Modulation of barrier function of bovine aortic and pulmonary

Transcription

1 Br. J. Pharmaol. (1992), 17, '." Mamillan Press Ltd, 1992 Modulation of barrier funtion of bovine aorti and pulmonary artery endothelial ells: dissoiation from ytosoli alium ontent 'Kevin W. Buhan & 2William Martin Department of Pharmaology, University of Glasgow, Glasgow G12 8QQ Keywords: 1 Barrier funtion and ytosoli free alium ontent [a2]i was measured in monolayers of bovine pulmonary artery endothelial ells (BPAE) and bovine aorti endothelial ells (BAE). 2 ombin (1 u ml-') inreased albumin transfer aross monolayers of BPAE but not BAE, yet indued biphasi inreases in [a2]i in both endothelial ell types, onsisting of a rapid, initial phasi omponent whih deayed to a lower, more sustained plateau phase. 3 4P-Phorbol 12-myristate 13-aetate (;.3-3 nm) inreased albumin transfer aross monolayers of BPAE and BAE, but had no effet on basal levels of [a2], in either endothelial ell type. 4 Treatment of BPAE and BAE with forskolin (3 gm), an ativator of adenylate ylase, had no effet on resting transfer of albumin, but inhibited that stimulated by (6 nm). It also inhibited the thrombin (1 u ml')-indued inrease in albumin transfer aross monolayers of BPAE, but enhaned the plateau phase of the assoiated inrease in (a2],. 5 Treatment of BPAE and BAE with either atriopeptin 11 (1 nm), an ativator of partiulate guanylate ylase, or 8 bromo yli GMP (3 pm) had no effet on resting or (6 nm)-stimulated transfer of albumin. Both agents did, however, inhibit the thrombin (1 u ml-')-indued inrease in albumin transfer aross monolayers of BPAE, but had no effet on the assoiated inrease in [a2],. 6 These data suggest a dissoiation between the ability of agents that inrease or derease albumin transfer and their effets on [a2]j. onsequently, ativation of protein kinase may be the major stimulus for trans-endothelial transfer of maromoleular solutes. Endothelial barrier funtion is enhaned by elevation of either yli AMP or yli GMP ontent. yli AMP appears to at by inhibiting the ations of protein kinase, while yli GMP may at to inhibit a key step proximal to ativation of this enzyme. Endothelium; vasular permeability, ytosoli alium; fura-2; yli AMP; yli GMP; protein kinase ; phorbol esters; thrombin; atrial natriureti fator Introdution The vasular endothelium is the interfae between the blood and the interstitium and fulfils the essential funtion of regulating the exhange of fluid, solutes and ells between these two ompartments. This barrier funtion is subjet to dynami regulation, and is modulated by many fators in vivo. For example, inreased transfer is stimulated by transendothelial endoytosis (Palade, 196), or, following stimulation by inflammatory mediators suh as histamine and bradykinin at post-apillary venules, by endothelial ell ontration and onsequent formation of inter-endothelial gaps (Majno & Palade, 1961; Svensjb et al., 1979). onversely, baffler funtion an be enhaned by P-adrenoeptor agonists, a property utilised to limit vasular leakage indued by inflammatory mediators (Mariniak et al., 1978; Svensjd et al., 1979). Important new insights into the mehanisms regulating inflammatory oedema have been gained by the development of endothelial ell ulture systems. For example, histamine and thrombin have been shown to inrease maromoleular transfer aross endothelial monolayers obtained from human umbilial vein (Rotrosen & Gallin, 1986; Killakey et al., 1986) and bovine pulmonary artery (Minnear et al., 1989; Lum et al., 1989). The preise nature of the effetor pathways linking reeptor oupation to inreases in maromoleular transfer are I Present address: Glaxo Group Researh, Park Road, Ware, Herts, SG12 ODP. 2 Author for orrespondene. not, however, fully eluidated. It has been proposed that elevation of ytosoli alium is the primary trigger on the basis that histamine-indued inreases in maromoleular transfer and alium mobilisation our over a similar onentration-range (Rotrosen & Gallin, 1986). Furthermore, the alium ionophore, A23187, indues maromoleular transfer aross endothelial monolayers (Shasby et al., 1985; Gudgeon & Martin, 1989) and thrombin-indued transfer is inhibited following inhibition of alium influx by lanthanum, or buffering of intraellular alium with quin-2 (Lum et al., 1989). alium may not be the only trigger, however, sine phorbol esters are known to indue endothelial ontration (Antonov et al., 1986; Grigorian & Ryan, 