HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kajimura, M. Articles by Curry, F. E. Search for Related Content PubMed PubMed Citation Articles by Kajimura, M. Articles by Curry, F. E.

Endothelial cell shrinkage increases permeability through a Ca²⁺-dependent pathway in single frog mesenteric microvessels

M. Kajimura and F. E. Curry

MS 9025 Received 1 December 1998; accepted after revision 24 March 1999.

	ABSTRACT

Top Abstract Introduction Methods Results Discussion References

1. We tested whether calcium (Ca²⁺)-dependent mechanisms were essential for our previous observation that a change in the endothelial cell (EC)-extracellular matrix (ECM) attachment caused an increase in microvessel hydraulic permeability (L_p) after exposure to hypertonic solutions in single perfused mesenteric microvessels in pithed frogs (Rana pipiens).

2. In microvessels where integrin-dependent EC-ECM attachments were disrupted by pretreatment with the peptide Gly-Arg-Gly-Asp-Thr-Pro (GRGDTP; 0·3 mmol l^-1), we measured microvessel L_p after exposure to hypertonic solutions under experimental conditions that reduced Ca²⁺ influx into endothelial cells.

3. High K⁺ solutions (59·7 and 100 mmol l^-1 K⁺) were used to depolarize the endothelial membrane and therefore to reduce the electrochemical driving force for Ca²⁺ influx through conductive Ca²⁺ channels. These solutions abolished the increase in L_p caused by hypertonic solutions in the microvessels pretreated with GRGDTP.

4. We previously suggested that the removal of albumin from the perfusate may reduce EC-ECM attachment because hypertonic solutions increased the L_p of microvessels above that due to removal of albumin alone. This additional increase in L_p was attenuated by the 59·7 mmol l^-1 K⁺ solution and was completely abolished by the 100 mmol l^-1 K⁺ solution.

5. Bumetanide, an inhibitor of the Na⁺-K⁺-2Cl^- co-transporter and one of the mechanisms of regulatory volume increase after exposure to hypertonic solutions in endothelial cells, did not change the response of microvessels to high K⁺ solutions.

6. Our findings indicate that Ca²⁺ entry into endothelial cells via passive conductance channels is necessary to increase microvessel L_p after exposure to hypertonic solutions in microvessels where EC-ECM attachments are disrupted.

	INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

The aim of this study was to investigate further the hypothesis that a change in the endothelial cell (EC)- extracellular matrix (ECM) attachment may modify the increase in capillary permeability caused by endothelial cell shrinkage. In intact microvessels of the frog mesentery, we have shown that endothelial cell shrinkage after exposure to a hypertonic solution (330 versus 230 mosmol l^-1 in frog Ringer solution) did not increase hydraulic permeability (L_p) when vessels were perfused with a normal Ringer solution containing bovine serum albumin (BSA). However, when vessels were pretreated with a hexapeptide containing the Arg-Gly-Asp (RGD) sequence, which competes for the binding sites between endothelial cell integrins and the basement membrane, and reduces the number of functional attachment sites between the endothelial cell and the subcellular matrix, exposure to hypertonic solutions caused a biphasic increase in L_p (Kajimura et al. 1997). The simplest interpretation of these experiments was that a reduction in endothelial cell volume resulted in a change in the geometry of the principal water pathway across the capillary to increase permeability only when the attachment of the endothelial cell to the basement membrane was compromised.

The mechanisms whereby EC-ECM attachment modulates microvessel permeability are not well understood. One possibility is that, in the presence of decreased attachment of the endothelial cell to the basement membrane, a reduction in the volume of endothelial cells directly modifies the resistance to water flow through the junction between cells by widening the junction or reducing the length of the junction overlap. A second possibility is that changes in endothelial cell attachment modify the way the endothelial cells respond to applied forces (in this case, a change in cell volume). We expected that the second possibility may involve Ca²⁺ dependent signalling pathways because these modulate responses of endothelial cells to shear force (Davies, 1995), and the attachment and detachment of cells to basement membranes (e.g. neutrophil detachment from a fibronectin or vitronectin substratum during migration, Hendey & Maxfield, 1993). The principal purpose of this paper was to distinguish between a direct effect at the EC-ECM attachment site and an effect mediated by intracellular Ca²⁺. We repeated experiments similar to those where vessels were exposed to hypertonic solutions after being treated with the RGD peptide, but this time the experiments were conducted in solutions in which the K⁺ concentration was high (high K⁺ solution). High K⁺ solutions are known to depolarize endothelial cell membrane and therefore to decrease the driving force for Ca²⁺ influx through passive conductance pathways (He & Curry, 1991). We argued that if the principal action of the RGD peptide to modify microvessel permeability involves compromising the integrin binding sites as a simple extracellular effect, high K⁺ solutions should not modify the permeability response. On the other hand if the action is dependent on Ca²⁺-dependent processes, the L_p increase is predicted to be attenuated by high K⁺ solutions.

We have also shown that removal of BSA from the perfusate enhanced the extent of an increase in microvessel permeability caused by cell shrinkage (Kajimura et al. 1997). Based on this observation, we would also suggest that the effect of the BSA removal may involve a modification of cell attachment. This could also be either a direct effect due to extracellular albumin removal or an effect mediated by intracellular Ca²⁺. As above, we argued that if the effects are direct, then they should not be significantly modified by high K⁺ solutions. Preliminary results of this work have been presented in abstract form (Kajimura et al. 1994; Kajimura & Curry, 1995).

	METHODS

Top Abstract Introduction Methods Results Discussion References

General methods

The preparation of both the frog mesentery and the micropipettes used to perfuse individual microvessels have been described in detail elsewhere (Curry et al. 1983). Briefly, the mesentery of a brain-pithed, male leopard frog (Rana pipiens, 6-7·5 cm in length supplied by J. M. Hazen, Alburg, VT, USA) was gently arranged on the surface of a transparent pillar (diameter 1 cm). This allowed trans-illumination of the mesenteric microvasculature. The upper surface of the mesentery was superfused continuously with frog Ringer solution at 16-18°C. The microvessels chosen for study were mostly venous capillaries (diameters 10-35 µm) though some were true capillaries. The tissue was observed with an inverted microscope (Leitz Diavert) using a × 6 objective lens (NA, 0·18). A video camera (Pulinix TM-7) was attached to the top of the microscope and the output from the camera was displayed on video monitors and recorded.

Animals were handled in accordance with the protocols approved by the Animal Care and Use Committee of the University of California, Davis.

Measurement of hydraulic permeability (L_p)

The modified Landis micro-occlusion technique was used for the measurement of L_p (Michel et al. 1974). The fluid flux across a segment of a single perfused microvessel was measured from the motion of a flow marker immediately after occlusion of the vessel. Transcapillary fluid flux per unit area (J_v/S) was calculated as:

where dl/dt is the initial velocity of the flow marker after the microvessel is occluded, r is the vessel radius, and l is the length between the flow marker and the point of occlusion. J_v/S was plotted against hydrostatic pressure. The regression of J_v/S with pressure yields L_p as the slope. In this study, however, J_v/S was measured at a single pressure (30 cmH₂O) to detect the time course of the L_p change. L_p was calculated as:

L_p = (J_v/S)/(P_c - _c),

where P_c is the capillary pressure, and _c is the effective oncotic pressure of the perfusate (0·36 cmH₂O was used as _c for the perfusate containing 1 mg ml^-1 BSA). The interstitial pressure was negligible and the interstitial oncotic pressure was assumed to be the same as that of the superfusate.

