Endothelin-1 Reduces Microvascular Fluid Permeability through Secondary Release of Prostacyclin in Cat Skeletal Muscle

Microvascular Research 63, 50 – 60 (2002) doi:10.1006/mvre.2001.2365, available online at http://www.idealibrary.com on

Endothelin-1 Reduces Microvascular Fluid Permeability through Secondary Release of Prostacyclin in Cat Skeletal Muscle Peter Bentzer,* Staffan Holbeck,* ,† and Per-Olof Gra¨nde* ,† *Departments of Physiological Sciences and †Department of Anesthesia and Intensive Care, University of Lund and University Hospital of Lund, Lund, Sweden Received February 12, 2001; published online October 29, 2001

The aim of the study was to analyze effects of various plasma concentrations of the vasoconstrictor endothelin-1 on microvascular fluid permeability and on transcapillary fluid exchange. We also analyzed whether the permeability-reducing substance prostacyclin is involved in the permeability effects of endothelin-1, as prostacylin is suggested to be released via ET B receptor stimulation. The study was performed on an autoperfused cat calf muscle preparation, and a capillary filtration coefficient (CFC) technique was used to estimate variations in microvascular fluid permeability (conductivity). Intraarterial infusion of endothelin-1 in low doses (5 and 10 ng/min/100 g muscle) caused transcapillary absorption, whereas higher doses (20 – 40 ng/min/100 g) induced filtration despite further vasoconstriction. Low-dose endothelin-1 had no significant effect on CFC, while CFC was reduced to at most 55% of baseline at higher doses (P < 0.01). Simultaneous local intraarterial infusion of the prostacyclin synthesis inhibitor tranylcypromine restored CFC to 114% of baseline (P < 0.01) and further increased vascular resistance. A low, nonvasodilator dose of prostacyclin given intravenously counteracted the tranylcypromine effect on CFC. The decreased CFC induced by a high dose of endothelin-1 was counteracted by the ET B receptor antagonist BQ-788 with no change in vascular resistance (P < 0.05). We conclude that the

50

decreased CFC following high doses of endothelin-1 can be attributed to a decrease in microvascular hydraulic conductivity, mediated by secondary release of prostacylin via stimulation of the ET B receptor. Endothelin-1 may induce edema through postcapillary vasoconstriction. © 2001 Elsevier Science Key Words: capillary filtration coefficient; capillary permeability; endothelium; endothelin-1; ET B receptor antagonist; whole organ; prostacyclin; transcapillary fluid exchange; microvessels; edema.

INTRODUCTION Endothelin-1 is the main endothelium-derived contracting factor discovered to date. It is a 21-amino-acid peptide synthesized by the endothelium and acts via two G-protein-coupled receptor types, the ET A and ET B receptors (Goto et al., 1996). In the vasculature, both the ET A and ET B receptors are present on the smooth muscle cells and activation of these receptors elicits vasoconstriction, while the ET B receptor is the predominant receptor on the endothelial cells and mainly elicits vasodilatation. In addition to its vasoconstricting effects, endothelin-1 may promote leukocyte adhesion and platelet aggregation. A role for endothelin-1 in several pathophysiological conditions, 0026-2862/01 $35.00 © 2001 Elsevier Science All rights reserved.

51

Endothelin-1 and Microvascular Permeability

like brain ischemia, brain trauma, myocardial infarction, and sepsis, has been suggested (Goto et al., 1996; Wanecek et al., 2000; Barone et al., 2000). These conditions are all associated with increased levels of endothelin-1 in plasma or cerebrospinal fluid (Ziv et al., 1992; Donckier et al., 1991; Hama et al., 1997). The pathophysiological conditions in which increased plasma concentrations of endothelin-1 have been observed are associated with increased transcapillary protein leakage and tissue edema, symptoms that can be related both to increased microvascular hydrostatic pressure and to increased microvascular permeability. Studies on permeability effects of endothelin-1, however, have not confirmed any general endothelin-induced permeability effects. In the pulmonary and myocardial circulations, endothelin-1 has been suggested either to increase or not to affect permeability, and several studies have suggested that endothelin-1 induces edema formation mainly through an increase in microvascular hydrostatic pressure (Barnard et al., 1991; Rodman et al., 1992; Helset et al., 1993; Dupuis et al., 1997). Endothelin-1 has been reported to increase the permeability for proteins in cat skeletal muscle (Porter et al., 1999, 2000). In contrast to these results, it was recently shown that endothelin-1 induced a decrease in fluid permeability in rat mesenteric venules (Victorino et al., 1999), an effect suggested to be mediated through activation of the ET B receptor (Victorino et al., 2000). The vascular effects of endothelin-1 can be referred to a direct effect on vascular smooth muscle cells and on endothelial cells, but may also be an effect of secondary release of vasoactive substances. Secondary release of prostacyclin and nitric oxide from the endothelium is suggested to contribute to ET B receptorinduced vasodilation (Hirata et al., 1993; Mc Murdo et al., 1994; Goto et al., 1996). Prostacyclin is a main product of arachidonic acid metabolism in the endothelial cell, and its platelet and leukocyte antiaggregating and antiadhesion effects are of importance for preservation of a normal microcirculation (Erlandsson et al., 1991; Vane and Botting, 1995). At high doses it also induces vasodilatation. In addition, low-dose prostacyclin was shown to have permeability-reducing properties and was recently suggested to be a

physiological modulator of this parameter (Mo¨ller and Gra¨nde, 1999; Bentzer et al., 1999). Based on these considerations, the present in vivo study evaluated the effects of various doses of endothelin-1 on microvascular fluid permeability (hydraulic conductivity), transcapillary fluid exchange, and vascular resistance and to what extent the effects could be referred to a secondary release of prostacyclin. The study was performed on a denervated autoperfused cat skeletal muscle preparation, and changes in microvascular fluid permeability (conductivity) were evaluated by using a capillary filtration coefficient (CFC) method.