1987) and maromoleular transfer (Gudgeon & Martin, 1989; Lynh et al., 199). These ations of phorbol esters are probably mediated by stimulation of protein kinase sine they are mimiked by syntheti diaylglyerols but not by inative phorbol esters and are bloked by H7, an inhibitor of protein kinase. The ability of P-adrenoeptor agonists to enhane barrier funtion in vivo has also been demonstrated in endothelial monolayers ultured from human umbilial vein, bovine pulmonary artery and bovine and porine aorta (Gudgeon & Martin, 1989; Minnear et al., 1989; Martin & Luk, 1991; Langeler & Van Hinsbergh, 1991). Enhanement of barrier funtion probably results from elevation of endothelial yli AMP ontent sine it is mimiked by other stimulants of adenylate ylase (Stelzner et al., 1989; Yamada et al., 199; Langeler & Van Hinsbergh, 1991), namely, forskolin, holera toxin and iloprost, a stable analogue of prostaylin. It is

2 ENDOTHELIAL BARRIER FUNTION 933 also enhaned by membrane permeant analogues of adenosine 3': 5'-yli monophosphate (yli AMP), and by theophylline and isobutylmethylxanthine, whih inhibit phosphodiesterase (asnoha et al., 1989; Gudgeon & Martin, 1989; Stelzner et al., 1989; Yamada et al., 199). Enhanement of barrier funtion has also been reported following treatment with membrane permeant analogues of guanosine 3':5'-yli monophosphate (yli GMP) or elevation of endothelial yli GMP ontent by atrial natriureti fators and sodium nitroprusside, whih stimulate partiulate and soluble guanylate ylase, respetively (Yamada et al., 199; Lofton et al., 1991). The aim of this study was to determine if elevation of yli AMP or yli GMP ontent inhibits agoniststimulated maromoleular transfer aross monolayers of endothelial ells ultured from bovine pulmonary artery and aorta by bloking either alium mobilisation or the stimulation of protein kinase. A preliminary aount of these findings has already been published (Buhan & Martin, 1991a). Methods Isolation of bovine aorti and pulmonary artery endothelial ells Bovine aorti endothelial ells (BAE) were isolated as desribed previously (Buhan & Martin, 1991b). Briefly, bovine thorai aortae were removed shortly after death and flushed with sterile saline ontaining benzyl peniillin (1 units ml 1) and streptomyin (plg ml-'). Following ligation of the proximal end and annulation of the distal end with an adaptor onneted to a syringe, 6 ml of the same saline solution was infused into the lumen, and the aorta was transported bak to the laboratory, where all subsequent proedures were arried out in a laminar flow hood. Following ligation of the interostal arteries, 2 ml of a sterile ollagenase solution (.1%; Type II; Sigma: in Dulbeo's modifiation of Eagle's Medium (DMEM)) was infused into the lumen and the vessel inubated for 25 min at 37 in an atmosphere of 5% O2 in air. The vessel was then gently massaged and the endothelial ells harvested by entrifugation (2 g; 4 min; 1 ) and resuspended in omplete ulture medium (DMEM ontaining 1% foetal alf serum, 1% newborn alf serum, 4 mm glutamine, 2 units ml-' benzyl peniillin and 2 pg ml-' streptomyin). Following a seond entrifugation, the ell pellet was resuspended in 5 ml of omplete ulture medium and seeded into 3 separate 8 m2 ulture flasks (Gibo). The ells were grown in an atmosphere of 5% O2 in air, and typially reahed onfluene within 4-6 days. Bovine pulmonary artery endothelial ells (BPAE) were isolated by a similar method to that outlined above for BAE. The pulmonary artery was removed, flushed with sterile saline and the proximal end and one of the two distal branhes were ligated. The other distal branh was annulated with an adaptor onneted to a syringe and 2 ml of the sterile saline solution was infused into the vessel. At the laboratory, 1O ml of sterile ollagenase solution (.1% in DMEM) was infused into the lumen and the vessel inubated and the ells harvested as indiated above for BAE. The ells were grown in ulture similarly to BAE exept that thymidine (1O pm) was added to enhane growth (Laskey et al., 199). Tissue ulture materials were obtained from Gibo (Paisley, U.K.) unless otherwise indiated. Measurement of endothelial barrier funtion Upon reahing onfluene, eah flask of BAE or BPAE was washed with 2 x 2 ml of sterile saline and inubated with 1 ml of a solution of trypsin (.5%)/ethylene diamine tetraaeti aid (EDTA;.2%) (Flow Laboratories) until the ells had detahed, usually 2-4 min. The ell suspension was then