Solutions

Frog Ringer solution (NK⁺; 2·4 mmol l^-1 K⁺) solution was used as the bathing solution for the dissection of the mesentery and for the initial superfusates. The composition of isotonic NK⁺ solution was (mmol l^-1): 111 NaCl, 2·4 KCl, 1·0 MgCl₂, 1·1 CaCl₂, 0·195 NaHCO₃, 5·5 glucose, and 5 Hepes acid and Na-Hepes. The pH was adjusted to 7·4 by the ratio of Hepes acid to base. The osmolarity of the isotonic Ringer solution was 230 mosmol l^-1 as determined by freezing point depression (Advanced Wide-Range Osmometer 3W2, Advanced Instruments, Inc., Norwood, MA, USA). The hypertonic Ringer solution was prepared by increasing NaCl concentration from 111 to 166 mmol l^-1 to raise the osmolarity to 330 mosmol l^-1. The perfusate contained bovine serum albumin (BSA; A-4378, Fraction V, Sigma) at 1 mg ml^-1 unless otherwise stated. This concentration (1 mg ml^-1) of BSA was chosen because: (i) it is found to maintain the normal property of the endothelial barrier of single perfused microvessels in frog mesentery (Huxley & Curry, 1985); (ii) we previously found that a 100 mosmol l^-1 increase in the osmolarity of both perfusate and superfusate does not increase L_p when the perfusate contains BSA at this concentration (Kajimura et al. 1997); and (iii) it was important to keep a relatively high RGD peptide /BSA ratio. To eliminate osmotic gradients across the capillary wall, microvessels were perfused with a solution prepared by dialysing perfusate in 8000 MW cut-off dialysis tubing (Spectro/Por; Spectrum, CA, USA) against three 2-litre changes of Ringer solution of equal osmolarity over 24 h at 5°C. Although 1 mg ml^-1 BSA will exert an osmotic pressure of 0·4 cmH₂O across the dialysis membrane, tending to draw water into the dialysis bag and dilute the protein, the actual changes in protein composition of the perfusate and the net filtration force for water flow across the capillary wall due to the use of a dialysed protein in the perfusate was estimated to be negligible.

High K⁺ solutions (57·9 and 100 mmol l^-1 K⁺, designated as 57·9 and 100 mM K⁺, respectively) were prepared by equimolar replacement of NaCl with KCl. The calcium (Ca²⁺) free solution was prepared as standard frog Ringer solution except for omission of Ca²⁺. EGTA (25 µmol l^-1) was added to the superfusates to chelate Ca²⁺ from the water source and EGTA (50 µmol l^-1) was added to Ringer-albumin perfusates ([BSA] = 1 mg ml^-1) to chelate Ca²⁺ bound to BSA. The concentrations of EGTA were based on the study by He & Curry (1993) in this laboratory where Ca²⁺ was titrated to less than 1 µmol l^-1 by EGTA. The final concentrations of EGTA were 10-30 µmol l^-1 in Ringer solution and 50-130 µmol l^-1 in Ringer-albumin solution ([BSA] = 10 mg ml^-1) in their study (He & Curry 1993).

The peptide GRGDTP (Peninsula Laboratories, San Diego, CA, USA) was made as a stock solution (3 mmol l^-1) in Ringer solution and stored at -70°C until use. The stock solution was stored for no more than one week. On the day of the experiment the one-tenth dilution was made into BSA-Ringer solution to bring the final concentration of GRGDTP to 0·3 mmol l^-1. Bumetanide (ICN Biochemicals, Inc., Aurora, OH, USA) was made as a stock solution (50 mmol l^-1) in 95 % ethanol. On the day of the experiment the first dilution, one-tenth dilution, was made into BSA-free Ringer solution. Then the second dilution was made using BSA-Ringer solution to bring the final concentration of bumetanide to 50 µmol l^-1. This concentration was well above the concentration of 10 µmol l^-1 which effectively blocked the regulatory volume increase caused by the cell shrinkage in cultured bovine aortic endothelial cells (O'Donnell, 1993).

Depolarization by high K⁺ solutions

In order to reduce Ca²⁺ influx we used high K⁺ solutions which depolarize the endothelial cell membrane. This reduced the driving force for the Ca²⁺ entry to the cell. The two groups of high K⁺ solutions were 57·9 and 100 mM K⁺. These high K⁺ solutions depolarized the frog endothelial cell membranes from a mean value of -52, to -27 mV and -12 mV, respectively, (He & Curry, 1995). In hypertonic solutions using NaCl as the osmotic agent, there is an additional change in membrane potential due to different ionic composition of the solution, but this is expected to be small, since the endothelial cell membrane seems to act as if the conductance to K⁺ is close to 8·5 times that of Na⁺ when the membrane potential is in the range -50 to 0 mV (He & Curry, 1995).

Experimental protocol

RGD pretreated vessel. Microvessels chosen for the experiment had brisk blood flow and no white cells sticking or rolling along the wall. Each vessel was initially perfused with a control solution (isotonic, 230 mosmol l^-1) which contained bovine serum albumin (BSA, 1 mg ml^-1). Usually occlusions were made every 30 s over approximately 5 min to establish a baseline L_p (designated as L_p0). The pipette was removed and re-cannulated with a new pipette filled with a solution containing GRGDTP (0·3 mmol l^-1) and BSA (1 mg ml^-1). The vessel was perfused with the solution for 10 min; this is referred to as the 'RGD pretreatment'. During the pretreatment, multiple occlusions were made to estimate L_p. We then replaced the superfusate with a hypertonic solution (330 mosmol l^-1) and re-cannulated the vessel with the hypertonic solution that contained GRGDTP (0·3 mmol l^-1) and BSA (1 mg ml^-1). To follow changes in L_p after exposing the vessel to hypertonic solutions, both the superfusate line and the perfusion pipette containing the hypertonic solution were carefully set in place before changing solution composition. With this arrangement there was usually a time lapse of approximately 10-20 s between the time when the mesentery was exposed to a hypertonic solution (i.e. the time of the superfusate switch to a hypertonic solution) and when the lumen of the vessel was exposed to a hypertonic perfusate (i.e. the time of the re-cannulation of the vessel with a hypertonic perfusate). Then the vessel was occluded as soon as possible after the re-cannulation to detect any rapid initial change in L_p. To test the effect of depolarization on the increase in L_p due to exposure to hypertonic solutions, the above protocol was carried out with either 57·9 or 100 mM K⁺ solution and these data were compared with our previous experiments with the NK⁺ solution (Kajimura et al. 1997). In an additional 10 vessels, the experiments were carried out as above under both normal and depolarizing conditions.

BSA-free Ringer perfused vessel. To determine whether conditions that reduce Ca²⁺ influx can inhibit the L_p increase caused by a hypertonic solution in the vessels perfused with BSA-free Ringer solution, the experiments were carried out under depolarizing conditions as follows: (i) baseline L_p was measured with the isotonic solution of high K⁺ concentration; (ii) changes in L_p were determined after removing BSA from the perfusate; and (iii) both superfusate and perfusate were replaced with a hypertonic solution and the L_p response measured for approximately 5 min. The experiments were carried out as above with either 57·9 or 100 mM K⁺ solution. In three vessels measurements were made in both NK⁺ and 100 mM K⁺ solutions.