MATERIAL AND METHODS Material and Anesthesia The study was approved by the Local Ethics Committee for animal subjects. Twenty-six male cats (3.9 to 5.3 kg body wt) from a professional breeder were used. The animals were treated in accordance with the Guidelines of the National Institutes of Health for care and use of laboratory animals. Anesthesia was induced with ␣-chloralose (40 mg/kg), urethane (10 mg/kg), and a small dose of pentobarbital (3– 4 mg/ kg) and continued with a continuous infusion of ␣-chloralose (5 mg/h/kg body wt). The animals were mechanically ventilated with air (Servo 900, Siemens– Elema, Solna, Sweden) through a tracheal cannula, keeping a normal end-tidal PCO 2 of 36 – 41 Torr (4.8 – 5.5 kPa) (Eliza CO 2 analyzer, Gambro Engstro¨m, Bromma, Sweden). Body temperature was kept at 37.5°C with a heating pad controlled by an esophageal thermistor. Blood gases and hematocrit were measured regularly (i-Stat, Princeton, NJ) to ensure that these parameters stayed within normal limits.

Skeletal Muscle Preparation The observations were made on a well-established model of the denervated lower leg muscles of the cat right hindlimb (calf muscles) (Kongstad and Gra¨nde, 1998). The gastrocnemius muscle is the largest muscle

© 2001 Elsevier Science All rights reserved.

52

of the preparation, but other muscles, such as the soleus and the tibialis muscles, are also included. The paw and the skin were removed, and the muscle preparation was isolated from the body, with the femur cut, the marrow cavity plugged, and the lymph vessels ligated. Great care was taken with ligatures and electrocautery to achieve complete hemostasis. The muscle was autoperfused from the animal via an arterial shunt from the right femoral to the right popliteal artery. Venous blood was drained from the popliteal vein via a tube, with its outlet open to atmospheric pressure and placed above an open funnel connected to the right external jugular vein. Mean arterial pressure in the preparation was kept constant throughout each experiment by the adjustment of a screw clamp placed around the arterial shunt. The constant arterial pressure varied among cats from 75 to 100 mm Hg. The screw clamp was controlled by an electric motor regulated by a feedback circuit via a pressure transducer. By this feedback control mean arterial pressure was kept at a constant level, uninfluenced by the CFC procedure per se (Bentzer et al., 2001). A stabilizing period of about 90 min passed after completion of the preparation before any measurements were taken.

Determination of Total Vascular Resistance Arterial inflow pressure and venous outflow pressure of the muscle were monitored via T-tubes in the shunts close to the cannulated popliteal artery and vein. Blood flow to the muscle was continuously recorded using an ultrasonic flow meter (Transonic Systems, T 106 Inc., Ithaca, New York) in the arterial shunt, calibrated by measurement of venous blood flow with a graded cylinder. Total vascular resistance was recorded continuously by electronically dividing signals representing arterial–venous pressure fall with blood flow. Vascular resistance is expressed in PRU (mm Hg/ml/min/100 g muscle). The muscle weight of the preparation was obtained by linear interpolation from previous data showing that the weight of the skeletal muscles in the preparation is 1.45% of the body weight for a 6.0 kg cat and 1.55% for a 4.0 kg cat (Kongstad and Gra¨nde, 1998). All parameters were recorded on a Grass Polygraph (Grass Instruments, Quincy, MA).

© 2001 Elsevier Science All rights reserved.

Bentzer, Holbeck, and Gra¨nde

Determination of Muscle Volume Changes The muscle preparation with intact blood supply was enclosed in a temperature-controlled (37°C) plethysmograph sealed with a two-component silicon elastomer (Silicon-Kautschuck RTV-M400, Wacker Chemie, Munich, Germany). The plethysmograph was filled with Ringer–acetate solution. Ciprofloxacine antibiotic (Ciproxin, Bayer) at a dose of 125 mg was added to the Ringer–acetate solution to avoid bacterial growth in the plethysmograph. The plethysmograph was in a fluid connection via a tube to an open waterfilled reservoir placed on a force transducer. The volume changes of the muscle were continuously recorded by monitoring the change in weight of the reservoir. The zero baseline of the reservoir and arterial and venous pressures were defined at the midlevel of the muscle. Before start of the experimental interventions, the venous pressure in the muscle was set at 6 –9 mm Hg by vertical variation of the level of the outlet of the venous tube so that an approximately isovolumetric state was achieved in the preparation.

Determination of Capillary Filtration Coefficient CFC is defined as the filtration rate induced by a fixed increase in transcapillary hydrostatic pressure and is dependent on the fluid permeability (hydraulic conductivity) and the surface area available for the induced filtration (Folkow and Mellander, 1970). The CFC method used in the present study has been carefully evaluated and it has been shown that CFC is independent of variations in vascular tone and the number of perfused capillaries in a large range of tonus and plasma flow variations (Kongstad and Gra¨nde, 1998; Bentzer et al., 2001). CFC was calculated from the rate of increase in tissue volume per minute following a fixed decrease in tissue pressure of 5 mm Hg, as obtained by a decrease in the intraplethysmographic pressure, in turn obtained by a lowering of the reservoir. The decreased tissue pressure causes an initial fast increase in tissue volume, mainly representing an increase in venous blood volume, followed by a slow increase in tissue volume, representing the induced transcapillary fluid filtration. Capillary filtra-

53

Endothelin-1 and Microvascular Permeability

tion coefficient values were calculated from the average curve slope between the third, and fourth minute after the tissue pressure decrease and related to the volume curve slope just preceding the tissue pressure decrease. At this time point the curve slope represents only the induced filtration, as shown previously at normal or reduced vascular tone (Watson and Wolf, 1989; Kongstad and Gra¨nde, 1998), which in the present study was shown to be valid also at a significantly increased vascular tone (see Results and Fig. 1). CFC measurements were always performed at a constant vascular resistance both during registration of the control volume slope and during registration of the induced filtration in order to eliminate blood volume capacitance effects in the volume registration.