added to 2 ml of newborn alf serum to inativate the trypsin, and twie spun (2 g; 4 min; l') followed by resuspension in 5 ml of omplete medium for BAE and omplete medium ontaining thymidine (1 pm) for BPAE; 1 p1 of the ell suspensions was then added to eah of 48 Transwell membrane assemblies (ostar; 6.5 mm diameter; 3 pm pore size). These were then plaed in 24 well plates, with eah well ontaining 1 ml of omplete medium for BAE or omplete medium ontaining thymidine (1 pm) for BPAE, and inubated for a further 2-4 days. For experimentation, membrane assemblies with ells attahed were washed twie by immersion in Krebs solution ontaining (mm): Nal 118, K1 4.8, al2 1.8, MgSO4 1.2, KH2PO4 1.2, NaHO3 2.4, HEPES (N-2-hydroxyethylpiperazine-N'-2 ethanesulphoni aid) 5 at 37' at ph 7.4, and transferred to 24 well plates. Thereafter, 6 p1 of the same Krebs solution was plaed in eah of the wells whih formed the lower hamber and 1 p1 of Krebs ontaining trypan blue-labelled albumin (4%) was plaed above the endothelial monolayer. These volumes were hosen so as to avoid reation of a hydrostati gradient aross the monolayer. Drugs were then added to the top and bottom hambers and the plates plaed on an orbital shaker and inubated under an atmosphere of air at 37'. Any monolayers demonstrating visible leakage within the first 5 min were disarded and those remaining (97%) were inubated for 9 min. At the end of this time, a 1 p1 aliquot was removed from eah of the lower hambers and transfer of trypan blue-labelled albumin aross the monolayers quantified by measuring optial density at 59 nm. In the Results, the transfer of albumin is expressed as a perentage of that whih would have been ahieved at equilibrium. The trypan blue-labelled albumin omplex was prepared by a ding trypan blue (18 mg) and bovine serum albumin (4 g; ftation V; Sigma) to 1 ml of Krebs solution. Preipitation with trihloroaeti aid (6%) showed that the trypan blue was > 99.8% albumin bound. Measurement of intraellular free alium ([ait]) [a2], was measured as previously desribed (Buhan & Martin, 1991b). Briefly, monolayers of first passage BAE and BPAE grown on glass overslips were inubated for 45 min at 37' with the penta-aetoxymethyl ester form of fura-2 (2 pm) in HEPES (2 mm)-buffered DMEM (Northumbria Biologials) ontaining 1% bovine serum albumin (fration V; Sigma). A overslip ontaining fura-2 loaded ells was then transferred to HEPES (1 mm)-buffered Krebs solution, idential to that used for permeability studies exept that KH2PO4 was omitted, for 2 min at room temperature to maximize onversion to the alium-sensitive aid form of fura-2. The overslip was then suspended aross the diagonal of a quartz uvette ontaining HEPES (1 mm)-buffered Krebs solution in a Perkin Elmer LS3B fluorimeter and maintained at 37 with ontinuous stirring. The beam irradiated the ells without passing through the overslip. The exitation monohromator was omputer-driven between 34 and 38 nm every 3.8 s and fluoresene emission was olleted at 59 nm. Bakground auto-fluoresene was determined at the end of eah experiment by permeabilizing the ells to divalent ations with ionomyin (1 pm) and adding Mn2 (2 mm) to quenh intraellular fura-2 fluoresene. Following subtration of auto-fluoresene, the orreted fluoresene values obtained following exitation at 34 nm were divided by those obtained at 38 nm, giving a orreted ratio (R). [a"]j was then alulated by the omputer by the equation of Grynkiewiz et al. (1985): [a2] = IQ X (R-RI) Sf2 (Rma-R) 51,2

3 934 K.W. BUHAN & W. MARTIN The maximal (Rm,,) and minimal (R.,j.) fluoresene ratios were determined to be 16.3 and.8, respetively. Sf and Sb2 are the fluoresene values obtained following exitation at 38 nm in the absene of alium and the presene of saturating levels of alium, respetively, and the ratio of these two values was alulated to be 7.3. The Kd for the fura-2-alium omplex was assumed to be 225 nm at 37T. Drugs Atriopeptin II, 8 bromo yli 3':5' guanosine monophosphate (8 bromo yli GMP), 4m-phorbol 12,13-dideanoate (4ax-PDD), 4p-phorbol 12-myristate 13-aetate () and thrombin (bovine) were obtained from Sigma, Poole, Dorset. Forskolin, fura-2 penta-aetoxymethyl ester and ionomyin were obtained from Novabiohem, ambridge, U.K. Solutions of drugs were made in distilled water exept for forskolin, fura-2-aetoxymethyl ester and ionomyin whih were dissolved in dimethylsulphoxide and and 4a-PDD whih were dissolved in 1% ethanol. Statistial