Statistical analysis

Average values are reported as means ± S.E.M. throughout, unless specified. The significance of differences between means was calculated using either the Wilcoxon signed rank test (paired) or the Mann-Whitney U test (unpaired). Tests of significance were set at the 5 % level (P < 0·05).

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Effect of high K⁺ solutions on increased L_p in response to hypertonic solutions in RGD pretreated vessels

To determine whether the mechanism that increases L_p in response to hypertonic solutions in RGD pretreated vessels involves Ca²⁺ influx to the endothelial cells, we compared the L_p after exposure of microvessels to hypertonic solutions under normal conditions to those under depolarized conditions with high K⁺ solutions. Figure 1 shows a paired experiment where the L_p changes were examined with NK⁺ solution and then with 57·9 mM K⁺ solution. The high K⁺ solution abolished the transient increase in L_p induced by a hypertonic solution (330 versus 230 mosmol l^-1 in frog Ringer solution) when the vessel had been treated with the RGD peptide. The control L_p measured with NK⁺ solution was (3·6 ± 1·0) × 10^-7 cm s^-1 cmH₂O^-1. When the vessel was perfused with GRGDTP (0·3 mmol l^-1), no significant change in L_p was observed. Subsequent exposure to a hypertonic solution caused a transient increase in L_p to a peak value of 33·8 × 10^-7 cm s^-1 cmH₂O^-1 within 1 min followed by a sustained increase (more than two times higher than control). When this vessel was exposed to a hypertonic solution for the second time, but this time in 57·9 mM K⁺ solution, the hypertonic solution did not increase L_p at all. This pattern of changes in L_p was seen in all 10 paired experiments of this kind (Table 1). In the five vessels (vessel nos 6-10) the order of the conditions was reversed, i.e. the depolarizing condition preceded the normal condition. It was found that the order of perfusion did not change the result.

		View larger version [in this window] [in a new window]
Figure 1. Effect of depolarization on hypertonic-induced increase in Lp: a single paired experiment on an RGD pretreated vessel Measurements of Lp are shown from multiple cannulations of a single vessel as a function of time. The measurements were made under the normal condition ([K+] = 2·4 mmol l-1, NK+) then under the depolarizing condition ([K+] = 57·9 mmol l-1, 57·9 mM K+). Whereas a hypertonic solution (330 mosmol l-1 as opposed to the isotonic solution of 230 mosmol l-1; seen as shaded areas) caused a transient increase in Lp in NK+ solution, it did not change Lp in 57·9 mM K+ solution. All data are from single occlusions. The arrows indicate time of re-cannulations. Dashed line indicates a baseline Lp. RGD represents the application of GRGDTP (0·3 mmol l-1).

Table 1. Paired measurements of L_p under normal and depolarizing conditions

Figure 2 shows the results from three experimental groups, i.e. NK⁺, 57·9 and 100 mM K⁺ groups (the data of Fig. 2A has been published elsewhere (Kajimura et al. 1997) and we present here the same data for comparison). In the 57·9 and 100 mM K⁺ groups the mean values for control L_p were (2·0 ± 0·4) × 10^-7 cm s^-1 cmH₂O^-1 (range from 0·6 × 10^-7 to 3·5 × 10^-7 cm s^-1 cmH₂O^-1) and (2·2 ± 0·3) × 10^-7 cm s^-1 cmH₂O^-1 (range from 0·8 × 10^-7 to 4·1 × 10^-7 cm s^-1 cmH₂O^-1), respectively. In contrast to the response to hypertonic solutions under normal conditions (L_p transiently increased 4·6-fold within 1 min and then fell but remained elevated more than 2-fold over the next 6 min as seen in Fig. 2A), the depolarizing condition in solutions of either 57·9 mM K⁺ (n = 7) or 100 mM K⁺ (n = 12) abolished the hypertonic-induced increase in L_p in RGD-pretreated vessels. Thus the effect of hypertonic solutions on capillary permeability was significantly attenuated under these depolarizing conditions (compared with the NK⁺ group; Mann-Whitney U test, P < 0·01). These results are not consistent with the hypothesis that the RGD peptide acts only at an extracellular site, independent of Ca²⁺ influx.

		View larger version [in this window] [in a new window]
Figure 2. Effect of depolarization on hypertonic-induced increase in Lp in RGD pretreated vessels In each experimental group, measurements of Lp were made: (i) in a control perfusate (); (ii) in the presence of GRGDTP (0·3 mmol l-1) in the perfusate (); and (iii) after exposure to a hypertonic solution with GRGDTP in the perfusate (). The ratio of test Lp to a baseline Lp (Lp/Lp0; mean ± S.E.M.) is plotted as a function of time. The time intervals are means ± S.D. A, in NK+ solutions (n = 10). B, in 57·9 mM K+ solutions (n = 7). C, in 100 mM K+ solutions (n = 12). The shaded areas indicate when the hypertonic solution was applied. * P < 0·05 compared with a baseline Lp (Wilcoxon signed rank test). Figure 2A was redrawn from Kajimura et al. (1997).

In the NK⁺ solution the RGD pretreatment caused a small but significant increase in L_p (1·27-fold increase in the first 4 min). In the 57·9 and 100 mM K⁺ solutions, however, no significant increase in L_p was seen during the RGD pretreatment.

Effect of extracellular Ca²⁺ removal on Lp

The ligand-binding function of integrin adhesion receptors is known to be dependent on extracellular divalent cations (for review, see Hynes, 1992; Smyth et al. 1993; Luscinskas & Lawler, 1994). We therefore investigated whether removal of Ca²⁺ might enable a direct study of integrin-mediated cell adhesion, independently of Ca²⁺ influx, which would be reduced in low Ca²⁺ solutions. The experiment consisted of: (i) a baseline measurement of L_p; (ii) measurement of Lp during superfusion and perfusion with a solution that was nominally free of Ca²⁺; and (iii) measurement of Lp after exposure of the vessel to a hypertonic solution that was also nominally free of Ca²⁺. The perfusate contained BSA at a concentration of 1 mg ml^-1. In each of the seven vessels (mean control Lp, (3·0 ± 0·9) × 10^-7 cm s^-1 cmH₂O^-1, range from 0·5 × 10^-7 to 7·4 × 10^-7 cm s^-1 cmH₂O^-1), removal of extracellular Ca²⁺ did not change Lp. The mean ratio LpCa²⁺ free /Lp0 for these seven vessels was 0·97 ± 0·15. This result was similar to that with other Ca²⁺-free solutions where excess EGTA was avoided (He & Curry, 1993). Exposure to hypertonic solutions did not increase Lp in these vessels. In fact, there was a significant decrease in Lp within 1-3 min of exposure to hypertonic solutions (Fig. 3; Wilcoxon signed rank test, P < 0·05). These results are also consistent with the conclusion that Ca²⁺ influx is needed to increase microvessel permeability after exposure to hypertonic solutions. Any effect of Ca²⁺ removal alone to change endothelial cell adhesion is not sufficient to cause increased permeability after cell shrinkage with hypertonic solutions.

		View larger version [in this window] [in a new window]
Figure 3. Effect of extracellular Ca2+ removal on Lp Lp was first measured in a control solution containing BSA (1 mg ml-1) to establish a baseline (not shown). Both perfusate and superfusate were then replaced with a Ca2+-free solution. Each vessel was perfused with this isotonic Ca2+-free solution for 10 min () and then exposed to the hypertonic Ca2+ free solution (). The ratio of test Lp to a baseline Lp (Lp/Lp0; mean ± S.E.M.) was plotted. The time intervals are means ± S.D. The shaded area indicates when the hypertonic solution was applied. * P < 0·05 compared with baseline Lp (Wilcoxon signed rank test, n = 7).