Drugs The animals were anticoagulated with heparin (1000 U/kg body wt) before vessel cannulation, supplemented by 200 U/h/kg. The animals received continuous volume substitution at a rate of 5 ml/h/kg by iv infusion of a mixture of Ringer–acetate (Pharmacia, Stockholm, Sweden) and 20% albumin (Pharmacia) at a ratio of 10:1. Prostacyclin (Flolan, Wellcome, London, UK) was dissolved in a glycine buffer vehicle. There are no vehicle-induced circulatory effects at the infusion rates used in this study (Jahr et al., 1995). Endothelin-1 (Sigma Chemical Co., St. Louis, MO), the prostacyclin synthesis inhibitor tranylcypromine (Gryglewski et al., 1976) (ICN, Aurora, Ohio), and the sodium salt of the ET B receptor antagonist BQ-788 (Dimethylpiperidino-MeLeu-DTrp (COOMe)-D-Nle 260601, Peptides International, Louisville, KY) (Ishikawa et al., 1994) were dissolved in isotonic saline. All substances given intraarterially were administered at a low infusion rate, ⬍0.04 ml/min (Model 11, Harvard Apparatus, S. Natick, MA), to avoid rheologic interference with the normal blood stream.

Experimental Protocol The following parameters were recorded continuously: arterial pressure, venous pressure, arterial blood flow, total vascular resistance, and tissue volume variations. Except for the methodological exper-

iments analyzing the time point at which the volume curve slope during the CFC procedure represents only filtration, the study consisted of four main series of experiments with the objective of studying the effects on vascular resistance, transcapillary fluid exchange, and CFC of (1) local intraarterial infusion of endothelin-1 at infusion rates of 5, 10, 20, and 40 ng/min/100 g muscle; (2) intraarterial infusion of endothelin-1 at a dose of 30 – 40 ng/min/100 g muscle, followed by simultaneous intraarterial infusion of tranylcypromine at a dose of 6 ␮g/min/100 g muscle; (3) intraarterial infusion of tranylcypromine at the same dose as above followed by simultaneous infusion of endothelin-1 at a dose of 30 – 40 ng/min/100 g muscle; and (4) intraarterial infusion of endothelin-1 at a dose of 30 – 40 ng/min/100 g followed by a simultaneous infusion of the ET B receptor antagonist BQ-788. BQ-788 was given intraarterially at a dose of 13 ␮g/min/100 g, producing a plasma concentration shown to cause effective blockade of the ET B receptor (Ishikawa et al., 1994). The washout time elapsing between each series of experiments was 1–2 h, a time long enough for adequate recovery of the recorded parameters. A new control was established before each series. All measurements were performed during the steady state of vascular resistance and blood flow, normally achieved 10 –20 min after the start of each infusion rate. Each animal was used in one or two series of experiments.

Statistics Statistical analyses were performed with one-factor repeated-measures analyses of variance. If the null hypothesis was rejected, differences were isolated using the Student–Newman–Keuls test. Values for transcapillary fluid exchange were not normally distributed and were analyzed by repeated measures analysis on ranks followed by the Student–Newman– Keuls test. Results are presented as mean values ⫾ SE.

© 2001 Elsevier Science All rights reserved.

54

FIG. 1. Changes in the volume curve slope over time during the first 6 min following a decrease in tissue pressure by 5 mm Hg at control (upper curve) (n ⫽ 6) and during infusion of endothelin-1 (ET-1) (lower curve) at a dose of 30 – 40 ng/min/100 g (n ⫽ 6).

RESULTS Figure 1 shows the tissue volume curve after a transmural pressure increase of 5 mm Hg at normal and increased vascular tone, the latter obtained by intraarterial infusion of endothelin-1 at a dose of 30 – 40 ng/ min/100 g muscle. Muscle volume changes were recorded for 6.5 min after the decrease in tissue pressure, and each time point represents the mean value of the volume curve slope from 30 s before to 30 s after each whole minute. The figure summarizes data from three cats, of which two recordings were performed during control conditions and two recordings during the infusion of endothelin-1 in each cat. Vascular resistance at control was 17.7 ⫾ 2.0 PRU and increased to 54.1 ⫾ 6.6 PRU during the infusion of endothelin-1. Blood flow decreased from 3.8 ⫾ 0.2 ml/min/100 g at control to 1.3 ⫾ 0.1 ml/min/100 g. As can be seen, there was a large decrease in volume curve slope over time during the first 2 min following the tissue pressure decrease both at control conditions and during endothelin-1 infusion. From the third minute the volume curve slope shows an equally large and much slower decrease in both groups. This shows that the curve slope between the third and fourth minute can be used for measurement of CFC also at a raised vascular tone (see Materials and Methods) as done in the present study.

© 2001 Elsevier Science All rights reserved.