analysis Results are expressed as the mean ± s.e.mean and omparisons were made by Student's t test or the Mann-Whitney test when there was unequal variane in samples. A probability of.5 or less was onsidered signifiant. Results Effets of thrombin and a phorbol ester on endothelial barrier funtion Resting transfer of trypan blue-labelled albumin aross monolayers of BPAE and BAE in the 9 min inubation period was typially 2-1% (Figures 1-5). Treatment with thrombin (1 u ml-') during the 9 min period inreased albumin transfer aross monolayers of BPAE, but not BAE, whereas 4p-phorbol 12-myristate 13-aetate (; 6 nm) inreased albumin transfer aross monolayers of both endothelial ell types (Figure 1). The ability of to a a 3 - b 25, For For m 2-4* o b r Fr 3 w LUI For For Figure 1 The transfer of trypan blue-labelled albumin aross monolayers of (a) bovine aorti endothelial ells (BAE) and (b) bovine pulmonary artery endothelial ells (BPAE) during a 9 min inubation period in the absene of drugs () and following stimulation with thrombin (; 1 u ml-') or 4P-phorbol 12-myristate 13- aetate (; 6 nm). Values given are means and vertial bars indiate the s.e.mean of 6 observations. P<.1 indiates a signifiant differene from ontrol (). For For Figure 2 The effets of forskolin (For; 3 jum) on resting transfer of trypan blue-labelled albumin and on transfer stimulated by 4pphorbol 12-myristate 13-aetate (; 6nM) aross monolayers of (a) bovine aorti endothelial ells (BAE) and (b and ) bovine pulmonary artery endothelial ells (BPAE) during a 9 min inubation. The effet of forskolin on albumin transfer stimulated by thrombin (; I u ml-') aross monolayers of bovine pulmonary artery endothelial ells is also shown. Values given are means and vertial bars indiate the s.e.mean of 6-12 observations. P<O.1; *P<.1 indiates a signifiant differene from ontrol () or between groups joined by a braket.

4 ENDOTHELIAL BARRIER FUNTION 935 inrease albumin transfer was not shared with the inative phorbol ester, 4x-phorbol 12,13-dideanoate (6 nm; data not shown). Effets of yli nuleotides on endothelial barrier funtion Forskolin (3pM), whih ativates the atalyti subunit of adenylate ylase (Seaman & Daly, 1981), had no effet on resting transfer of albumin aross monolayers of BPAE or BAE, inhibited the inrease in albumin transfer stimulated by thrombin (1 u ml-') and (6 nm) aross BPAE, and inhibited the inrease stimulated by (6 nm) aross BAE (Figure 2). a Neither atriopeptin II (1 nm), an ativator of partiulate guanylate ylase, nor 8 bromo yli GMP (3 gim), a membrane permeant analogue of yli GMP, had any effet on resting or (6 nm)-stimulated transfer of albumin aross monolayers of BAE (Figure 3) or BPAE (Figure 4), but both inhibited that stimulated by thrombin (1 u ml-') aross monolayers of BPAE (Figures 3 and 4). alium mobilization in endothelial ells In monolayers of BPAE and BAE the basal level of [a2]i was 16 ± 4 nm (n = 11) and 98 ± 4 nm (n = 127), respetively. ombin (1 u ml-') indued a biphasi elevaa * -., - b 2 - APII APII.E 4- - b BrGMP 8BrGMP. 4-o ẉ _ m w - Qi 1 5 Ul - 2 APII T APII BrGMP 8BrGMP A v-- APII APII Figure 3 The effets of atriopeptin II (APII; 1nM) on resting transfer of trypan blue-labelled albumin and on transfer stimulated by 4P-phorbol 12-myristate 13-aetate (; 6 nm) aross monolayers of (a) bovine aorti endothelial ells (BAE) and (b and ) bovine pulmonary artery endothelial ells (BPAE) during a 9 min inubation. The effet of atriopeptin II on albumin transfer stimulated by thrombin (; 1 u ml 1) aross monolayers of bovine pulmonary artery endothelial ells is also shown. Values given are means and vertial bars indiate the s.e.mean of 6 observations. P<.1; *P<.1 indiates a signifiant differene from ontrol () or between groups joined by a braket. 8BrGMP 8BrGMP Figure 4 Effets of 8 bromo yli GMP (8BrGMP; 3fLM) on resting transfer of trypan blue-labelled albumin and on transfer stimulated by 4,-phorbol 12-myristate 13-aetate (; 6 nm) aross monolayers of (a) bovine aorti endothelial ells (BAE) and (b and ) bovine pulmonary artery endothelial ells (BPAE) during a 9 min inubation period. The effet of 8 bromo yli GMP on albumin transfer stimulated by thrombin (; I u ml ') aross monolayers of bovine pulmonary artery endothelial ells is also shown. Values given are means and vertial bars indiate the s.e.mean of 6 observations. *P<.5; P<.1 indiates a signifiant differene from ontrol () or between groups joined by a braket.

