to MCh (10⁶ M). In top pulmonary arteries, MCh induced a relaxation that reached a maximum by 2.5 min and was maintained for the entire 10-min observation period. By contrast, in bottom pulmonary arteries, MCh caused transient relaxation that peaked at 1 min and was followed by contraction that was fully developed by 10 min. The difference between top and bottom vessels in endothelium-mediated relaxation to MCh was observed over a wide range of concentrations (Fig. 3). Because the timing of the MCh-induced responses differed between top and bottom pulmonary arteries, the MCh concentration-response curve was constructed by plotting data obtained at 5 min, a time by which the relaxation of top vessels is stable and the contraction of bottom vessels is near maximal (Fig. 2). In endothelium-intact top arteries, low concentrations of MCh (10⁸ M to 10⁶ M) caused a concentration-dependent relaxation. Higher concentrations caused less relaxation or even contraction. In top vessels, removal of endothelium abolished relaxation and caused a concentration-dependent contraction. In endothelium-intact bottom vessels, MCh caused a concentration-dependent contraction that was enhanced by endothelium removal. Thus, in response to MCh, top vessels relax in an endothelium-dependent manner, whereas bottom vessels contract.

View larger version (12K): [in this window] [in a new window]   Fig. 2.   Examples of responses to methacholine (MCh; 106 M) in endothelium-intact arterial rings from top (caudodorsal region; A) and bottom (cranioventral region; B) of the lung. In response to MCh, top artery relaxed, whereas bottom artery had only a small transient relaxation followed by a profound long-lasting contraction. NE, norepinephrine.

View larger version (29K): [in this window] [in a new window]   Fig. 3.   Noncumulative concentration (shown by brackets) response curves to MCh (A; n = 7 horses); calcium ionophore A-23187 (B; n = 5 horses); bradykinin (BK; C; n = 5 horses); and sodium nitroprusside (SNP; D; n = 6 horses) in endothelium-intact and -denuded top and bottom pulmonary arteries. Measurements were made 2 min (BK and A-23187) or 5 min (MCh and SNP) after the agonist was added. MCh caused a concentration-dependent relaxation of endothelium-intact top vessels and a concentration-dependent contraction of endothelium-intact bottom arteries. BK and A-23187 caused a concentration-dependent relaxation of endothelium-intact top arteries that was greater than that of bottom vessels. Removal of endothelial cells abolished the relaxation of top and bottom pulmonary arteries to BK and A-23187; MCh-induced relaxation of top vessels was abolished, and contraction of bottom vessels was enhanced by endothelial cell removal. SNP caused identical relaxations in all rings. Data are means ± SE. * Significant difference (P < 0.05) between endothelium-intact and -denuded bottom vessels;  significant difference (P < 0.05) between endothelium-intact and -denuded top vessels; § significant difference (P < 0.05) between top and bottom vessels.

The difference in endothelium-mediated relaxation between top and bottom arteries was also observed in response to the receptor-dependent and -independent agonists BK and A-23187. These agents caused a concentration-dependent relaxation of endothelium-intact top and bottom pulmonary arteries (Fig. 3, B and C). In comparison to top arteries, the concentration-response curves of bottom vessels to BK and A-23187 were significantly shifted to the right. Endothelium removal abolished the relaxation of both top and bottom pulmonary arteries to BK and A-23187.

Endothelium-independent response. The endothelium-independent vasodilator SNP caused a concentration-dependent relaxation of top and bottom pulmonary arteries with and without endothelium (Fig. 3D). There were no significant differences in the SNP-induced relaxation between top and bottom vessels (91.8 ± 3.5% vs. 89.3 ± 3.5% at 10⁶ M). Removal of EC did not change the relaxation response of top and bottom vessels to SNP.

