INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CAPILLARY DIAMETERS under pathophysiological conditions may be reduced below those of the traversing undeformed red and white blood cells (16) and, therefore, may jeopardize cell passage through the microcirculation as well as nutritive blood flow. The major cause of the reduction of capillary patency is endothelial swelling, as observed in hemorrhagic shock (14, 16) or systemic blood acidosis, but not at low flow conditions per se (15). Swelling in these experiments could be prevented by amiloride analogs, as an indication that a Na⁺/H⁺ exchanger (NHE) participates in the swelling process. NHEs have been widely studied as membrane mechanisms involved in the regulation of intracellular pH (pH_i) and are present in endothelial cells (4, 10). Five NHE isoforms (NHE1-NHE5) represent the major acid-extruding transporters in the absence of bicarbonate (HCO₃). In vivo, however, HCO₃ is usually available, and HCO₃-dependent systems contribute to pH_i regulation; among these, the Na⁺-dependent Cl/HCO₃ exchanger (NCBE) has been found activated in acidosis in bovine aortic endothelium (4). In addition, Na⁺-HCO₃ cotransport is activated in corneal endothelium (1, 2, 10). Na ⁺-independent Cl/HCO₃ exchange (CBE) is known to maintain pH_i after alkalinization. All three systems can be inhibited by DIDS.

There are very few in vitro studies concerned with cell-volume homeostasis of endothelial cells. Those available mostly deal with mechanisms involved in cell-volume regulation after exposure to osmotic stress (e.g., 11, 19, 23). In other cell types such as glia, swelling mechanisms during extracellular acidosis have been studied in detail (9, 12, 20, 22, 24). The purpose of the current in vitro study was, therefore, to evaluate the postulated swelling effect of extracellular lactacidosis on endothelial cell volume under closely controlled extracellular conditions and to characterize the transport systems involved.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cell culture. Bovine aortic endothelial (BAE) cells (Stanford University) (5) were maintained as monolayers in petri dishes with the use of DMEM containing 25 mmol/l HCO₃, 10% bovine fetal calf serum (FCS), 100 IU/ml penicillin G, and 50 µg/ml streptomycin. The cells were cultured at 37°C in humidified room air containing 5% CO₂. Subcultivation was performed 2-3 times/wk by washing with physiological saline and trypsinization (0.05% trypsin and 0.02% EDTA). They were subsequently suspended in medium containing FCS for inactivation of trypsin. After centrifugation, the cells were again washed in physiological saline to remove FCS. For experiments, cells were used once they had reached a confluent stage, usually 5-6 days after subcultivation. To investigate endothelial cell swelling, suspended cells were then transferred into an incubation chamber, which allowed for close control of the extracellular environment (12, 13, 24). The chamber was equipped with two apertures for measurement of pH and temperature by respective electrodes. A gas-permeable silicon rubber tubing within the chamber served as a membrane oxygenator, providing cells with a mixture of O₂, CO₂, and N₂ for experiments with HCO₃ buffering or with O₂ and N₂ in HEPES-buffered experiments. A magnetic stirrer prevented cell sedimentation.

Measurement of cell volume. Cell volume was determined by CASY1 technology (Coulter technique employing the "pulse area analysis" signal-processing technique; Schärfe System, Reutlingen, Germany) (24). It combines an established particle measurement technique, the "resistance measuring principle," with the pulse area analysis signal-processing technique. For measurement, the suspended cells are introduced into the measuring cell through a capillary of predefined geometry at a constant-stream velocity. During the measurement, a current is supplied to the capillary via two platinum electrodes. The capillary filled with electrolytes has a defined electrical resistance. While passing the capillary, the cells displace electrolyte solution in proportion to cell volume. Because intact cells have isolation properties, resistance along the capillary rises. The measuring signal is scanned by CASY1 at a frequency of 1 MHz. CASY1 captures the amplitude and width of the pulse and determines the integral of the measuring signal (pulse area analysis). This procedure allows for measurements with a dynamic range of >1:32,000 in volume. To store this wide volume range with high resolution, a multichannel analyzer with 512,000 volume-linear channels is used. Each of 512,000 registers contains the number of cells that have produced the corresponding pulse area value while passing through the measuring pore. From this volume-linear original size distribution, a diameter-linear size distribution with a resolution level of 1,024 channels is computed. All subsequent measuring parameters are determined on the basis of this size distribution.