Bentzer, Holbeck, and Gra¨nde

Start of the intraarterial endothelin-1 infusion induced a transient vasodilator response lasting for 4 –5 min, followed by a vasoconstrictor response. Figure 2 shows vascular resistance effects of increasing doses of endothelin-1 at the steady state and effects on CFC relative to the control values before start of the infusion (n ⫽ 7). Total vascular resistance and CFC at control were 14.4 ⫾ 1.5 PRU and 0.0088 ⫾ 0.0004 ml/min/mm Hg/100 g, respectively. As seen, there is a dose-dependent increase in vascular resistance, which is significant at infusion rates above 10 ng/ min/100 g. There was no effect on CFC at an endothelin-1 infusion rate of 5 and 10 ng/min/100 g, and a significant decrease at higher doses. During the infusion of low doses (5 and 10 ng/min/100 g) of endothelin-1, there was a small transcapillary absorption relative to control at the steady state ( p ⬍ 0.05). This absorption gradually declined at increasing doses of endothelin-1 and turned to transcapillary fluid filtra-

FIG. 2. Effects of increasing doses of endothelin-1 on total vascular resistance (top) and on capillary filtration coefficient (CFC) (bottom) at the steady state relative to the control values before start of the infusion (n ⫽ 7). There was a significant decrease in CFC relative to control during the endothelin-1 infusion at a dose of 20 ng/min/100 g (P ⬍ 0.01), with a further decrease in CFC at the higher dose of 40 ng/min/100 g (P ⬍ 0.05). Vascular resistance increased significantly during infusion of endothelin-1 at a dose of 20 ng/min/100 g (P ⬍ 0.05) and increased further at the higher dose (P ⬍ 0.05). **P ⬍ 0.01, *P ⬍ 0.05.

Endothelin-1 and Microvascular Permeability

FIG. 3. Effects of increasing doses of endothelin-1 on transcapillary fluid exchange ( J V) relative to control. Endothelin-1 induced absorption at the lower doses (P ⬍ 0.05), whereas it induced filtration at the higher doses (*P ⬍ 0.05) (n ⫽ 7).

tion relative to control at the highest dose of endothelin-1 (Fig. 3, P ⬍ 0.05). In the second series of experiments (Fig. 4) (n ⫽ 9), endothelin-1 was given intraarterially at a dose in the range of 30 to 40 ng/min/100 g, aiming at an increase in vascular resistance of about 200% of control. The vascular resistance increased to 226 ⫾ 18% of the control value of 10.8 ⫾ 1.6 PRU (P ⬍ 0.05). This dose of endothelin-1 caused a significant decrease in CFC to 55 ⫾ 4% of control from a control value of 0.0102 ⫾ 0.0008 ml/min/mm Hg/100 g (P ⬍ 0.01). After the start of a simultaneous intraarterial infusion of the prostacyclin synthesis inhibitor tranylcypromine at a dose of 6 ␮g/min/100 g, total vascular resistance increased further to a value of 297 ⫾ 38% of control, and this occurred simultaneously with a restitution of CFC to a value of 114 ⫾ 8% of the control values prevailing before the start of endothelin-1 (P ⬍ 0.01). In two of the experiments, prostacyclin was now given intravenously at a dose of 1 ng/min/kg body wt during the endothelin-1 and the tranylcypromine infusions. The prostacyclin infusion had no effect on total vascular resistance, but reduced CFC to 56% of the initial control. This CFC level was similar to that prevailing before the start of tranylcypromine. In the third series of experiments (n ⫽ 6, Fig. 5), tranylcypromine was started before the endothelin-1 infusion using the same doses as above. In these experiments, the mean value of vascular resistance during tranylcypromine was 135 ⫾ 16% of the control

55

value of 13.9 ⫾ 4.0 PRU before the start of the tranylcypromine infusion, and mean CFC increased to 117 ⫾ 2% (P ⬍ 0.05) of the initial control value of 0.0080 ⫾ 0.0005 ml/min/mm Hg/100 g. The start of the simultaneous endothelin-1 infusion increased vascular resistance further to 247 ⫾ 18% of the initial control (P ⬍ 0.05), while CFC remained unchanged, showing a mean value of 121 ⫾ 3% of the initial control. In the fourth series of experiments (n ⫽ 7), an endothelin-1 infusion given at the high dose of 30 – 40 ng/min/100 g reduced CFC to 62 ⫾ 4% of a control value of 0.0085 ⫾ 0.0002 ml/min/mm Hg/100 g (P ⬍ 0.05) and increased vascular resistance to 261 ⫾ 16% of 18.9 ⫾ 1.7 PRU (P ⬍ 0.05). Following the start of the simultaneous infusion of the ET B receptor antagonist BQ-788, CFC increased to 106 ⫾ 5% of control (P ⬍ 0.05), while vascular resistance remained unchanged (Fig. 6).

FIG. 4. Effects of a high dose of endothelin-1 (ET-1) (30 – 40 ng/ min/100 g) given alone intraarterially followed by a simultaneous infusion of the prostacyclin synthesis inhibitor tranylcypromine (Trc) (6 ␮g/min/100 g) on total vascular resistance (top) and on capillary filtration coefficient (CFC) (bottom) relative to the control values before start of the infusion (n ⫽ 9). This high dose of ET-1 caused a decrease in CFC (P ⬍ 0.01), and simultaneous infusion of Trc restituted CFC to a value similar to that of the control (P ⬍ 0.01). ET-1 increased the vascular resistance (P ⬍ 0.05), and simultaneous infusion of Trc caused further increase in vascular resistance (P ⬍ 0.05). *P ⬍ 0.05, **P ⬍ 0.01.

