Introduction

Top Abstract Introduction Methods Results Discussion References

A cardinal sign of acute inflammation is the leakage of plasma from the vascular compartment into the extracellular space. The increase in vascular permeability induced by mediators such as histamine and substance P generally is attributed to the formation of gaps between endothelial cells of postcapillary venules, as originally identified by transmission electron microscopy (Majno and Palade, 1961; Majno, 1992; Wu and Baldwin, 1992; McDonald, 1994).

More recently, localization of endothelial gaps by light microscopy has facilitated the quantification of gap number and size in blood vessels. For example, after silver nitrate staining, endothelial cell borders of normal postcapillary venules in the rat trachea are visualized as uniform, thin black lines, and no gaps are present (McDonald, 1994). By comparison, after substance P, the lines are interrupted by silver deposits about 1 µm in diameter, which mark the location of intercellular gaps (McDonald, 1994; Hirata et al., 1995). Morphometric analyses have shown that changes in number and location of such gaps correspond to the time course of leakage and location of the leaky sites (McDonald, 1994; Hirata et al., 1995; Baluk et al., 1997). When substance P is preceded by the beta-2 adrenoceptor agonist formoterol, the number of gaps and amount of leakage are both reduced the same amount (Baluk and McDonald, 1994). These findings suggest that the antileakage action of beta-2 adrenoceptor agonists results from the inhibition of gap formation and provides a reference for investigating the action of novel substances that reduce plasma leakage.

The overall aim of this investigation was to determine the mechanism of the antileakage effect of a group of peptides called "mystixins," which were discovered in the course of studies of the anti-inflammatory action of CRF (Thomas et al., 1993). These peptides have amino acid sequences of -Arg-Lys-Leu-(Leu/Met)-A*-Ile-(Leu/D-Leu)-NH₂, where A* is an anisolylated glutamic derivative or an aromatic residue. Mystixins potently inhibit plasma leakage in several models of inflammation and tissue injury. For example, the undecapeptide mystixin-11 inhibits heat-induced edema in skin for at least an hour with a median effective dose of 50 µg/kg i.v. (Thomas et al., 1993). Although chemically similar to CRF, which also has anti-edema properties, mystixins do not release adrenocorticotropin or act through known CRF receptors (Kolobov et al., 1995). From the more than 120 undeca-, octa- and heptapeptide mystixin analogs which have been synthesized and tested, the heptapeptide mystixin-7, in which D-Thi = -thienyl-D-Ala, was selected as a prototype for further investigation (Thomas et al., 1993; Wei and Thomas, 1993, 1996). Some studies were also done with the undecapeptide mystixin-11 (Thomas et al., 1993).

We approached the problem in four steps. First, we tested the efficacy of the antileakage action by determining the dose-response relationship for the inhibition by mystixin-7 of substance P-induced Evans blue leakage in rat trachea and skin (McDonald, 1994; Hirata et al., 1995; Thurston et al., 1996; Baluk et al., 1997). Second, we determined whether a reduction in the number of endothelial gaps was involved by examining the effect of leakage-inhibiting doses of mystixin-7 and mystixin-11 on the number and size of endothelial gaps made visible in tracheal blood vessels by silver nitrate staining. Third, we assessed the possible involvement of systemic hypotension in the antileakage action by measuring the effect of mystixin-7 on arterial blood pressure and heart rate for 60 min after injection of mystixin-7. Finally, we examined whether mystixin-7 can reduce vessel leakiness independent of its hypotensive effect by measuring the leakiness of vessels after aldehyde fixation.

The results of this study demonstrate that, unlike beta-2 adrenoceptor agonists, mystixin-7 reduces plasma extravasation without inhibiting endothelial gap formation.

	Methods

Top Abstract Introduction Methods Results Discussion References

Materials. Mystixin-7 (MW 1041) and mystixin-11 (MW 1356) were synthesized by standard solid-phase methods and purified by column chromatography as described previously (Thomas et al., 1993; Wei and Thomas, 1993, 1996). Mystixin-7 and mystixin-11 were dissolved in 100 µl of 0.1 N acetic acid and then diluted in 0.9% NaCl for i.v. injection at a volume of 1 ml/kg body weight.