Experimental groups. Experiments were performed after a 20-min control period utilized for baseline measurements of cell volume, cell viability, and medium osmolality at an extracellullar pH (pH_e) of 7.4. Data from three cell-volume measurements obtained during this control phase (15, 10, and 5 min before the experimental phase) were averaged, and all measurements, including the three baseline samples, were expressed as a percentage of the reference value thus obtained. This permitted the highlighting of volume changes/constancy during the control phase as well as during experimental conditions. Unstable cell volume or an impaired cell viability (trypan blue exclusion) during the control phase led to discharge of the cells. In a control group, the pH_e of the medium was maintained at 7.4 for 1 h (n = 3). In four experimental groups, pH_e was lowered from an initial 7.4 (control phase) to either 6.8, 6.6, 6.4, or 6.0 by addition of isotonic lactic acid (350 mmol/l) to the HCO₃-buffered cell suspension (5-8 experiments/group). The required volumes of lactic acid were read from a titration curve determined beforehand. CO₂ loss from buffering during acidosis was compensated for by increasing the PCO₂ in the membrane oxygenator under control of medium pH and PCO₂ (ABL Radiometer, Copenhagen, Germany). Cell volume and viability were monitored for 25 min during lactacidosis. Osmolality was assessed under control conditions as well as after induction of lactacidosis (Osmomat 030, Gonotec). Respective experiments were repeated in HEPES-buffered medium (40 mmol/l) in the virtual absence of HCO₃ (n = 5-6/group). To verify the data from experiments with BAE cells, the study was also performed with human umbilical cord endothelial cells (HUVEC) at pH 6.4 and 6.0 (n = 6/group) in HCO₃-buffered media.

Statistics. Data are expressed as means ± SE. As a parametric test, a one-way repeated-measures analysis of variance (ANOVA) was used, and, as a nonparametric test, a repeated-measures ANOVA on ranks according to Friedman was used. Experimental groups were compared by ANOVA (Sigmastat, Jandel Scientific). Differences were considered significant if P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Under control conditions, at pH = 7.4 and 37°C, volume and viability of the BAE cells remained stable for up to 70 min; longer periods were not studied. The average BAE cell volume was 1,232.3 ± 19.2 µm³ in HCO₃-buffered medium. Titration of the suspension from pH_e = 7.4 to pH_e = 6.8 by isotonic lactic acid did not lead to a significant increase of cell volume (Fig. 1). Reduction of pH_e to 6.6 caused mild swelling (102.9 ± 2.2% of baseline volume) after 5 min, followed by volume recovery.

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Average cell-volume changes of bovine aortic endothelial (BAE) cells induced by changes of extracellular pH (pHe) after administration of isotonic lactic acid to bicarbonate (HCO3)-buffered media. pHe was changed at time 0 after at least 15 min of incubation of the suspended cells at baseline conditions (pHe = 7.4). Cell volumes are expressed as a percentage of the average of the 3 baseline measurements. * P < 0.05 vs. baseline conditions.

Lowering of pH_e to 6.4 or 6.0, respectively, led to immediate cell swelling. The initial swelling after 3 min at pH_e = 6.4 reached 106.7 ± 5.6% of control. Maximal swelling (107.0 ± 2.2%) could be observed after 5 min. An attempt to regulate cell volume was seen after the swelling maximum (Fig. 1), although a complete recovery of cell volume could not be found. At 10 min after onset of lactacidosis, cell volume had stabilized and remained at a constant level.

At pH_e = 6.0, volume increased to 109.5 ± 2.9% after 3 min and to a maximum of 111.7 ± 4.0% after 5 min. The cell-volume recovery was negligible, and volume remained at an increased level of 110.5 ± 3.6% of control. Further reductions of pH_e to 5.6 or 4.0 in individual experiments did not lead to further increases of cell volume, i.e., maximal swelling occurred at pH_e = 6.0. However, at pH_e levels below 6.0, cell viability decreased fast, which may have prevented the detection of further cell-volume increases.

Endothelial cells suspended in HEPES-buffered medium had a mean cell volume of 1,200.4 ± 12.6 µm³, i.e., cells were only negligibly smaller than in HCO₃-buffered medium. In HEPES-buffered medium, lactacidosis again induced significant increases of cell volume if pH_e was reduced to or below 6.6 (Fig. 2). During the whole observation period, swelling had approximately the same time course as in HCO₃-buffered media, although the degree of swelling was significantly attenuated.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Average cell-volume changes of BAE cells induced by changes of pHe after administration of isotonic lactic acid to HEPES-buffered media. pHe was changed at time 0 after at least 15 min of incubation of the suspended cells at baseline conditions (pHe = 7.4). Cell volumes are expressed as a percentage of the average of the 3 baseline measurements. * P < 0.05 vs. baseline conditions. + P < 0.05 vs. pHe = 6.8.