Effect of depolarization on hypertonic-induced increase in Lp in BSA-free Ringer solution perfused microvessels

Having found that depolarizing conditions block an increase in Lp caused by hypertonic solutions in RGD pretreated vessels, we next investigated its effect in BSA-free Ringer solution perfused vessels. We have previously observed that exposure to hypertonic solutions (330 mosmol l^-1) induced a sustained increase in Lp when the vessels were perfused with BSA-free Ringer solution (Fig. 4B; NK⁺). This increase was in addition to the sustained increase due to BSA removal from the perfusate. To determine whether this additional increase in permeability was dependent on Ca²⁺ influx, we investigated the effect of the hypertonic solution under depolarizing conditions on BSA-free Ringer solution perfused vessels. Figure 4A shows a single experiment of this kind where responses to hypertonic solutions were examined under both normal and depolarizing conditions. Exposure to 100 mM K⁺ solution abolished the sustained increase in Lp caused by a hypertonic solution (330 versus 230 mosmol l^-1). The control Lp measured with NK⁺ solution was (4·8 ± 0·2) × 10^-7 cm s^-1 cmH₂O^-1. When BSA was removed from the perfusate, there was a sustained increase in Lp. Subsequent exposure to a hypertonic solution caused a greater increase in Lp, which was also sustained over the 6 min (4·8-fold increase relative to control). In 100 mM K⁺ solution, however, neither removal of BSA nor the hypertonic solution caused an increase in Lp.

		View larger version [in this window] [in a new window]
Figure 4. Effect of depolarization on hypertonic-induced increase in Lp in BSA-free Ringer solution perfused vessels A, measurements of Lp were first made in NK+ solution ([K+] = 2·4 mmol l-1) with the vessel perfused: (i) with a control solution containing BSA (1 mg ml-1, ); (ii) with no BSA in the perfusate (); (iii) with a hypertonic solution containing no BSA (). Then the superfusate was replaced with 100 mM K+ solution ([K+] = 100 mmol l-1) and a similar protocol to (i), (ii) and (iii) was followed but with the perfusate made in 100 mM K+ solution. All data are from single occlusions. The arrows indicate time of re-cannulations. Dashed line indicates a baseline Lp. B, time courses of changes in Lp after exposure to hypertonic solutions from each of the three experimental groups were compared. They were: NK+ (, n = 9); 57·9 mM K+ (, n = 10); and 100 mM K+ (, n = 8). The ratio of test Lp to a mean Lp measured with the perfusate containing no BSA (Lp/LpBSA free; mean ± S.E.M.) is plotted as a function of time. The time intervals are means ± S.D. Shaded area indicates when hypertonic solutions were applied. * P < 0·05 compared with mean LpBSA free (Wilcoxon signed rank test). The NK+ data in B were redrawn from Kajimura et al. (1997).

Figure 4B summarizes time courses of changes in Lp after exposure of microvessels to hypertonic solutions under depolarizing conditions examined in 57·9 and 100 mM K⁺ solutions; these responses were compared with our previous data obtained under normal conditions for comparison (designated as NK⁺) (Kajimura et al. 1997). In contrast to the effect of hypertonic solutions under normal conditions (a sustained increase in Lp over 5 min) the 57·9 mM K⁺ solution partially attenuated the hypertonic-induced increase in Lp and the response was transient, unlike the sustained increase observed in NK⁺ solution. In the 10 vessels investigated in 57·9 mM K⁺ solution, the mean Lp value measured was (3·7 ± 0·8) × 10^-7 cm s^-1 cmH₂O^-1 (range from 1·3 × 10^-7 to 9·1 × 10^-7 cm s^-1 cmH₂O^-1). Lp increased to a mean peak value of 2·7 ± 0·7 times control (significantly smaller than that in NK⁺ solution, comparison made using Mann-Whitney U test, P < 0·05) after 1 min of exposure to hypertonic solutions then fell to 0·9 ± 0·3 times control after 4·5 min.

Although 57·9 mM K⁺ solution was only partially effective in blocking the hypertonic-induced permeability increase, 100 mM K⁺ solution completely abolished the response. In the eight vessels studied in 100 mM K⁺ solution, the mean value of Lp was (2·9 ± 0·3) × 10^-7 cm s^-1 cmH₂O^-1 (range from 1·4 × 10^-7 to 3·7 × 10^-7 cm s^-1 cmH₂O^-1). Hypertonic solutions did not increase Lp at all. In fact, there was a significant reduction in Lp below the Lp measured with BSA-free perfusate (Wilcoxon signed rank test, P < 0·05). These results cannot be accounted for by a simple extracellular mechanism acting independently of Ca²⁺ influx into endothelial cells. They suggest that the Ca²⁺- dependent increase in Lp due to exposure to hypertonic solutions may be added to the Ca²⁺-dependent increase in Lp due to 'Ringer solution perfusion' (BSA-free perfusate).

Effect of bumetanide on Lp response to hypertonic solutions in high K⁺ solutions

We examined other mechanisms whereby high K⁺ solutions may modify the response of endothelial cells. Not only does the high K⁺ solution depolarize the membrane and therefore decrease the electrochemical driving force for Ca²⁺ entry into endothelial cells, but high extracellular K⁺ concentration also changes the chemical driving force on the electroneutral Na⁺-K⁺-2Cl^- co-transporter to favour greater solute uptake into endothelial cells (see Appendix). This may increase the capability of cells to undergo the process of regulatory volume increase (RVI). Thus after cell shrinkage, the cells in high K⁺ solutions may restore their volume more rapidly than those cells in normal K⁺ solutions, thereby attenuating the effect of cell shrinkage to increase permeability. If cells rapidly restore their volume, the micro-occlusion technique may not detect a rapid Lp increase associated with cell shrinkage (for instance changes occurring within 5 s after exposure to a hypertonic solution). To examine this possibility we attempted to block the Na⁺-K⁺-2Cl^- co-transporter by a specific inhibitor, bumetanide, and then exposed the microvessels to a hypertonic solution in 57·9 mM K⁺ solution. We argued that if blocking the Na⁺-K⁺-2Cl^- co-transporter modifies the time course of Lp increase caused by hypertonic solutions, i.e. the pattern of the increase would be changed from transient to sustained, the results would be evidence in favour of the 57·9 mM K⁺ solution changing the chemical driving force on the electroneutral Na⁺-K⁺-2Cl^- co-transporter to favour greater solute uptake into endothelial cells.

The experiments were conducted in 57·9 mM K⁺ solution at all times and the perfusate did not contain BSA at any time. The experimental protocol was as follows: (i) measurement of Lp with the perfusate containing bumetanide (50 µmol l^-1) for 10 min; and (ii) measurement of Lp after exposure to a hypertonic solution with bumetanide present in the perfusate. In six microvessels the mean value for Lp measured with the first perfusate was (2·5 ± 0·6) × 10^-7 cm s^-1 cmH₂O^-1 (range from 1·2 × 10^-7 to 4·3 × 10^-7 cm s^-1 cmH₂O^-1). When the vessels were then exposed to hypertonic solutions, Lp transiently increased 2·9-fold within the first half minute, fell back towards the control value and decreased significantly at 4·6 min after the initial exposure to hypertonic solutions (Fig. 5; Wilcoxon signed rank test, P < 0·05). This result suggests that the high K⁺ solution does not exert its main effect by modifying the driving force on the Na⁺-K⁺-2Cl^- co-transporter.