© 2001 Elsevier Science All rights reserved.

56

FIG. 5. Effects of the prostacyclin synthesis inhibitor tranylcypromine (Trc) given alone intraarterially (6 ␮g/min/100 g) followed by a simultaneous infusion of a high dose of endothelin-1 (ET-1) (30 – 40 ng/min/100 g) on total vascular resistance (top) and on capillary filtration coefficient (CFC) (bottom) relative to the control values before the start of the Trc infusion (n ⫽ 9). Trc increased CFC relative to control (P ⬍ 0.05), but no further change in CFC was observed following infusion of ET-1. There was an increase in vascular resistance during the simultaneous infusion of ET-1 and Trc (P ⬍ 0.05). *P ⬍ 0.05.

Bentzer, Holbeck, and Gra¨nde

by increases in vascular resistance and precapillary sphincter tone (Folkow and Mellander, 1970). There are studies, however, suggesting that CFC in skeletal muscle is unaltered by changes in vascular tone and the number of perfused capillaries (Watson et al., 1984; Gamble et al., 1997). This issue has been examined in detail in our laboratory, the results showing that CFC values from the skeletal muscle model used in the present study are unaltered during both pharmacologically induced decreases and increases in vascular resistance (Kongstad and Gra¨nde, 1998; Bentzer et al., 2001). In addition, it was also shown that a reduction in the number of perfused capillaries, accomplished by the injection of microspheres, did not significantly alter CFC (Bentzer et al., 2001). According to these results, it is unlikely that the observed decrease in CFC due to the vasoconstrictor endothelin-1 was caused by a reduction in the number of perfused capillaries. Such an interpretation also finds support from the present study, as (1) the decreased CFC during endothelin-1 infusion was restored by the prostacyclin synthesis inhibitor tranylcypromine, with a further increase in vascular resistance; (2) CFC was unaltered by endo-

DISCUSSION The present results show that endothelin-1 has a dose-dependent CFC-reducing effect, causing a decrease in CFC to about 55% of the control CFC value at the highest dose given, simultaneously with an increase in total vascular resistance. Endothelin-1 induced absorption at the lower and filtration at the higher doses. Infusion of the prostacyclin synthesis inhibitor tranylcypromine simultaneously with infusion of the high dose of endothelin-1 more than restored CFC. The CFC-reducing effect of a high dose of endothelin-1 was counteracted by infusion of the ET B receptor antagonist BQ-788. CFC is dependent both on the fluid permeability of the capillary wall and on the surface area available for the filtration induced by the applied increase in transcapillary pressure. The conventional view is that the latter factor is influenced by changes in the number of perfused capillaries, which in turn may be decreased

© 2001 Elsevier Science All rights reserved.

FIG. 6. Effects of a high dose of endothelin-1 (ET-1) (30 – 40 ng/ min/100 g) given alone intraarterially followed by a simultaneous intraarterial infusion of the ET B receptor antagonist BQ-788 (13 ␮g/min/100 g) on total vascular resistance (top) and on capillary filtration coefficient (CFC) relative to control (bottom) (n ⫽ 7). ET-1 increased vascular resistance and decreased CFC relative to control (P ⬍ 0.05), and simultaneous infusion of BQ-788 restituted CFC (P ⬍ 0.05) with no effect on vascular resistance. *P ⬍ 0.05.

Endothelin-1 and Microvascular Permeability

thelin-1 in the experiments in which endothelin-1 was given after the start of tranylcypromine, although vascular resistance increased by approximately 250%; and (3) CFC was restituted by the ET B receptor blocker BQ-788 without any simultaneous effect on vascular tone. The filtration induced by the decrease in tissue pressure causes a plasma flow-dependent increase in plasma colloid osmotic pressure, which may cause an underestimation of CFC (Renkin, 1984). As mentioned above, we have previously shown that CFC in the present model is not altered by decreases in plasma flow (Bentzer et al., 2001), which suggests that a flow rate-dependent increase in plasma colloid osmotic pressure cannot explain our results. We therefore conclude that the observed changes in CFC by endothelin-1 is not an effect of alterations in number of perfused capillaries, vascular tone, or blood flow, but a consequence of alterations in microvascular fluid permeability. Previous studies on isolated rat mesenteric venules perfused with an artificial perfusate have suggested that fluid permeability is reduced by endothelin-1 (Victorino et al., 1999, 2000), findings in agreement with results in the present study. Such a response in isolated microvessels does not, however, necessarily reflect the net effect on fluid permeability in a whole vascular bed. The present results are thus important, as they show that the net effect of endothelin-1 in a blood-perfused whole organ preparation is decreased fluid permeability. The effects of endothelin-1 on fluid permeability have previously been investigated by several groups in lung preparations with differing results. Thus, endothelin-1 either decreased or did not change fluid permeability in the rat lung (Barnard et al., 1991; Rodman et al., 1992; Helset et al., 1993), whereas fluid permeability was unaltered in rabbit lung and dog lung (Barnard et al., 1991). That these results are not fully compatible with our results or the results presented by Victorino et al. (1999, 2000) may be explained by variation in receptor subtype populations between organs/species (Barnard et al., 1991; Seo et al., 1994) or alternatively by differences in perfusate composition (Rodman et al., 1992; Helset et al., 1993). To our knowledge, only two previous studies have investigated endothelin-1 effects on permeability in skeletal muscle. In these studies, performed on cat