Experiments with HUVECs were in agreement with above observations (Fig. 3). Again, swelling occurred at pH_e  6.6 (data not shown).

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Cell-volume changes of human umbilical cord endothelial cells (HUVEC) after reduction of pHe to 6.0 in HCO3-buffered media. pHe was changed by addition of isotonic lactic acid at time 0 after at least 15 min of incubation of the suspended cells at baseline conditions (pHe = 7.4). Cell volumes are expressed as a percentage of the average of the 3 baseline measurements. In acidosis, all measurements were significantly elevated compared with physiological pHe (P < 0.05).

Inhibition of acidosis-induced endothelial cell swelling. In the absence of HCO₃, the increase of cell volume at pH_e = 6.0 could have been completely prevented if the cells had been pretreated with EIPA, the specific inhibitor of Na⁺/H⁺ exchange (Fig. 4). Under baseline conditions, the presence of EIPA had no effect on cell volume.

View larger version (23K): [in this window] [in a new window]   Fig. 4.   Effect of ethylisopropylamiloride (EIPA) on cell swelling of BAE cells in HEPES-buffered media. Suspensions were pretreated with EIPA (50 µmol/l) 15 min before administration of lactic acid (). Compared with data from the untreated group (), the acidosis-induced cell-volume increase could be completely prevented. Besides, EIPA had no effect proper on cell volume or viability during the control period.

View larger version (26K): [in this window] [in a new window]   Fig. 5.   Changes in BAE cell volume after lactic acid administration to a level of pHe = 6.0 in the presence () and absence () of DIDS (1 mmol/l) in HCO3-buffered media. The data obtained in the group pretreated with DIDS were similar to those obtained in HEPES-buffered medium in the absence of DIDS (HCO3 free; ).

View larger version (21K): [in this window] [in a new window]   Fig. 6.   Kinetics of BAE cell swelling in the presence and absence of Na+. Compared with HEPES-buffered media (i.e., HCO3 free; ), acidosis-induced cell swelling was nearly absent in Na+- and HCO3-free medium ().

View larger version (24K): [in this window] [in a new window]   Fig. 7.   Effects of ouabain on acidosis-induced swelling of BAE cells in HEPES-buffered media (pHe = 6.0; ). Ouabain was added to the cells 10 min before acidosis induction. , data obtained in the absence of ouabain. Swelling kinetics are slowed in the presence of ouabain. Inset: baseline cell volume was unaffected by ouabain at physiological pH (pHe = 7.4).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The data accumulated herein indeed confirm former in vivo observations that extracellular acidosis may promote endothelial swelling (15). To understand the mechanisms of endothelial swelling, we have to assume the following sequence of events.

Extracellular acidosis leads to an influx of acid equivalents into the cell as described for astrocytes (12, 17, 18, 20, 22). The degree of acidosis studied here is within a range expected to occur in the parenchyma of heart (21), muscle (25), or brain (reviewed in Ref. 12) during ischemia. In HCO₃/CO₂-buffered media, a fast acid entry is likely. It is caused by CO₂ influx, which diffuses easily through the cell membrane, forming carbonic acid and subsequently H⁺ and HCO₃ in the cytosol. The electrochemical gradient for H⁺ favors the development of cytosolic acidosis already at physiological pH_e levels. Hence, a normal pH_i can only be maintained by active pH regulation involving ion transporters, ion channels, and metabolic processes.

Under CO₂/HCO₃-free conditions, the situation is more complex. One might expect enhanced intracellular acidosis compared with baseline conditions, because the cells are HCO₃ depleted. However, because CO₂ diffusion over the cell membrane is thought to be the major mechanism of intracellular acidification, omission of CO₂/HCO₃ may rather result in an attenuated intracellular acidosis, as seen in glial cells (20).

The observed swelling is assumed to result from pH_i regulatory systems such as NHE: the uptake of Na⁺ in exchange for H⁺, which comes from intracellular buffers and is therefore thought to be osmotically inactive (3), increases the osmotic load of the cell and is followed by influx of water. Inhibition of NHE either by EIPA (Fig. 4) or the absence of Na⁺ (Fig. 6) hence prevents swelling in HCO₃-free media. The use of ouabain to partially depolarize the cells results in a delay or slowdown of the swelling response (Fig. 7). This presumably is related to a gradual rise in intracellular Na⁺, which suppresses Na⁺ entry through the NHE.