Mechanism(s) of the regional difference in endothelium-mediated relaxation. Relaxation was the common feature of the response to endothelium-dependent agonists, and contraction of bottom arteries was observed only for MCh, but not BK or A-23187 (Fig. 3). Therefore, we decided to focus our studies on the mechanism(s) of the regional difference in endothelium-mediated relaxation to MCh. The NO-synthase inhibitors L-NNA (3 × 10⁵ M), and L-NMMA (10⁴ M) inhibited the relaxation of top vessels and completely abolished the transient relaxation of bottom vessels (Fig. 4). Incubation with L-Arg (Fig. 4), but not D-Arg (data not shown), prevented the inhibition of the MCh-induced relaxation by L-NNA in endothelium-intact top and bottom pulmonary arteries. MB inhibited the relaxation of both top and bottom pulmonary arteries (Fig. 4).

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Effect of N-nitro-L-arginine (L-NNA; 3 × 105 M; n = 7 horses) and inhibition of its effect by L-arginine (L-Arg; 102 M), and effect of methylene blue (MB; 105 M; n = 7 horses) on MCh-induced relaxation of endothelium-intact top (A) and bottom (B) pulmonary arteries. Effect of L-Arg was assessed by incubating vessels with L-Arg 20 min before L-NNA incubation (20 min). Data are means ± SE. * L-NNA and MB significantly (P < 0.05) inhibited the MCh-induced relaxation of both top and bottom pulmonary arteries, and pretreatment with L-Arg prevented the L-NNA-induced inhibition of relaxation (relaxation not different from control).

Compared with the major effects of NO-synthase inhibition on endothelium-dependent relaxations to MCh, cyclooxygenase inhibition had only trivial effects. Meclo inhibited slightly the relaxation of top, but not bottom, vessels. Furthermore, the combination of Meclo + L-NNA was no more effective than L-NNA alone in inhibiting the MCh-induced relaxation of top and bottom arteries.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

This study demonstrates a fundamental difference in the endothelium-dependent responses of top and bottom vessels of the equine lung. MCh, the prototypical endothelium-dependent vasodilator, caused top vessels to relax, consistent with the effect of cholinergic agonists in most vascular beds (3, 4). In contrast, bottom vessels contracted after a brief minimal relaxation (Fig. 2). The elimination of relaxation by EC removal demonstrates that the endothelium is required for the MCh-induced relaxation of both top and bottom pulmonary arteries. Furthermore, EC removal enhanced MCh-induced contraction of bottom arteries (Fig. 3), suggesting that the endothelium modulates the contraction of arteries from the bottom of the lung and may be a regulator of smooth muscle tone in both top and bottom pulmonary arteries. The differences in the MCh-induced endothelium-dependent relaxation and contraction between top and bottom vessels were maintained over a wide range of MCh concentrations (Fig. 3) and demonstrate that these differences are concentration dependent and may reflect variations in endothelial function. Differences in endothelium-dependent relaxation were also observed in response to BK and A-23187 (Fig. 3). The responses of the bottom pulmonary arteries were significantly shifted to the right compared with those of top vessels, indicating a decreased sensitivity of bottom pulmonary arteries to these agonists. Because these differences in relaxation between top and bottom vessels occur in response to a variety of agonists that are both receptor dependent (MCh and BK) and independent (A-23187), they must reflect fundamental regional differences in the biological responses of EC from top and bottom pulmonary arteries.

Because contraction of bottom vessels was observed only with MCh (Fig. 3), the rest of our discussion will focus only on the mechanisms of regional differences in endothelium-dependent relaxation. The endothelium is actively involved in vascular regulation by releasing an endothelium-derived relaxing factor, which in most species is NO (3, 4, 13). NO diffuses from the EC to the smooth muscle where it activates guanylate cyclase to increase cGMP, resulting in vasodilation (3, 8).