5 936 K.W. BUHAN & W. MARTIN tion of [a2]i in both ell types onsisting of a large initial peak at around 3s whih then fell to a more sustained plateau within 5 min (Figure 5); the peak and plateau levels were 35 ± 51 nm and 239 ± 21 nm (n = 9), respetively, for BPAE, and 291 ± 3 nm and 18 ± 13 nm (n = 2), respetively, for BAE. In ontrast, (1-1 nm) had no effet on the basal level of [a2]i in BPAE or BAE. Effets of yli nuleotides on thrombin-indued alium mobilisation in BPAE Pretreatment of BPAE with forskolin (3pM; 5 min) had no effet on basal levels of [a2]i or on the magnitude of the initial transient elevation of [a2]i indued by thrombin (1 u ml-') (data not shown). Addition of forskolin (3 gm) during the plateau phase of the inrease in [a2]i indued by thrombin (1 u ml -) did, however, lead to a further rapid inrease in [a2]i of 67 ± 7 nm (n = 6), whih remained stable for at least 5 min (Figure 5). -i N (a U. 3 1 BAE BPAEG BPAE 3 1 BPAE 1 uml- 2-4 nof% - 1 um-', For uml- 13 FM f- I Apil 1 nm min 8BrGMP 3 1umV-1 p.m Figure 5 Individual traes illustrating the effets of thrombin (; 1 u ml-') on [a2ji in bovine aorti endothelial ells (BAE) and bovine pulmonary artery endothelial ells (BPAE). The effets of adding forskolin (For; 3 gm), atriopeptin II (APII; 1 nm) or 8 bromo yli GMP (8BrGMP; 3pM) during the plateau phase of the inrease in [a2]i indued by thrombin (; u ml-') in bovine pulmonary artery endothelial ells are also shown. Eah trae is representative of at least 5 separate observations. Pretreatment of BPAE for 5 min with either atriopeptin 11 (1 nm) or 8 bromo yli GMP (3 gam) had no effet on basal levels of [a2]i or on the magnitude of the initial transient elevation of [a2]j indued by thrombin (1 u ml-') (data not shown). Atriopeptin II (1 nm) and 8 bromo yli GMP (3 JAM) were both also without effet when added during the plateau phase of the inrease in [a2] indued by thrombin (I u ml-') (Figure 5). Disussion The major new finding in this study is that hanges in endothelial barrier funtion an be dissoiated from hanges in levels of ytosoli alium ([a2]i). Evidene for this omes from the observation that thrombin stimulates alium mobilization in both BPAE and BAE, yet inreases albumin transfer aross monolayers of only BPAE. Furthermore, the phorbol ester,, had no effet on basal levels of [a2]i in BPAE or BAE, yet stimulated albumin transfer aross monolayers of both endothelial ell types. It is likely that the ability of phorbol esters to stimulate transendothelial transfer of maromoleular solutes results from ativation of protein kinase, sine this ation is not shared with phorbol esters whih do not ativate this enzyme, but is mimiked by syntheti diaylglyerdls, and bloked by H7 (Gudgeon & Martin, 1989; Lynh et al., 199). It is possible, therefore, that ativation of protein kinase represents the major pathway by whih inflammatory mediators indue plasma leakage. At present, it is not lear how ativation of protein kinase inhibits endothelial barrier funtion, but it is likely to result from endothelial ontration (Antonov et al., 1986; Grigorian & Ryan, 1987) and formation of inter-endothelial gaps. In vasular smooth musle, ativation of protein kinase indues ontration by inreasing the sensitivity of the ontratile proteins to alium (Itoh et al., 1988), and it is possible that a similar mehanism operates in the endothelium. Alternatively, ativation of protein kinase may be responsible for the loss of peripheral bands of F-atin and the resultant disruption of ell-ell ontats (Garia et al., 1986; Minnear et al., 1989) in a manner similar to that desribed