57

skeletal muscle, it was concluded that endothelin-1 decreased the reflection coefficient (i.e., increased permeability) for macromolecules by a mechanism related to increased vascular pressure and/or increased shear stress following activation of the ET A receptors on vascular smooth muscle cells (Porter et al., 1999, 2000). There may be several reasons for the apparent discrepancy between the present study and these previous studies. The preparation used by Porter et al. (1999, 2000) was perfused at a constant flow. This implies that an increase in vascular resistance during infusion of endothelin-1 caused a large increase in microvascular pressure, which may increase permeability by a pressure-induced injury of the endothelium. In contrast, the present preparation is perfused at a constant arterial pressure, which implies only minor variations in capillary pressure following increases in vascular resistance. Differences in the perfusate composition may also explain the difference in results as blood components, including platelets, are suggested to be important for the synthesis of prostacyclin (Gryglewski et al., 1991). From the doses of endothelin-1 given in the present study and the estimated plasma flow values of 1–3 ml/min/100 g, it can be calculated that the lowest dose of endothelin-1 results in a plasma concentration of 1–2 ng/ml. The highest dose given corresponds to a plasma concentration of 30 – 40 ng/ml and is similar to that used in most previous studies on permeability effects of endothelin-1 (Barnard et al., 1991; Rodman et al., 1992; Porter et al., 1999, 2000). Normal plasma concentrations of endothelin-1 in man are shown to be in the range of 2– 6 pg/ml and may increase two to eight times above basal values during pathophysiologic conditions (Donckier et al., 1991). This could be taken to indicate that the doses used by us and others will result in plasma concentrations clearly above normal. However, it is reasonable to assume that plasma concentrations of endothelin-1 are several orders of magnitude lower than those prevailing close to the endothelial wall. Endothelin-1 has a very short plasma half clearance time (1–2 min), clearance taking place rapidly by passage through the lungs, liver, and kidney (Wanecek et al., 2000), and increases in tissue content of endothelin-1 have been shown to be several times as high as those in plasma (Serneri et al., 1995).

© 2001 Elsevier Science All rights reserved.

58

Furthermore, the threshold for increases in vascular resistance by endothelin-1 is reported to be about 3 ng/ml (Rodman et al., 1992), whereas blockade of endothelin receptors decreases blood pressure at endothelin-1 plasma concentrations far below those needed to elicit vasoconstriction (Ohlstein et al., 1993). Based on these considerations, we believe that the doses of endothelin-1 used in the present study result in concentrations close to the endothelial wall that lie within the physiological and pathophysiological range. Strong support for the hypothesis that the endothelin-induced decrease in CFC is not a direct endothelin-1 effect but rather an effect mediated through the secondary release of prostacyclin can be obtained from the present results for the following reasons: (1) the prostacyclin synthesis inhibitor tranylcypromine counteracted the endothelin-induced CFC reduction; (2) restoration of plasma prostacyclin levels during tranycypromine infusion counteracted the tranylcypromine effect on CFC; and (3) endothelin-1 had no effect on CFC if the tranylcypromine infusion was started before the endothelin infusion. The induced increase in CFC by tranylcypromine alone before start of the simultaneous endothelin-1 infusion supports the newly presented hypothesis that basal endothelial production of prostacyclin may be a modulator of microvascular permeability (Mo¨ller and Gra¨nde, 1999). The results of the present study indicate that interactions between prostacyclin and endothelin-1 may contribute to control of microvascular fluid permeability. The ET B receptor is a G-protein-coupled receptor, and its activation may lead to release of prostacyclin and nitric oxide from endothelial cells (Hirata et al., 1993; Mc Murdo et al., 1994). As mentioned, endothelin-1 was recently suggested to decrease fluid permeability in isolated rat mesenteric venules through activation of the ET B receptor, but the mechanism mediating this effect was not investigated (Victorino et al., 2000). The present finding that blockade of the ET B receptor counteracted the endothelin-1-induced decrease in CFC, and to the same degree as obtained by tranylcypromine, suggests that ET B receptor stimulation exerts its permeability-reducing effect through release of prostacyclin. This novel finding is in agree-

© 2001 Elsevier Science All rights reserved.

Bentzer, Holbeck, and Gra¨nde

ment with the observation that the selective ET B agonist IRL 1620 reduces CFC in cat skeletal muscle (Ekelund et al., 1994). As expected, the vasoconstrictor endothelin-1 induced transcapillary absorption relative to the control, which most likely was an effect of reduced hydrostatic capillary pressure. The reason this absorption was not increased further with increasing doses of endothelin-1 despite a further increase in vascular tone may have been the simultaneous decrease in CFC. However, that absorption turns toward filtration relative to control at the highest dose of endothelin-1 must be compatible with an increase in hydrostatic capillary pressure, most likely an effect of an increase in postcapillary resistance by activation of ET A receptors, as suggested previously (Dahlo¨f et al., 1990; Barnard et al., 1991; Ekelund et al., 1993). Thus, even though microvascular fluid permeability is reduced by endothelin-1, the net effect on fluid exchange in vivo may be increased edema formation (Dahlo¨f et al., 1990; Rodman et al., 1992; Ekelund et al., 1993). Such an increase in fluid filtration will cause increased convective transport of protein, an effect that may explain the increased protein extravasation observed in several studies following infusion of endothelin-1 and the observation that ET A receptor blockade could counteract such an effect (Filep et al., 1993; Sirois et al., 1992). Recent studies showing that blockade of the ETA receptor can prevent edema formation both in endotoxemia and following brain trauma suggest that an endothelin-1-induced increase in capillary pressure is of importance for edema formation also in pathophysiological conditions (Barone et al., 2000; Schmeck et al., 2000). In conclusion, the present study demonstrates that endothelin-1 induces a marked decrease in microvascular fluid permeability in cat skeletal muscle, an effect most likely mediated through secondary release of prostacyclin via activation of the ET B receptor. Endothelin-1 may induce edema through increased hydrostatic capillary pressure due to postcapillary vasoconstriction.