The central role of NO and guanylate cyclase/cGMP in the relaxation of both top and bottom pulmonary arteries was demonstrated by our observation that the NO-synthase inhibitors L-NNA and L-NMMA, and the guanylate cyclase inhibitor MB, eliminated the MCh-induced relaxation (Fig. 4). Furthermore, the inhibitory effect of L-NNA on MCh-induced relaxation was prevented by L-Arg (Fig. 4) but not D-Arg. These data suggest that the relaxation of top and bottom vessels, although different in magnitude, is mediated by the same mechanism, the endothelial release of NO and activation of the guanylate cyclase/cGMP system. In some vascular beds, arterial dilation may be mediated by the endothelial release of prostanoids (3, 11, 12). This is an unlikely mechanism, as MCh only slightly inhibited the relaxation of top but not bottom vessels.

One possible mechanism for the regional difference in endothelium-mediated relaxation is a difference in responsiveness of vascular smooth muscle to NO. SNP, like NO, stimulates guanylate cyclase, resulting in increased intracellular cGMP and relaxation of vascular smooth muscle (8). We observed identical relaxation responses to SNP in top and bottom vessels. Therefore, the regional differences in endothelium-mediated responses do not result from differences in responsiveness of the smooth muscle to increases in cGMP but must be due to differences in the magnitude of the endothelial production or release of NO.

Regional differences in endothelium-mediated relaxation between top and bottom vessels could also be caused by a difference in inactivation of NO by oxygen radicals (6, 14, 20) or by NO diffusion to the vascular smooth muscle. Our data cannot rule out these possibilities.

Zellers and Vanhoutte (21) reported a decreased endothelium-dependent relaxation of large (5-7 mm OD) compared with small (2-3 mm OD) porcine pulmonary arteries in response to acetylcholine and BK but not to A-23187 or SNP. These studies suggest that the heterogeneity in endothelium-mediated relaxation between large and small vessels results from differences in receptor function (e.g., number or affinity of receptors, or signal-transduction pathways), whereas our data indicate that differences in endothelium-mediated relaxation between top and bottom pulmonary arteries are receptor independent and mediated by NO.

The above study, reporting differences in vasodilation between central and peripheral arteries (21), raises the possibility that the increased endothelium-mediated relaxation of our top pulmonary arteries results from the fact that top arteries are more peripheral than bottom vessels. Top arteries were ~1.5-2 times farther from the hilum than were bottom arteries; therefore, regional differences in relaxation may have been simply distance dependent rather than gravity dependent. In pilot studies (n = 2), we sampled different branches of pulmonary arteries from the top and bottom of the lung that allowed comparisons of vessels with equal diameter and distance from the hilum. Because top arteries always relaxed more than vessels from the bottom of the lung, we think that it is unlikely that the regional differences in endothelium-dependent relaxation described herein result from differences in the distance of the sampled vessels from the hilum.

What is the physiological significance of the regional differences in endothelial function in the lung? The endothelium-dependent agonists used in this study are unlikely to constitute a physiological stimulus to activate EC in vivo. However, like the arterial relaxation caused by these agonists, increases in blood flow, such as occur during exercise (16), result in active arterial dilation through shear forces that activate EC to produce NO and/or prostanoids (10-12). If regional differences in endothelium-dependent flow-mediated vasodilation similar to those observed in this study occur during exercise, then the greater dilation of caudodorsal pulmonary arteries may be responsible for preferential redistribution of pulmonary blood flow to dorsal lung regions in exercising horses (1). Therefore, the greater endothelium-mediated relaxation of top pulmonary arteries may be a mechanism that offsets the effect of gravity on vascular pressures and distribution of blood flow in the lung and results in gravity-independent distribution of blood flow in exercising animals (1, 7, 15). This mechanism may also explain the gravity-independent preferential perfusion of caudodorsal lung regions in anesthetized (2, 5, 9) and awake resting mammals (1, 7). There is evidence that EC regulate smooth muscle tone even in resting mammals (19) through a basal shear-induced NO release. If top pulmonary arteries release more NO in response to shear, it is possible that preferential dilation of top pulmonary arteries counteracts the effect of gravity on the distribution of blood flow, resulting in the lack of a dorsoventral perfusion gradient.