Experimental procedures. Male pathogen-free F344 rats (200-250 g b.wt.) were purchased from Simonsen Laboratories (Gilroy, CA). All experimental protocols were approved by the Committees on Animal Research at the University of California, San Francisco, and the University of California, Berkeley. Rats anesthetized with sodium pentobarbital (50 mg/kg i.p. plus supplements) were pretreated with mystixin-7 (20, 40, 80, 100, or 160 µg/kg i.v.) or vehicle (n = 6 rats/group). Sixty minutes later Evans blue dye in 0.9% NaCl (30 mg/kg i.v.; EM Sciences, Cherry Hill, NJ) was injected, followed immediately by substance P (5 µg/kg i.v.; Peninsula Laboratories, Belmont, CA). Five minutes after the substance P, the vessels were fixed by perfusion of 1% paraformaldehyde in 0.05 M citrate buffer, pH 3.5, for 2 min through the left ventricle. The trachea and skin of the dorsal hind foot were removed, gently blotted, and weighed. Evans blue was extracted with formamide (Sigma Chemical Company, St. Louis, MO) and measured by spectrophotometry; the Evans blue content was expressed as nanograms of dye per milligram of tissue wet weight (Saria and Lundberg, 1983). For dose-response studies, the amount of Evans blue in the trachea and skin above the baseline value (no mystixin or substance P) was expressed as a percentage of the mean value for substance P-induced leakage after vehicle pretreatment.

Silver nitrate staining of endothelial gaps. Anesthetized rats were pretreated with mystixin-7 (100 µg/kg i.v.) or vehicle, and 60 min later received substance P (5 µg/kg i.v.) or vehicle (1 ml/kg). Alternatively, some rats received mystixin-11 (400 µg/kg i.v.) or vehicle, followed 5 min later by substance P. Monastral blue (30 mg/kg i.v., Sigma) was injected immediately before the substance P to mark sites of leakage (McDonald, 1994; Baluk and McDonald, 1994). Three minutes after the substance P the vasculature was perfused with fixative and the endothelial cell borders were stained with silver nitrate as described previously (McDonald, 1994). The number of endothelial gaps was determined from the number of silver deposits visible in digital color video images of specimens viewed at a projected magnification of 2,600. Counts were made on 25 silver-stained endothelial cells of postcapillary venules (diameter, 20-40 µm) in each trachea and were expressed as silver deposits per endothelial cell (n = 6 rats/group). Similarly, the size of 25 silver deposits was measured at a projected magnification of 3,700 (n = 6 rats/group).

Blood pressure and heart rate. Recordings were made in 10 pentobarbital-anesthetized rats from 15 min before to 60 min after an injection of mystixin-7 (100 µg/kg i.v., 7 rats) or vehicle (3 rats). A polyethylene catheter was inserted into a cannulated carotid artery and connected to a solid-state BPA-190 blood pressure transducer and analyzer interfaced with a PC computer-based data acquisition DMSI 220/1 software (Micro-Med, Louisville, KY). Data for heart rate and systolic, diastolic and mean arterial pressure were collected from individual rats every 10 sec, then averaged during 1-min intervals and presented as the means of values for groups of rats.

Leakage from fixed blood vessels. Anesthetized rats pretreated with mystixin-7 (100 µg/kg i.v.) or vehicle were injected 60 min later with substance P (5 µg/kg i.v.) or vehicle (1 ml/kg). Five minutes after the substance P, the vasculature was fixed by perfusion of 1% paraformaldehyde in phosphate-buffered saline, pH 7.4, for 2 min. Immediately thereafter, Evans blue solution was perfused for 1 min at a constant flow of 200 ml/kg/min at a pressure of 120 to 140 mm Hg. The perfusate contained 0.8 mg/ml Evans blue, which corresponded to the concentration in blood after a dose of 30 mg/kg i.v., assuming a plasma volume of 3.75% of body weight (Aukland and Fadnes, 1973). The perfusate also contained 5% bovine serum albumin (type A 2153, Sigma) in phosphate-buffered saline to mimic a plasma protein concentration of 50 mg/ml (Rippe et al., 1978). The flow approximated the cardiac output (Baker et al., 1979). After Evans blue perfusion, the vasculature was perfused for 2 min with citrate-buffered 1% paraformaldehyde, and the content of Evans blue in the trachea and skin were measured as above.

Statistics. Values are presented as mean ± S.E.. Differences between means, assessed by Student's t test or analysis of variance and Scheffé's test, were considered significant when P < .05. Values for the ED₅₀ and 95% confidence limits of the inhibition of Evans blue leakage by mystixin-7 were calculated according to the log dose-probit method of Litchfield and Wilcoxon (Tallarida and Murray, 1987) with Sigmaplot software (Jandel Scientific, Inc., San Raphael, CA).