ACKNOWLEDGMENTS This study was supported by grants from the Swedish Medical Research Council (No. 11581), from the Knut and Alice Wallenberg

Endothelin-1 and Microvascular Permeability

Foundation, and from the Medical Faculty of Lund, Sweden. We sincerely thank Christine Wikstrand and Hele´n Davidsson for skilled technical assistance.

REFERENCES Barnard, J. W., Barman, S. A., Adkins, W. K., Longenecker, G. L., and Taylor, A. E. (1991). Sustained effects of endothelin-1 on rabbit, dog, and rat pulmonary circulations. Am. J. Physiol. 261, H479 –H486. Barone, F. C., Ohlstein, E. H., Hunter, A. J., Campbell, C. A., Hadingham, S. H., Parsons, A. A., Yang, Y., and Shohami, E. (2000). Selective antagonism of endothelin-A-receptors improves outcome in both head trauma and focal stroke in rat. J. Cardiovasc. Pharmacol. 36, S357–S361. Bentzer, P., Holbeck, S. and Gra¨nde, P.-O. (1999). Prostacyclin reduces microvascular fluid conductivity in cat skeletal muscle through opening of ATP-dependent potassium channels. J. Vasc. Res. 36, 516 –523. Bentzer, P., Kongstad, L., and Gra¨nde, P.-O. (2001). Capillary filtration coefficient is independent of number of perfused capillaries in cat skeletal muscle. Am. J. Physiol. 49, H2697–H2706. Dahlo¨f, B., Gustafsson, D., Hedner, T., Jern, S., and Hansson, L. (1990). Regional haemodynamic effects of endothelin-1 in rat and man: Unexpected adverse reaction. J. Hypertension 8, 811– 817. Donckier, J. E., Hanet, C., Berbinschi, A., Galanti, L., Roberts, A., Van Mechelen, H., Pouleur, H., and Ketelslegers, J.-M. (1991). Cardiovascular and endocrine effect of endothelin-1 at pathophysiological and pharmacological plasma concentrations in conscious dogs. Circulation 84, 2476 –2484. Dupuis, J., Goresky, C. A., Rose, C. P., Stewart, D. J., Cernacek, P., Schwab, A. J., and Simard, A. (1997). Endothelin-1 myocardial clearance, production, and effect on capillary permeability in vivo. Am. J. Physiol. 273, H1239 –H1245. Ekelund, U., Albert, U., Edvinsson, L., and Mellander, S. (1993). In-vivo effects of endothelin-1 and ETa receptor blockade on arterial, venous and capillary functions in skeletal muscle. Acta Physiol. Scand. 148, 273–283. Ekelund, U., Adner, M., Edvinsson, L., and Mellander, S. (1994). Effects of selective ETb-receptor stimulation on arterial, venous and capillary functions in cat skeletal muscle. Br. J. Pharmacol. 112, 887– 894. Erlandsson, M., Bergqvist, D., Persson, N. H., and Svensjo¨, E. (1991). Modification of postischemic increase of leukocyte adhesion and vascular permeability in the hamster by iloprost. Prostaglandins 41, 157–168. Filep, J. G., Sirois, M. G., Fo¨ldes-Filep, E., Rousseau, A., Plante, G. E., Fournier, A., Yano, M., and Sirois, P. (1993). Enhancement by endothelin-1 of microvascular permeability via the activation of ETa receptors. Br. J. Pharmacol. 109, 880 – 886. Folkow, B., and Mellander, S. (1970). Measurements of capillary filtration coefficient and its use in studies of the control of capil-