for a kidney epithelial ell line (Shliva et al, 1984). Our finding that forskolin, whih diretly ativates the atalyti subunit of adenylate ylase (Seaman & Daly, 1981), inhibits inreases in albumin transfer stimulated by aross monolayers of BPAE and BAE as well as that stimulated by thrombin aross BPAE is onsistent with previous reports of elevated levels of yli AMP enhaning endothelial barrier funtion in vivo and in vitro (Mariniak et al., 1978; Svensjd et al., 1979; Killakey et al., 1986; Minnear et al., 1989; Gudgeon & Martin, 1989; arson et al., 1989; Langeler & Van Hinsbergh, 1991). Furthermore, our observation that atriopeptin II, a stimulant of endothelial partiulate guanylate ylase (Shini et al., 1988; Martin et al., 1988), and 8 bromo yli GMP inhibit thrombin-stimulated transfer of albumin aross monolayers of BPAE is also onsistent with the ability of yli GMP to enhane barrier funtion (Yamada et al., 199; Lofton et al., 1991). The mehanisms by whih elevations of yli AMP or yli GMP enhane endothelial barrier funtion are not lear, but are unlikely to result from inhibition of alium mobilization, sine atriopeptin II and 8 bromo yli GMP had no effet on thrombin-indued alium mobilization and forskolin atually augmented this. This proposal is supported by the observation that elevation of yli AMP ontent inhibits histamine-indued transfer of albumin aross monolayers of human umbilial vein endothelial ells but does not blok the assoiated inrease in [a2j]i (arson et al., 1989). It is possible, however, that inhibition of barrier funtion is exerted through blokade of protein kinase, sine elevation of yli AMP ontent inhibits albumin transfer stimulated

6 ENDOTHELIAL BARRIER FUNTION 937 by both thrombin and. In ontrast, elevation of yli GMP ontent inhibits albumin transfer stimulated by thrombin, but not, suggesting a different mehanism of ation from yli AMP. One possible explanation for this is that yli GMP may blok the ability of thrombin to stimulate protein kinase, and onsistent with this is the ability of yli GMP to inhibit prodution of inositol (1,4,5) trisphosphate in porine aorti endothelium (Lang & Lewis, 1991). We do not favour this explanation, however, sine 8 bromo yli GMP and atriopeptin II had no effet on thrombin-indued mobilization of alium in BPAE, whih presumably involves hydrolysis of phosphatidylinositol-4,5- bisphosphate (Jaffe et al., 1987). It is possible, however, that albumin transfer is stimulated by ativation of protein kinase resulting from hydrolysis of phosphatidylholine and not phosphatidylinositol4,5-bisphosphate. This is suggested sine diaylglyerol prodution from the former soure is better sustained (Billah & Anthes, 199) and would be more onsistent with the relatively long time ourse (9 min) required to observe albumin transfer. If elevations of [a2"ji and sustained prodution of diaylglyerol are subjet to differential regulation in the endothelial ell, as in the neutrophil (ronstein et al., 1988; ronstein & Haines, 1992), then it might be possible to explain our ability to blok the inrease in albumin transfer but not [a2"]j stimulated by thrombin in BPAE. Diret assessment of the differential effets of elevated levels of yli GMP on hydrolysis of phosphatidylinositol-4,5-bisphosphate and phosphatidylholine will be required to test this hypothesis. A related problem that is also diffiult to explain at present is the ability of thrombin to stimulate albumin transfer aross monolayers of BPAE but not BAE. On the basis of the above sheme, it is possible that following stimulation with thrombin, only BPAE generates