	Results

Top Abstract Introduction Methods Results Discussion References

Effect of mystixin-7 on Evans blue leakage. In the absence of substance P, the baseline leakage of Evans blue during 5 min was 10.9 ± 1.5 ng/mg in the trachea and 1.3 ± 0.2 ng/mg in skin. After substance P (5 µg/kg i.v.) the Evans blue content increased to 77.0 ± 7.1 ng/mg in the trachea and 16.1 ± 3.3 ng/mg in skin. These values were 7 and 12 times greater than the corresponding baseline values. In rats pretreated with mystixin-7 60 min before the substance P, the leakage was decreased in a dose-dependent fashion in both organs (fig. 1). The decrease was greater in the skin. The 100 µg/kg dose of mystixin-7 reduced leakage by 49% in the trachea and 86% in the skin. The ED₅₀ values (95% confidence limits) of mystixin-7 were 130 (76-211) µg/kg in the trachea and 52 (27-100) µg/kg in the skin (fig. 1).

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Log-dose-probit plot of mystixin-7-induced inhibition of Evans blue leakage in rat trachea and skin after substance P. Mystixin-7 was injected i.v. 60 min before substance P. Evans blue values reflect the amount of leakage during 5 min, expressed as percent of value for vehicle pretreatment plus substance P (mean ± S.E.; n = 6 rats/group). All values had the amount of baseline leakage (no mystixin or substance P) subtracted. ED50 (95% confidence limits): trachea, 130 (76-211) µg/kg; skin, 52 (27-100) µg/kg.

Effect of mystixin-7 and mystixin-11 on silver-stained endothelial gaps. In the absence of substance P, no extravasated Monastral blue particles and no silver deposits were present in the endothelium of postcapillary venules of the trachea (fig. 2A). By comparison, after substance P, extravasated Monastral blue was abundant in vessel walls and many silver deposits were present at endothelial cell borders (fig. 2B). Mystixin-7 pretreatment 1 hr in advance reduced the substance P-induced leakage of Monastral blue, but not the number of silver deposits (figs. 2C and 3). Similarly, mystixin-7 pretreatment had no effect on the size of the silver deposits (diameter, 1.4 ± 0.1 µm after mystixin-7 vs. 1.4 ± 0.1 µm after vehicle; n = 6 rats/group).

View larger version (123K): [in this window] [in a new window]   Fig. 2.   Endothelial cell borders made visible in postcapillary venules of rat tracheal mucosa by silver nitrate staining. (A) No pretreatment or substance P. There are no silver deposits along black lines at the endothelial cell borders. (B) Vehicle pretreatment before substance P. Silver deposits (arrows), reflecting the location of endothelial gaps, are numerous at endothelial cell borders, and patches of extravasated Monastral blue are present in the vessel wall. (C) Mystixin-7 (100 µg/kg i.v.) pretreatment 60 min before substance P. Silver deposits (arrows) are numerous, but almost no extravasated Monastral blue is present. (D) Mystixin-11 (400 µg/kg i.v.) 5 min before substance P. Abundant silver deposits (arrows) and sparse Monastral blue resemble vessel morphology after mystixin-7. The scale bar in panel D applies to all micrographs, 10 µm.

View larger version (43K): [in this window] [in a new window]   Fig. 3.   Effect of mystixin-7 (100 µg/kg i.v.) or vehicle 60 min before substance P on the number of silver deposits (arrows) at endothelial cell borders, as an index of the number of endothelial gaps. Values are expressed as silver deposits per endothelial cell; mean ± S.E.; n = 6 rats/group.

Effect of mystixin-7 on blood pressure and heart rate. The mean arterial blood pressure of pentobarbital-anesthetized rats before treatment with mystixin was 111 ± 7 mm Hg, and the mean heart rate was 349 ± 21 bpm. In control rats treated with vehicle, blood pressure gradually declined during the 75-min course of the experiment from 124 ± 4 to 104 ± 7 mm Hg and heart rate declined from 400 ± 6 to 305 ± 23 bpm (fig. 4). Within 2 min of the injection of mystixin-7 (100 µg/kg i.v.), blood pressure decreased to 69 ± 8 mm Hg, which was 62% of the initial value. Blood pressure remained at that level for about 10 min, after which it gradually rose to a plateau of 90 to 100 mm Hg. At 60 min after mystixin-7 the mean blood pressure was 97 ± 6 mm Hg, which was 87% of the initial value. Heart rate increased to 376 ± 19 bpm at 3 min after mystixin-7 and then decreased to 299 ± 7 bpm at 10 min, after which it varied between 300 and 350 bpm (fig. 4).