59 lary exchange. In “Capillary Permeability: The Transfer of Molecules and Ions Between Capillary Blood and Tissue” (C. Crone and N. A. Lassen, Eds.), pp. 614 – 623. Munksgaard, Copenhagen. Gamble, J., Christ, F., and Gartside, I. B. (1997). The effect of passive tilting on microvascular parameters in the human calf: A strain gauge plethysmography study. J. Physiol. 498, 541–552. Goto, K., Hama, H., and Kasuya, Y. (1996). Molecular pharmacology and pathophysiological significance of endothelin. Jap. J. Pharmacol. 72, 261–290. Gryglewski, R. J., Bunting, S., Moncada, S., Flower, R. J., and Vane, J. R. (1976). Arterial walls are protected against deposition of platelet thrombi by a substance (prostaglandin x) which they make from prostaglandin endoperoxides. Prostaglandins 12, 685– 713. Gryglewski, R. J., Botting, R. M., and Vane, J. R. (1991). Prostacyclin: From discovery to clinical application. In “Cardiovascular Significance of Endothelium-Derived Vasoactive Factors” (G. M. Rybanyi, Ed.) pp. 3–37. Futura Publishing Co., New York. Hama, H., Kasuya, Y., Sakurai, T., Yamada, G., Suzuki, N., Masaki, T., and Goto, K. (1997). Role of endothelin-1 in astrocyte responses after acute brain damage. J. Neurosci. Res. 47, 590 – 602. Helset, E., Kjaeve, J., and Hauge, A. (1993). Endothelin-1-induced increases in microvascular permeability in isolated, perfused rat lungs requires leukocytes and plasma. Circ. Shock 39, 15–20. Hirata, Y., Emori, T., Eguchi, S., Kanno, K., Imai, T., Ohta, K., and Marumo, F. (1993). Endothelin receptor subtype b mediates synthesis of nitric oxide by cultured bovine endothelial cells. J. Clin. Invest. 91, 1367–1373. Ishikawa, K., Ihara, M., Noguchi, K., Mase, T., Mino, N., Saeki, T., Fukuroda, T., Fukami, T., Ozaki, S., Nagase, T., Nishikibe, M., and Yano, M. (1994). Biochemical and pharmacological profile of a potent and selective endothelin B-receptor antagonist, BQ-788. Proc. Natl. Acad. Sci. USA 91, 4892– 4896. Jahr, J., Ekelund, U., and Gra¨nde, P.-O. (1995). In vivo effects of prostacyclin on segmental vascular resistances, on myogenic reactivity, and on capillary fluid exchange in cat skeletal muscle. Crit. Care Med. 23, 523–531. Kongstad, L., and Gra¨nde, P.-O. (1998). The capillary filtration coefficient for evaluation of capillary fluid permeability in cat calf muscles. Acta Physiol. Scand. 164, 201–211. Mc Murdo, L., Thiemermann, C., and Vane, J. R. (1994). The endothelin ETB receptor agonist, IRL 1620, causes vasodilatation and inhibits ex vivo platelet aggregation in the anaesthetised rabbit. Eur. J. Pharmacol. 259, 51–55. Mo¨ller, A., and Gra¨nde, P.-O. (1999). Role of prostacyclin and nitric oxide in regulation of basal microvascular hydraulic permeability in cat skeletal muscle. J. Vasc. Res. 36, 245–252. Ohlstein, E. H., Douglas, S. A., Ezekiel, M., and Gellai, M. (1993). Antihypertensive effects of the endothelin receptor antagonist BQ-123 in conscious spontaneously hypertensive rats. J. Cardiovasc. Pharmacol. 22, S321–S324. Porter, P. L., McNamee, J. E., and Wolf, M. B. (1999). Endothelin-1 induces endothelial barrier failure in the cat hindlimb. Shock 11, 111–114.

© 2001 Elsevier Science All rights reserved.

60 Porter, L. P., McNamee, J. E., and Wolf, M. B. (2000). Interaction of endothelin-1 and nitric oxide in endothelial barrier failure in the cat hindlimb. Microcirculation 7, 347–356. Renkin, E. M. (1984). Control of microcirculation and blood–tissue exchange. In “Handbook of Physiology. The cardiovascular system: Microcirculation” (E. M. Renkin and C. C. Michel, Eds.) pp. 627– 687. Williams and Wilkins, Baltimore. Rodman, D. M., Stelzner, T. J., Zamora, M. R., Bonvallet, S. T., Oka, M., Sato, K., O’Brien, R. F., and McMurthry, I. F. (1992). Endothelin-1 increases the pulmonary microvascular pressure and causes pulmonary edema in salt solution but not blood-perfused rat lungs. J. Cardiovasc. Pharmacol. 20, 658 – 663. Schmeck, J., Heller, A., Groschler, A., Recker, A., Neuhof, H., Urbaschek, R., and Koch, T. (2000). Impact of endothelin-1 in endotoxin-induced pulmonary vascular reactions. Crit. Care Med. 28, 2851–2857 Seo, B., Oemar, B. S., Siebenmann, R., von Segesser, L., and Lu¨scher, T. F. (1994). Both ETa and ETb receptors mediate contraction to endothelin-1 in human blood vessels. Circulation 89, 1203–1208. Serneri, G. G., Modesti, P. A., Cecioni, I., Biagini, D., Migliorini, A., Costoli, A., Colella, A., Naldoni, A., and Paoletti, P. (1995). Plasma endothelin and renal endothelin are two distinct systems involved in volume homeostasis. Am. J. Physiol. 268, H1829 – H1837.

© 2001 Elsevier Science All rights reserved.

Bentzer, Holbeck, and Gra¨nde

Sirois, M. G., Filep, J. G., Rousseau, A., Fournier, A., Plante, G. E., and Sirois, P. (1992). Endothelin-1 enhances vascular permeability in conscious rats: Role of thromboxane A2. Eur. J. Pharmacol. 22, 119 –125. Vane, J. R., and Botting, R. M. (1995). Pharmacodynamic profile of prostacyclin. Am. J. Cardiol. 75, 3A–10A. Victorino, G. P., Wisner, D. H., and Tucker, V. L. (1999). Direct actions of endothelin-1 on single vessel hydraulic permeability. J. Trauma 47, 713–718. Victorino, G. P., Wisner, D. H., and Tucker, V. L. (2000). Basal release of endothelin-1 and the influence of the ETB receptor on single vessel hydraulic permeability. J. Trauma 49, 314 –319. Wanecek, M., Weitzberg, E., Rudehill, A., and Oldner, A. (2000). The endothelin system in septic and endotoxin shock. Eur. J. Pharmacol. 407, 1–15. Watson, P. D., Scott, D. R., and Wolf, M. B. (1984). Anoxia and capillary filtration coefficient in the isolated cat hindlimb. Am. J. Physiol. 246, H312–H315. Watson, P. D., and Wolf, M. B. (1989). Filtration coefficient in cat hindlimb using protein concentration changes. Am. J. Physiol. 256, H186 –H194. Ziv, I., Fleminger, G., Djaldetti, R., Achiron, A., Melamed, E., and Sokolovsky, M. (1992). Increased plasma endothelin-1 in acute ischemic stroke. Stroke 23, 1014 –1016.