		View larger version [in this window] [in a new window]
Figure 5. Effect of bumetanide perfusion on the time course of changes in Lp after exposure to a hypertonic solution in 57·9 mM K+ solution The first measurements () with the perfusate containing bumetanide (50 µmol l-1) were followed by measurements after exposure to a hypertonic solution (). The perfusate did not contain BSA and [K+] was 57·9 mmol l-1. The ratio of test Lp to the Lp measured during the first perfusion (Lp/LpBSA free; mean ± S.E.M.) is plotted as a function of time. The shaded area indicates when the hypertonic solution was applied. The time intervals are means ± S.D. * P < 0·05 compared with the value measured with isotonic BSA-free perfusate containing bumetanide (Wilcoxon signed rank rest, n = 6).

Control experiments for depolarizing conditions

We carried out a control experiment to examine the effect of the high K⁺ solutions on the Lp. In each vessel after a baseline Lp was established in NK⁺ solution, the superfusate was replaced with a high K⁺ solution and Lp was re-measured with BSA present in the high K⁺ perfusate (5 min) and then with BSA absent (5 min). The experiments were carried out as above with 57·9 and 100 mM K⁺ solutions.

In six vessels, the mean control Lp measured in NK⁺ solution was (3·9 ± 0·4) × 10^-7 cm s^-1 cmH₂O^-1. After the solution was changed to 57·9 mM K⁺ solution, the mean Lp measured with a perfusate that contained BSA was (3·5 ± 0·6) × 10^-7 cm s^-1 cmH₂O^-1 and after BSA was removed from the perfusate, it was (3·9 ± 0·3) × 10^-7 cm s^-1 cmH₂O^-1. These values were not different from the control (Wilcoxon signed rank test). When the same protocol was carried out in 100 mM K⁺ solution, a similar pattern was observed. Specifically, in another six vessels the mean Lp values under the control condition, after exposure to 100 mM K⁺ during perfusion with the perfusate that contained BSA, and with the BSA-free perfusate were ({FONT size 4·3 ± 0·6), (4·1 ± 0·8) and (4·9 ± 0·6) × 10^-7 cm s^-1 cmH₂O^-1, respectively. There was no significant difference amongst these values (Wilcoxon signed rank test). It was, therefore, found that the depolarizing condition alone did not appear to change baseline Lp but blocked the increase in permeability caused by BSA removal with no pronounced reduction in Lp (Fig. 6A and B). This is consistent with previous findings reported by He & Curry (1991, 1993). In their experiments, the mean ratios for Lp57·9 _mM/Lp0 and LpBSA _{free in 57·9 mM}/Lp0 were 0·9 ± 0·1 (He & Curry, 1991) and 1·0 ± 0·1 (He & Curry, 1993), respectively. These results are similar to the present ones.

		View larger version [in this window] [in a new window]
Figure 6. Effect of high K+ solutions on Lp before and after BSA removal The experiment consisted of: (i) a baseline measurement of Lp in NK+ solution containing BSA (10 mg ml-1); (ii) measurement of Lp during superfusion and perfusion with a high K+ solution for ~5 min; and (iii) measurement of Lp after BSA was removed from the perfusate in a high K+ solution for ~5 min. Ratio of test Lp to a baseline Lp (Lp/Lp0; mean ± S.E.M.) is plotted. Control, ; high K+ solutions with BSA, ; high K+ solutions with no BSA, . A, in 57·9 mM K+ solution (n = 6). B, in 100 mM K+ solution (n = 6).

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

Our results demonstrate that in the RGD pretreated vessels, increasing extracellular K⁺ concentration to depolarize the endothelial cell membrane potential abolished a transient increase in Lp upon exposure of intact microvessels to hypertonic solutions. These observations demonstrate that the action of the RGD peptide to modulate the response of endothelial cells to changes in cell volume is not a simple steric modification of attachment at extracellular sites, but rather is dependent on Ca²⁺ influx into endothelial cells forming the microvessel walls. A similar conclusion can be drawn for microvessels perfused with either Ca²⁺-free or albumin-free perfusates. Our observation that the additional sustained increase in Lp in microvessels perfused with albumin-free perfusates in response to hypertonic solutions was also dependent on Ca²⁺ influx indicates that any effect of albumin removal to change endothelial cell attachment does not, by itself, account for the additional increase in permeability. These investigations provide new insight into the mechanisms which determine the change in microvessel permeability after a reduction in endothelial cell volume (Kajimura et al. 1997). We review the principal assumptions in our experiments first.

Validity of high K⁺ protocol to reduce Ca²⁺ influx

One of the central assumptions of the experimental design was that solutions with increased K⁺ concentration reduced Ca²⁺ influx by depolarizing endothelial cell membranes. Previous studies in our laboratory have found that the initial rise in intracellular Ca²⁺ concentration ([Ca²⁺]_i) in the endothelial cells of frog mesenteric microvessels in situ exposed to inflammatory mediators was attenuated when the microvessels were bathed in high K⁺ solutions (He & Curry, 1991; He et al. 1996; Bates & Curry, 1997). Under the same conditions membrane potential (E_m) was measured using the membrane potential sensitive dye, bis-oxonol (He & Curry, 1995). The study showed that values of the E_m measured with the 57·9 and 100 mM K⁺ solutions were -26·6 and -11·6 mV, respectively, compared with the control E_m value of -51 mV measured in the normal K⁺ solution. These values are close to those predicted by the Goldman-Hodgkin-Katz (GHK) equation when membrane conductance to K⁺ is assumed to be 8·5 times greater than that to Na⁺. Further, the reduction in electrochemical driving force after depolarization accounted for the reduced Ca²⁺ influx (Curry, 1992; Michel & Curry, 1999). The present experiments differ from the previous studies using high K⁺ solutions to the extent that the intracellular ion composition changes when the cells are reduced in volume. As a result, the expected membrane potentials in the presence of high K⁺ solutions are changed slightly, but the expected change in Ca²⁺ influx is small. For example, when endothelial cells shrunk with hypertonic solutions the estimated reductions in Ca²⁺ influx were 49 and 63 % in the 57·9 and the 100 mM K⁺ hypertonic solutions, respectively, compared with 42 and 63 % in the normal volume state. The calculation takes into account the change in intracellular ion composition in the shrunk cell, assuming it acts as a perfect osmometer (O'Neill & Klein, 1992; O'Donnell, 1993).