In the presence of HCO₃, other pH_i regulatory transport systems are activated in extracellular acidosis in addition to NHE, i.e., Na⁺-HCO₃ cotransport or NCBE: an involvement of HCO₃-dependent transport processes is suggested by the larger volume increases seen in HCO₃-containing compared with HEPES-buffered media (Figs. 1 and 2) and by the reduction of swelling in DIDS-treated cells to the level seen in the absence of HCO₃ (Fig. 5). HCO₃-dependent transporters so far have not been discussed with respect to endothelial swelling, although their participation in the regulation of endothelial pH_i is known (1, 2, 4, 10); at physiological pH_e, HCO₃-dependent transport systems contribute more to pH_i homeostasis than NHE. In most cell types, the contribution of NHE to pH_i homeostasis increases with decreasing pH_i (4, 24).

It remains to be determined whether NCBE or Na⁺-HCO₃ cotransport is responsible for HCO₃-dependent swelling, because both are present in endothelium (1, 2, 4, 10). NCBE, however, is less likely to be involved in the swelling response, because uptake of HCO₃ by this transporter and the ensuing buffering of protons generates CO₂, which can readily leave the cell. Because Cl is exported in exchange for HCO₃, this would imply rather a loss of osmotic activity and, hence, cell shrinkage. Na⁺-HCO₃ cotransport, on the other hand, imports Na⁺ as osmotically active particles together with HCO₃ and, therefore, better explains the observed swelling response. Na⁺-HCO₃ cotransport is electrogenic and can function only if membrane potential permits.

In glial cells, despite the activation of pH_i regulatory mechanisms after induction of extracellular acidosis, pH_i does not normalize (17, 18, 20). With ongoing pH_i regulation, one would assume cell swelling to continue until pH_i has normalized. Because swelling kinetics are similar in endothelial cells (Figs. 1-3) and glia (12, 22) (cells swell on acidification with cell volume, reaching a new steady state after a few minutes), it is quite likely that endothelium is likewise acidified in extracellular acidosis. This has to be verified in future experiments. Mellergard et al. (17, 18) offered three possible explanations for the failing pH_i regulation at reduced extracellular pH in astroglia: 1) pH_i is not the regulated parameter, 2) the H⁺ extrusion capacity is reduced, and 3) H⁺ leak fluxes are too high. As an explanation, they favored a reduced acid-extrusion capacity by competitive inhibition of the NHE by extracellular protons.

Likewise unexplained so far is the observation that cell swelling ceases at a given plateau in glia (12, 20, 22) as well as in endothelium (Figs. 1 and 2), particularly if pH_i does not normalize (20). Under these conditions, pH_i regulatory mechanisms and swelling would be expected to continue ion transport until pH_i normalizes. A possible explanation might involve an opening of volume-regulated anion channels once cell volume increases (19), which would render cell membranes more permeable for Cl, HCO₃, and lactate. Recently, Voets et al. (23) elegantly demonstrated that a decrease of intracellular ionic strength rather than an increase of cell volume triggers the opening of such channels, at least in conditions such as hyposmotic endothelial swelling. All processes of the current project were studied in strict isotonicity, which, therefore, makes an involvement of volume-regulated anion channels less likely.

Another possible explanation would be the recently described control of NHE1 activity by an intracellular Na⁺ receptor (7). That receptor is thought to provide a general mechanism for regulating the intracellular Na⁺ concentration in epithelia, where its activation reduces NHE1 activity. In nonepithelial cells, NHE has also been reported to be inhibited by increased intracellular Na⁺ concentration (6). Inhibition of NHE would reduce Na⁺ influx and, hence, cell swelling and could explain the plateau phase observed at all levels of acidosis tested. The experiments employing ouabain to partially depolarize cells, however, only yielded a slowdown of the swelling response but did not affect the plateau phase. Inhibition of Na⁺-K⁺-ATPase by ouabain was presumably followed by a gradual increase of intracellular Na⁺ and should thereby inhibit NHE (and swelling) via activation of the intracellular Na⁺ receptor (7) earlier than in the absence of ouabain. This did not occur. Hence, further experiments including measurements of pH_i are required to verify details of the mechanisms of the swelling response in extracellular acidosis.

In conclusion, endothelial swelling in extracellular lactacidosis is a result of an activation of ion transport systems involved in pH_i regulation, in particular, NHE and Na⁺-HCO₃ cotransport. Under in vivo conditions, acidosis-induced endothelial swelling may hamper microcirculatory blood flow. Drugs that interfere with pH_i regulation can prevent the swelling response and, hence, may positively affect postischemic microcirculation.