diaylglyerol from phosphatidylholine in the sustained manner neessary to stimulate albumin transfer. Alternatively, the two ell types ould generate diaylglyerols, or ontain different forms of protein kinase (Thompson et al., 1991). In onlusion, the results of this study show that hanges in endothelial barrier funtion an be dissoiated from alterations in ytosoli alium ontent and suggest protein kinase as the primary regulator. We are grateful for support from the British Heart Foundation and the Royal Soiety. Referenes ANTONOV, A.S., LUKASHEV, M.E., ROMANOV, Y.A., TKAHUK, V.A., REPIN, U.S. & SMIRNOV, U.N. (1986). Morphologial alterations in endothelial ells from human aorta and umbilial vein indued by forskolin and phorbol 12-myristate 13-aetate: A synergisti ation of adenylate ylase and protein kinase ativators. Pro. Natl. Aad. Si. U.S.A., 83, BILLAH, M.M. & ANTHES, A.. (199). The regulation and ellular funtions of phosphatidylholine hydrolysis. Biohem. J., 269, BUHAN, K.W. & MARTIN, W. (1991a). Inhibition of thrombinindued inreases in endothelial permeability by forskolin and atriopeptin II. Br. J. Pharmaol., 14, 84P. BUHAN, K.W. & MARTIN, W. (1991b). Bradykinin indues elevations of ytosoli alium through mobilisation of intraellular and extraellular pools in bovine aorti endothelial ells. Br. J. Pharmaol., 12, ARSON, M.R., SHASBY, S.S. & SHASBY, D.M. (1989). Histamine and inositol phosphate aumulation in endothelium: AMP and a G protein. Am. J. Physiol., 257, L259-L264. ASNOHA, S.A., ESKIN, S.G., HALL, E.R. & MINTYRE, L.V. (1989). Permeability of human endothelial monolayers: effet of vasoative agonists and AMP. J. Appl. Physiol., 67, RONSTEIN, B.N. & HAINES, K.A. (1992). Stimulus-response unoupling in the neutrophil. Biohem. J., 281, RONSTEIN, B.N., KRAMER, S.B., ROSENSTEIN, E.D., KORHAK, H.M., WEISSMANN, G. & HIRSHHORN, R. (1988). Oupany of adenosine reeptors raises yli AMP alone and in synergy with oupany of hemoattratant reeptors and inhibits membrane depolarisation. Biohem. J., 252, GARIA, J.G.N., SIFLINGER-BIRNBOIM, A., BIZIOS, R., DEL- VEHIO, P.J., FENTON, J.W. & MALIK, A.B. (1986). ombinindued inrease in albumin permeability aross the endothelium. J. ell. Physiol., 128, GRIGORIAN, G.Y. & RYAN, U.S. (1987). Platelet-ativating fator effets on bovine pulmonary artery endothelial ells. ir. Res., 61, GRYNKIEWIZ, G., POENIE, M. & TSIEN, R.Y. (1985). A new generation of a2 indiators with greatly improved fluoresene properties. J. Biol. hem., 26, GUDGEON, J.R. & MARTIN, W. (1989). Modulation of arterial endothelial permeability: studies on an in vitro model. Br. J. Pharmaol., 98, ITOH, T., KUBOTA, Y. & KURIYAMA, H. (1988). Effets of a phorbol ester on aetylholine-indued a2 mobilisation and ontration in the porine oronary artery. J. Physiol., 397, JAFFE, E.A., GRULIH, J., WEKSLER, B.B., HAMPEL, G. & WAT- ANABE, K. (1987). orrelation between thrombin-indued prostaylin prodution and inositol triphosphate and ytosoli free alium in ultured human endothelial ells. J. Biol. hem., 262, KILLAKEY, J.J.F., JOHNSTONE, M.G. & MOVAT, H.Z. (1986). Inreased permeability of miroarrier ultured endothelial monolayers in response to histamine and thrombin. A model for the in vitro study of inreased vasopermeability. Am. J. Pathol., 122, LANG, D. & LEWIS, M.J. (1991). Inhibition of inositol 1,4,5,- triphosphate formation by yli GMP in ultured aorti endothelial ells of the pig. Br. J. Pharmaol., 12, LANGELER, E.G. & VAN HINSBERGH, V.W.M. (1991). Norepinephrine and iloprost improve barrier funtion of human endothelial ell monolayers: role of AMP. Am. J. Physiol., 