We investigated an alternative way that high extracellular K⁺ might have modified the response of the cells. Specifically, high K⁺ concentrations may increase the capacity of endothelial cells to restore their volume after exposure of microvessels to hypertonic solutions. Based on the observations of O'Donnell (1993) and O'Neill & Klein (1992), the most likely transport pathway for this effect would be the bumetanide-sensitive Na⁺-K⁺-2Cl^- cotransporter. We estimate a 3·9-fold increase in the net driving force for electrolyte uptake into endothelial cells in 57·9 mM K⁺ solutions relative to that in normal K⁺ solutions (see Appendix for details). If the reduced responses (i.e. reduction in magnitude of Lp increase) to hypertonic solutions observed in high K⁺ solutions were mainly the result of the increased capacity of the cells to volume regulate by this mechanism, then 57·9 mM K⁺ would not attenuate increased permeability after blocking the co-transporter using bumetanide. However, as seen in Fig. 5, addition of bumetanide did not significantly modify the pattern of Lp change in response to hypertonic solutions in vessels perfused with BSA-free Ringer solution in 57·9 mM K⁺ solutions. The possibility of other K⁺ dependent pathways that contribute to endothelial cell volume regulation cannot be eliminated, but appears unlikely on the basis of current knowledge of endothelial cell volume regulation. Thus a Ca²⁺ entry mechanism is likely to be the principal process whereby the high K⁺ solutions modify permeability.

Ca²⁺ influx is required to trigger the cell shrinkage-induced Lp increase

We argued that, if the primary action of the RGD peptide was to modify endothelial cell attachment to the basement membrane and thus the resulting change in cell shape during hypertonic shrinkage, then these mechanisms should not be modified by the reduction in Ca²⁺ influx. However, we found that high K⁺ solutions abolished the transient Lp increase observed in the normal K⁺ solution (Fig. 2). Thus the RGD effect is not simply due to its competitive inhibition of the binding site with the normal substrata at the extracellular site. The results suggest that the cell shrinkage-induced Lp increase is likely to involve the endothelial cell cytoskeleton, which would modify attachment sites between endothelial cells and the basement membrane as well as the attachment between adjacent endothelial cells. Our results also demonstrate that these changes are modulated by Ca²⁺ influx.

The mechanisms responsible for the Ca²⁺ influx associated with changes in cell attachment and cell volume are not well understood, but experiments in cultured endothelial cells suggest some possibilities. For example, there may be an increase in the Ca²⁺ conductance of endothelial cells upon cell shrinkage. Indeed an increase of this kind has been reported in the study of the endothelium of excised intact rat aorta (Marchenko & Sage, 1998). These authors report that hypertonic solutions (30-150 mosmol l^-1 above the normal osmolarity) caused a rise in endothelial [Ca²⁺]_i due to Ca²⁺ entry. The preliminary experiments of Marchenko & Sage (1998) also show that pretreatment of endothelium of rat aorta with cytochalasin B or D decreased the rise in [Ca²⁺]_i in response to hypertonic stress. The result suggests that the endothelial cell cytoskeleton is involved in the signalling pathways linking a change in cell volume to Ca²⁺ entry.

Another mechanism for calcium ion entry into endothelial cells in our experiments may involve a more complex interaction of RGD peptide with the cell membrane. Sjaastad et al. (1994) showed a large transient increase in [Ca²⁺]_i triggered by integrin interaction with RGD-containing peptides immobilized on beads in Madin-Darby canine kidney epithelial cells in vitro. They also showed that this transient increase was partly due to Ca²⁺ influx. Thus, endothelial cells exposed to the RGD peptide may also have had their intracellular Ca²⁺ increased before the exposure to hypertonic solutions. We have previously observed a small but significant increase in Lp (by 27 %) in vessels perfused with an isotonic solution containing BSA and the RGD peptide (Kajimura et al. 1997). This increase may result from an increase in [Ca²⁺]_i during pretreatment with the RGD peptide in our experiments. The depolarizing condition caused by high K⁺ solutions would reduce the extent of this increase due to an influx pathway. Under normal conditions, an initial Ca²⁺ increase due to exposure to RGD peptide may precondition the endothelial cells to respond to the additional hypertonic stress. The same type of mechanism may explain the response of vessels exposed to albumin-free solutions, which respond with an increase in intracellular Ca²⁺ concentration. Thus part of the explanation of the action of high K⁺ solutions may involve attenuation of Ca²⁺ influx during pretreatment with RGD peptide or BSA-free Ringer solution perfusion.

Our experiments with the Ca²⁺-free bathing solution were also designed to separate extracellular effects further from intracellular mechanisms. If the EC-ECM attachment site is compromised by the extracellular Ca²⁺ depletion, an increase in Lp would be expected upon cell shrinkage in the Ca²⁺-free condition. This did not occur (Fig. 3). The result suggests Ca²⁺-free conditions may have attenuated a conformational change which depended on Ca²⁺ influx. Thus, if the principal action of removal of Ca²⁺ was at integrin binding sites, these observations also conform to the hypothesis that changes in both extracellular and intracellular aspects of EC-ECM adhesion contribute to increased permeability. There is evidence to support a role for Ca²⁺ to maintain binding functions (Hynes, 1987; Ruoslahti & Pierschbacher, 1987). Crystallographic studies of the cation-binding loops of parvalbumins, homologous to the putative cation-binding sites of integrin -subunits, demonstrate that Ca²⁺ is thermodynamically preferred to bind to the loop. One problem with this interpretation is that Mg²⁺ also binds to the cation-binding loops (Declercq et al. 1991). Thus it is possible that Mg²⁺ may have been able to maintain the 'normal' conformation of integrins in the absence of Ca²⁺ to such a degree that there was not sufficient change in attachment in our nominally Ca²⁺-free conditions to initiate a change in permeability during cell volume reduction.

For the BSA-free Ringer solution perfused vessels we observed a graded effect of depolarization (as opposed to the complete abolition by either the depolarizing conditions in the RGD pretreated vessels), i.e. the cell shrinkage-induced sustained increase in Lp observed with NK⁺ solution was only partly attenuated with 57·9 mM K⁺ solution and completely abolished with 100 mM K⁺ solution (Fig. 4). These results are consistent with the idea that Ca²⁺ entry into endothelial cells is necessary for the additional increase in Lp in BSA-free Ringer solution perfused microvessels. Our laboratory has already demonstrated that the increase in Lp with BSA-free Ringer solution depends on Ca²⁺ entry. The simplest explanation of action of hypertonic solutions to increase Ca²⁺ influx is that the conductance of one or more passive Ca²⁺ channels is increased. Either the conductance of the Ca²⁺ channel already open due to albumin removal from the perfusate is increased further, or an additional conductive channel for Ca²⁺ entry into the endothelium due to changes in cell volume and/or shape is opened. That 100 mM K⁺ Ringer solution reduces the Lp of microvessels exposed to both BSA-free perfusion and increased osmolarity to values below those measured with BSA-free perfusion alone shows that the effects of BSA-free perfusion and exposure to hypertonic solutions are additive.

Relation of present study to mechanisms which change permeability

The standard model of mechanisms to increase microvessel permeability suggests that Ca²⁺-dependent contractile mechanisms in the endothelium can disrupt cell-cell attachment and cell-matrix attachments sufficiently to cause gaps between the cells at the sites of the intercellular junction and also through endothelial cells near the junctions (Neal & Michel, 1995). We began studies of the effect of endothelial cell volume decrease to test the idea that an external mechanical disturbance on the cells (volume reduction) might modify the intercellular junctions directly (i.e. independently of Ca²⁺ dependent mechanisms). We previously suggested (Kajimura et al. 1997) that, upon a reduction in cell volume, the cells more loosely attached to the ECM may experience pronounced cell-shape changes which disrupt the geometry of the junction between the cells if they tend to round up, e.g. decreasing the mean depth of the clefts from luminal to abluminal surface of the endothelium for exchange. This speculation could be supported by the models of attached endothelial cells based on the principles of tensional integrity described by Ingber et al. (for review, see Ingber, 1994). According to these principles, endothelial cells under continuous tension spread when attachments are made on a rigid substratum, but round up when sites of attachment are released. The new information from the present studies is that an applied force (generated by volume reduction) does not directly modify junctional geometry. Rather, Ca²⁺ influx into endothelial cells modulates the link between the imposed changes in cell shape and changes in endothelial cytoskeleton and permeability.