View larger version (27K): [in this window] [in a new window]   Fig. 4.   Effect of mystixin-7 (100 µg/kg; filled circles) or vehicle (open circles) injected i.v. (at arrows) on mean arterial blood pressure (mm Hg) and heart rate (bpm) of pentobarbital-anesthetized rats. Values are means for 1-min intervals (n = 3 vehicle-treated rats and 7 mystixin-treated rats).

Effect of mystixin-7 on leakage from fixed vessels. Baseline leakage of Evans blue from aldehyde-fixed vessels was similar that found in vivo (trachea, 11.3 ± 1.8 after fixation vs. 10.9 ± 1.5 ng/mg in vivo; skin, 2.7 ± 0.5 after fixation vs. 1.3 ± 0.2 ng/mg in vivo). When substance P was injected before fixation, the amount of leakage from fixed vessels was approximately 4 times the baseline value (fig. 5). Under these conditions, the amount of leakage in the trachea (45 ± 5.6 ng/mg) was greater than that in the skin (11 ± 1.7 ng/mg). When rats were pretreated with mystixin-7 (100 µg/kg i.v.) 60 min before the substance P, the amount of leakage was reduced 49% in the trachea and 79% in skin, indicating that mystixin-7 pretreatment decreased leakage in fixed tissues perfused at constant pressure (fig. 5).

View larger version (25K): [in this window] [in a new window]   Fig. 5.   Amount of Evans blue leakage from aldehyde-fixed vessels perfused in situ. Mystixin-7 (100 µg/kg i.v.) or vehicle were injected 60 min before substance P. The fixative was perfused for 2 min beginning 5 min after substance P and followed by Evans blue perfused for 1 min. The baseline group had no mystixin or substance P. Values are mean ± S.E.; n = 6 rats/group. *Significantly different from baseline group; significantly different from vehicle + substance P group; P < .05.

	Discussion

Top Abstract Introduction Methods Results Discussion References

This study confirmed the ability of mystixin to reduce plasma leakage in the rat trachea and skin and provided new insights as to its mechanism of action. In light of our previous studies of the antileakage action of beta-2 adrenoceptor agonists (Baluk and McDonald, 1994), it was surprising to find that the antileakage action of mystixin-7 and mystixin-11 was not caused by a reduction of endothelial gap formation, as judged from the silver deposits at endothelial cell borders after silver nitrate staining. Mystixin-7 caused a rapid fall in arterial blood pressure, but this was transient and the pressure was only slightly below normal at the time when the antileakage action was measured. Additional evidence suggesting that hypotension is not the sole explanation for the antileakage effect came from the demonstration of reduced leakage from aldehyde-fixed blood vessels after mystixin pretreatment. The antileakage effect of mystixin appeared within 5 min and lasted for at least 60 minutes. Further experiments are required to determine how permanent the effect is.

Effect of mystixin on plasma leakage. The ED₅₀ of 52 µg/kg i.v. for mystixin-7 in reducing substance P-induced leakage in skin is similar to previous estimates of the ED₅₀ of 50 µg/kg i.v. for mystixin-11 analogs in reducing heat-induced edema in the skin (Thomas et al., 1993). Substance P produced 5 times as much leakage in the trachea as in the skin, but mystixin-7 had a more potent antileakage action in the skin than in the trachea. The greater leakage in the trachea may be related to a greater blood flow, vascularity or sensitivity to substance P. The organ-related differences in efficacy of mystixin may be caused by differences in the perfusion of these tissues, number of mystixin receptors or properties of the extracellular matrix and interstitial tissue. Despite these differences, the finding that the slopes of the dose-response curves for mystixin-7 were parallel in skin and trachea is consistent with the same type of receptor being activated by mystixin in these organs.

Effect of mystixin on endothelial gaps. Previous light, scanning, and transmission electron microscopic studies of the rat trachea have shown that silver deposits correspond to sites of plasma leakage at endothelial gaps (McDonald, 1994; Hirata et al., 1995; Baluk et al., 1997). In the present study, both mystixin-7 and mystixin-11 inhibited plasma leakage, but neither peptide reduced the number or size of endothelial gaps, as judged from the number or size of silver deposits. Light microscopic silver-stained preparations made it possible to quantify a large sample of endothelial gaps, but scanning and transmission electron microscopic studies will be necessary to determine whether mystixin causes any structural changes within endothelial gaps.