26, LASKEY, R.E., ADAMS, D.J., JOHNS, A., RUBANYI, G.M. & VAN BREEMEN,. (199). Membrane potential and Na -K pump ativity modulate resting and bradykinin-stimulated hanges in ytosoli free alium in ultured endothelial ells from bovine atria. J. Biol. hem., 265, LOFTON,.E., BARON, D.A., HEFFNER, J.E., URRIE, M.G. & NEWMAN, W.H. (1991). Atrial natriureti peptide inhibits oxidant-indued inreases in endothelial permeability. J. Mol. ell. ardiol., 23, LUM, H., DEL-VEHIO, P.J., SHNEIDER, A.S., GOLIGORSKY, M.S. & MALIK, A.B. (1989). alium-dependene of the thrombinindued inrease in endothelial albumin permeability. J. Appl. Physiol., 66, LYNH, J.J., FERRO, T.J., BLUMENSTOK, F.A., BROKENAUER, A.M. & MALIK, A.B. (199). Inreased endothelial albumin permeability mediated by protein kinase ativation. J. lin. Invest., 85, MAJNO, G. & PALADE, M.D. (1961). Studies on inflammation: I. The effet of histamine and serotonin on vasular permeability: An eletron-mirosopi study. J. Biophys. Biohem. ytol., 11, MARINIAK, D.L., DOBBINS, D.E., MAIEJKO, J.J., SOTT, J.B., HADDY, F.J. & GREGA, G.J. (1978). Antagonism of histamineindued edema formation by ateholamines. Am. J. Physiol., 234, H18-H185. MARTIN, W. & LUK, J.F. (1991). Enhanement of arterial endothelial barrier funtion by P-adrenoeptor agonists. Br. J Pharmaol., 12, 13. MARTIN, W., WHITE, D.G. & HENDERSON, A.H. (1988). Endothelium-derived relaxing fator and atriopeptin II elevate yli GMP levels in pig aorti endothelial ells. Br. J. Pharmaol., 93, MINNEAR, F.L., DE MIHELE, M.A., MOON, D.G., RIEDER,.L. & FENTON, J.W.II. (1989). Isoproterenol redues thrombin-indued pulmonary endothelial permeability in vitro. Am. J. Physiol., 257, H1613-H1623. PALADE, G.E. (196). Transport in quanta aross the endothelium of blood apillaries. Anat. Reord, 136,

7 938 K.W. BUHAN & W. MARTIN ROTROSEN, D. & GALLIN, J.I. (1986). Histamine type I oupany inreases endothelial ytosoli alium, redues F-atin, and promotes albumin diffusion aross ultured endothelial monolayers. J. ell Biol., 13, SHINI, V., GRANT, N.J., MILLER, R.. & TAKEDA, K. (1988). Morphologial haraterisation of ultured bovine aorti endothelial ells and the effets of atriopeptin II and sodium nitroprusside on ellular and extraellular aumulation of yli GMP. Eur. J. ell Biol., 47, SHLIVA, M., NAKAMURA, T., PORTER, K.R. & EUTENEUER, U. (1984). A tumor promoter indues rapid and oordinated reorganisation of atin and vinulin in ultured ells. J. ell Biol., 99, SEAMAN, K. & DALY, J.W. (1981). Ativation of adenylate ylase by the diterpene forskolin does not require the guanine nuleotide regulatory protein. J. Biol. hem., 256, SHASBY, D.M., LIND, S.E., SHASBY, S.S., GOLDSMITH, J.. & HUNN- INGHAKE, G.W. (1985). Reversible oxidant-indued inreases in albumin transfer aross ultured endothelium: Alterations in ell shape and alium homeostasis. Blood, 65, STELZNER, T.J., WEIL, J.V. & O'BRIEN, R.F. (1989). Role of yli adenosine monophosphate in the indution of endothelial barrier properties. J. ell. Physiol., 139, SVENSJO, E., ARFORS, K.-E., RAYMOND, R.M. & GREGA, G.J. (1979). Morphologial and physiologial orrelation of bradykinin-indued maromoleular efflux. Am. J. Physiol., 236, H6-H66. THOMPSON, N.T., BONSER, R.W. & GARLAND, L.G. (1991). Reeptor-oupled phospholipase D and its inhibition. Trends Pharmaol. Si., 12, YAMADA, Y., FURUMIHI, T., FURUI, H., YOKOI, T., ITO, T., YAMAUHI, K., YOKOTA, M., HAYASHI, H. & SAITO, H. (199). Roles of alium, yli nuleotides, and protein kinase in regulation of endothelial permeability. Arterioslerosis, 1, (Reeived April 24, 1992 Revised June 26, 1992 Aepted July 27, 1992)