To date the mechanism of the specific interactions between EC-ECM attachments and intracellular Ca²⁺-dependent processes to regulate permeability has not been investigated. The structure of the integrin-mediated cell-ECM attachment sites has been well investigated and many localized constituent proteins of cell-ECM attachment sites have been identified in studies in vitro. The integrins are transmembrane heterodimers which bridge between ECM proteins and the cell interior. The intracellular aspect of the cell consists of various intermediary proteins, such as -actinin, vinculin, and talin, and these proteins link integrins to the actin cytoskeleton (for review, see Burridge et al. 1988; Albelda & Buck, 1990; Sastry & Horwitz, 1993). Many studies have shown that the engagement of integrins with their specific extracellular ligands induces the phosphorylation of tyrosine (Guan et al. 1991; Kornberg et al. 1991). One of the major substrates for integrin-induced tyrosine phosphorylation is a protein tyrosine kinase, focal adhesion kinase (FAK) (Schaller et al. 1992). Romer et al. (1994) showed that FAK is phosphorylated on tyrosine during migration and tyrphostin, a specific tyrosine kinase inhibitor, inhibited endothelial cell migration during wound repair by inhibition of tyrosine phosphorylation on FAK in human umbilical vein endothelial cells. They suggest that there is a switching mechanism by which cell adhesion and/or detachment events are regulated during migration, i.e. when cells actively adhere to their ligands tyrosine kinases are upregulated, and when cells detach from the ligands tyrosine phosphatases are upregulated. It is possible such a mechanism is regulated by [Ca²⁺]_i in endothelial cells.

One striking observation suggesting that cell adhesion and/or detachment events may be regulated by influx of Ca²⁺ is provided by Hendey & Maxfield (1993). They measured repeated increases in neutrophil [Ca²⁺]_i which correlated with cell spreading and motility. When extracellular Ca²⁺ was removed or buffered by EGTA, neutrophils were unable to detach from a fibronectin or vitronectin substratum, thereby losing their motility. Moreover buffering intracellular Ca²⁺ with quin2 or BAPTA substantially inhibited cell motility. These data strongly suggest that increased [Ca²⁺]_i is required for cell detachment from adherent substrata.

In summary, we have shown that the cell shrinkage-induced permeability increase after the microvessels had been treated with the RGD peptide to disrupt integrin dependent EC-ECM attachment is not simply due to steric disruption of integrin-ECM ligands, but also due to Ca²⁺-dependent processes which are modulated by Ca²⁺ influx. A similar conclusion was reached in the case of BSA-free Ringer solution perfused vessels. Our findings may be useful in understanding how changes in cell-ECM attachment may be one of the processes that modulate how tension that is developed within an endothelial cell is transduced into a change in the pathways regulating transcapillary exchange. Further experiments are required to investigate these models.

Estimates of the driving force for the Na⁺-K⁺-2Cl^- co-transporter in high K⁺ solutions

Estimates of intracellular concentration of each electrolyte after cell shrinkage

In order to estimate the driving force for the Na⁺-K⁺-2Cl^- co-transporter it is necessary to estimate the intracellular concentration of each electrolyte after cell shrinkage. If a cell behaves as a perfect osmometer with a semipermeable membrane, its volume should depend on the osmotic pressure of the surrounding medium. Using a classic equation of Lucké & McCutcheon, the relation when cells are placed in solutions of different osmolarities (Lucké & McCutcheon, 1932; Hoffmann, 1977) can be expressed as:

v = (vo - b)(o/) + b,

(A1)

where is the external osmotic pressure, v is the cell volume, b is the non-solvent volume of the cell (osmotically inactive volume), and _o and v_o are the original (isotonic) external osmotic and the original cell volume, respectively.

According to this relation the reduction in the volume of cell water with a change in osmolarity is given as:

(v - b)/(vo - b) = o/.

(A2)

We assume the cell shrinkage occurs instantaneously after the cell is put into a hypertonic solution and only water can move through the cell membrane. Under this assumption the mass of each of the osmotically active solutes is kept constant.

c(v - b) = co(vo - b), c = co(vo - b)/(v - b),

(A3)

where c and c_o are the concentrations of each solute after cell shrinkage, and before cell shrinkage, respectively.

Using eqn (A2) to substitute for (v_o - b)/(v - b) in eqn (A3) results in the expression:

c = co(/o).

(A4)

Let the intracellular concentrations of electrolytes, [Na⁺]_i, [K⁺]_i, [Cl^-]_i in the isotonic condition be 10, 120 and 30 mmol l^-1, respectively. In our experiments, values for _o and are 230 and 330 mosmol l^-1, respectively. Then following from eqn (A4), [Na⁺]_i, [K⁺]_i, and [Cl^-]_i in the hypertonic condition become 14, 172, and 43 mmol l^-1, respectively.

Driving force for the Na⁺-K⁺-2Cl^- co-transport

Since the co-transport is electroneutral, the thermodynamic driving force is the chemical potential difference for Na⁺, K⁺, and 2Cl^- (µ_Na + µ_K + 2µ_Cl).

µNa-K-2Cl = µNa + µK + 2µCl µNa-K-2Cl = RT ln([Na]i/[Na]o) + RT ln([K]i/[K]o + 2RT ln([Cl]i/[Cl]o).

(A5)

Using eqn (A5), µ_Na-K-2Cl (J mol^-1) in the respective solutions is calculated and values are shown in Table A1.

Table A1. Estimate of the Na⁺-K⁺-2Cl^- co-transport driving force in the hypertonic solution

	[K+]o	[Na+]o	[Cl-]o	µNa-K-2Cl	Ratio *
NK+	2·4	166	172·6	-2307	1·0
57·9 mM K+	57·9	110·5	172·6	-9028	3·9
100 mM K+	100	68·4	172·6	-9190	4·0