Effect of mystixin on the transmural driving force. The hypotensive effect of mystixin can reduce the transmural driving force for plasma leakage. A lowered pressure gradient across the endothelium would be expected to reduce the amount of leakage, even without a decrease in endothelial permeability. For instance, the severe hypotension observed after intravenous administration of certain flavone compounds may contribute to the reduction in plasma leakage induced by thermal injury in guinea pig skin (Bohr et al., 1949). However, it is still uncertain whether the residual systemic hypotension or possible reflex vasoconstriction observed in anesthetized rats after mystixin pretreatment is sufficient to have much effect at the level of postcapillary venules. The finding that blood pressure gradually declined in control rats suggests that the prolonged anesthesia may also contribute to the apparent hypotension. The issue of the true driving force could be answered by direct measurements of intravascular pressure and endothelial permeability in individual leaky microvessels. However, such measurements are technically difficult to make in the relevant microvessels, and to our knowledge, have not been made in inflamed vessels (Ballard et al., 1992; Kendall and Michel, 1995).

Possible effect of mystixin on gap conductance. Mystixin could also reduce plasma leakage by decreasing the conductance of endothelial gaps. The luminal surface of endothelial cells is coated with a glycocalyx, which can be seen after staining with lanthanum salts and ruthenium red, or certain plant lectins, and may be present within endothelial gaps (Clough, 1991; Thurston et al., 1996). The endothelial gaps are subdivided into multiple openings by fingerlike processes (Hirata et al., 1995; Baluk et al., 1997). If mystixin increases the amount of glycocalyx or decreases the sieving through this matrix, plasma leakage through endothelial gaps could be reduced. A decrease in conductance could also occur if endothelial gaps became clogged by proteinaceous material and/or cellular blood components, such as platelets, erythrocytes or leukocytes (Arfors et al., 1979).

Pharmacological features of the antileakage effect of mystixin. Mystixin peptides potently inhibit the edema evoked by physical stimuli such as trauma, heat or freezing, or by mediators such as substance P, bradykinin, C5a or epinephrine (Thomas et al., 1993; Wei and Thomas, 1993). Furthermore, mystixin can reduce leakage from microvascular beds in the brain or pulmonary circulation which are usually insensitive to inflammatory mediators (Thomas et al., 1993; Wei and Thomas, 1993). This broad spectrum of action makes it unlikely that mystixin acts as a competitive antagonist for the receptor of any specific mediator. When evaluated at a 10 µM concentration in standard competitive in vitro ligand-receptor binding assays, mystixin-7 does not displace binding for inflammatory mediators including substance P, neurokinin A, histamine (H₁), leukotriene B₄ or D₄, phorbol ester, platelet-activating factor, thromboxane A₂ or tumor necrosis factor-, or anti-inflammatory compounds including CRF, glucocorticoids, somatostatin, neuropeptide Y, neurotensin, vasoactive intestinal peptide or various other peptides, cytokines and growth factors including atrial natriuretic factor, angiotensin II, cholecystokinin, epidermal growth factor, gastrin-releasing peptide, interleukin-1, interleukin-6, interleukin-8, nerve growth factor and platelet-derived growth factor (Kolobov et al., 1995). The absence of an effect of mystixin-7 on known CRF receptors is documented further by evidence that this peptide does not release adrenocorticotropin from the pituitary and does not activate cAMP accumulation in cells transfected with CRF₁ or CRF₂ receptors. Furthermore, the antileakage action is not antagonized by -helical CRF(9-41), a specific CRF receptor antagonist (Thomas et al., 1993). Thus, although mystixin-7 and CRF both have antileakage actions, they appear to act through different mechanisms (Thomas et al., 1993; Wei and Thomas, 1996).

Conclusions. Mystixin-7 inhibits substance P-induced plasma leakage with an ED₅₀ of less than 150 µg/kg in the rat trachea and skin without reducing the number or size of endothelial gaps. Possible contributing factors to its antileakage action include hypotension, inhibition of changes in interstitial fluid pressure or decreased gap conductance. The identification of the cellular targets of mystixin-7 may give further insight into how plasma leakage is inhibited and suggest novel mechanisms that regulate the magnitude of inflammatory responses in the microcirculation.