	REFERENCES

Top Abstract Introduction Methods Results Discussion References

Albelda, S. M. & Buck, C. A. (1990). Integrins and other cell adhesion molecules. FASEB Journal 4, 2868-2880	[Abstract]
Bates, D. O. & Curry, F. E. (1997). Vascular endothelial growth factor increases microvascular permeability via a Ca2+-dependent pathway. American Journal of Physiology 273, H687-694	[Medline]
Burridge, K., Fath, K., Kelly, T., Nuckolls, G. & Turner, C. (1988). Focal adhesions: transmembrane junctions between the extracellular matrix and the cytoskeleton. Annual Reviews of Cell Biology 4, 487-525.	[Medline]
Curry, F. E. (1992). Modulation of venular microvessel permeability by Ca2+ influx into endothelial cells. FASEB Journal 6, 2456-2466	[Abstract]
Curry, F. E., Huxley, V. H. & Sarelius, I. H. (1983). Techniques in the microcirculation: measurement of permeability, pressure and flow. In Cardiovascular Physiology. Techniques in the Life Sciences ed. Linden, R. T., pp. 1-34. Elsevier.
Davies, P. F. (1995). Flow-mediated endothelial mechanotransduction. Physiological Reviews 75, 519-560	[Medline]
Declercq, J. P., Tinant, B., Parello, J. & Rambaud, J. (1991). Ionic interactions with parvalbumins. Crystal structure determination of pike 4.10 parvalbumin in four different ionic environments. Journal of Molecular Biology 220, 1017-1039	[Medline]
Guan, J. L., Trevithick, J. E. & Hynes, R. O. (1991). Fibronectin/integrin interaction induces tyrosine phosphorylation of a 120-kDa protein. Cell Regulation 2, 951-964	[Medline]
He, P. & Curry, F. E. (1991). Depolarization modulates endothelial cell calcium influx and microvessel permeability. American Journal of Physiology 261, H1246-1254	[Medline]
He, P. & Curry, F. E. (1993). Albumin modulation of capillary permeability: role of endothelial cell [Ca2+]i. American Journal of Physiology 265, H74-82	[Medline]
He, P. & Curry, F. E. (1995). Measurement of membrane potential of endothelial cells in single perfused microvessels. Microvascular Research 50, 183-198	[Medline]
He, P., Zhang, X. & Curry, F. E. (1996). Ca2+ entry through conductive pathway modulates receptor-mediated increase in microvessel permeability. American Journal of Physiology 271, H2377-2387	[Medline]
Hendey, B. & Maxfield, F. R. (1993). Regulation of neutrophil motility and adhesion by intracellular calcium transients. Blood Cells 19, 143-161	[Medline]
Hoffmann, E. K. (1977). Control of cell volume. In Transport of Ions and Water in Animals, ed. Gupta, B. L., pp. 285-333. Academic Press, New York.
Huxley, V. H. & Curry, F. E. (1985). Albumin modulation of capillary permeability: test of an adsorption mechanism. American Journal of Physiology 248, H264-274	[Medline]
Hynes, R. O. (1987). Integrins: a family of cell surface receptors. Cell 48, 549-554	[Medline]
Hynes, R. O. (1992). Integrins: versatility, modulation, and signaling in cell adhesion. Cell 69, 11-25	[Medline]
Ingber, D. E., Dike, L., Hansen, L., Karp, S., Liley, H., Maniotis, A., McNamee, H., Mooney, D., Plopper, G., Sims, J. & Wang, N. (1994). Cellular tensegrity: exploring how mechanical changes in the cytoskeleton regulate cell growth, migration, and tissue pattern during morphogenesis. International Review of Cytology 150, 173-224	[Medline]
Kajimura, M. & Curry, F. E. (1995). Albumin may modulate endothelial cell (EC) binding to the extracellular matrix (ECM) by a calcium dependent process. Microcirculation 2, 71, A40.
Kajimura, M., O'Donnell, M. E. & Curry, F. E. (1994). Integrin-extracellular matrix interaction modulates hydraulic conductivity (Lp) of single perfused frog mesenteric microvessels. Molecular Biology of the Cell, supplement 5, 54a.
Kajimura, M., O'Donnell, M. E. & Curry, F. E. (1997). Effect of cell shrinkage on permeability of cultured bovine aortic endothelia and frog capillaries. The Journal of Physiology 503, 413-425	[Abstract]
Kornberg, L. J., Earp, H. S., Turner, C. E., Prockop, C. & Juliano, R. L. (1991). Signal transduction by integrins: increased protein tyrosine phosphorylation caused by clustering of 1 integrins. Proceedings of the National Academy of Sciences of the USA 88, 8392-8396	[Abstract]
Lucké, B. & McCutcheon, M. (1932). The living cell as an osmotic system and its permeability to water. Physiological Reviews 12, 68-139.
Luscinskas, F. W. & Lawler, J. (1994). Integrins as dynamic regulators of vascular function. FASEB Journal 8, 929-938	[Abstract]
Marchenko, S. M. & Sage, S. O. (1998). Effects of shear stress and changes in osmolarity on endothelium of rat aorta. The Journal of Physiology 506.P, 17P.
Michel, C. C. & Curry, F. E. (1999). Microvascular permeability. Physiological Reviews 79 (in the Press).
Michel, C. C., Mason, J. C., Curry, F. E., Tooke, J. E. & Hunter, P. J. (1974). A development of the Landis technique for measuring the filtration coefficient of individual capillaries in the frog mesentery. Quarterly Journal of Experimental Physiology 59, 283-309.
Neal, C. R. & Michel, C. C. (1995). Transcellular gaps in microvascular walls of frog and rat when permeability is increased by perfusion with the ionophore A23187. The Journal of Physiology 488, 427-437	[Abstract]
O'Donnell, M. E. (1993). Role of Na-K-2Cl cotransport in vascular endothelial cell volume regulation. American Journal of Physiology 264, C1316-1326	[Medline]
O'Neill, W. C. & Klein, J. D. (1992). Regulation of vascular endothelial cell volume by Na-K-2Cl cotransport. American Journal of Physiology 262, C436-444	[Medline]
Romer, L. H., McLean, N., Turner, C. E. & Burridge, K. (1994). Tyrosine kinase activity, cytoskeletal organization, and motility in human vascular endothelial cells. Molecular Biology of the Cell 5, 349-361	[Abstract]
Ruoslahti, E. & Pierschbacher, M. D. (1987). New perspectives in cell adhesion: RGD and integrins. Science 238, 491-496	[Medline]
Sastry, S. K. & Horwitz, A. F. (1993). Integrin cytoplasmic domains: mediators of cytoskeletal linkages and extra- and intracellular initiated transmembrane signalling. Current Opinion in Cell Biology 5, 819-831	[Medline]
Schaller, M. D., Borgman, C. A., Cobb, B. S., Vines, R. R., Reynolds, A. B. & Parsons, J. T. (1992). pp125FAK a structurally distinctive protein-tyrosine kinase associated with focal adhesions. Proceedings of the National Academy of Sciences of the USA 89, 5192-5196	[Abstract]
Sjaastad, M. D., Angres, B., Lewis, R. S. & Nelson, W. J. (1994). Feedback regulation of cell-substratum adhesion by integrin-mediated intracellular Ca2+ signaling. Proceedings of the National Academy of Sciences of the USA 91, 8214-8218	[Abstract]
Smyth, S. S., Joneckis, C. C. & Parise, L. V. (1993). Regulation of vascular integrins. Blood 81, 2827-2843	[Medline]

Acknowledgements

We are grateful to Dr Martha E. O'Donnell for many helpful suggestions on this research and especially for her help with designing the bumetanide experiments. We also thank Dr Peter M. Cala for his advice on the calculation of the driving force for the Na⁺-K⁺-2Cl^- co-transport. This research was conducted during the tenure of a Regents Fellowship of the University of California and an International Fellowship of the American Association of University Women (M. K.). This work was supported by the NIH grant, Merit Award R37 HL 28607 (F. E. C.).

Corresponding author

M. Kajimura: Section of Cellular and Integrative Biology, Division of Biomedical Sciences, Imperial College School of Medicine, The Sir Alexander Fleming Building, South Kensington, London SW7 2AZ, UK.

Email: m.kajimura{at}ic.ac.uk

This article has been cited by other articles:

				B. Nilius and G. Droogmans Ion Channels and Their Functional Role in Vascular Endothelium Physiol Rev, October 1, 2001; 81(4): 1415 - 1459. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kajimura, M. Articles by Curry, F. E. Search for Related Content PubMed PubMed Citation Articles by Kajimura, M. Articles by Curry, F. E.