INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

EXPOSURE OF ANIMAL CELLS to hypertonicity induces an immediate cell shrinkage by osmosis usually followed by an early regulatory volume increase (RVI) mediated by the uptake of inorganic ions and accompanying water (19). Thus the consequence of this response to the extracellular pressure is an abnormally high concentration of inorganic ions inside the cells. This early phase of RVI is followed in some cells by the substitution of various "compatible osmolytes" in place of some of the inorganic ions, which enables the cells to maintain a cell volume adequate for their survival in the adverse hypertonic environment (5). The usual explanation of this is the need to overcome the deleterious effects of high salt concentrations on macromolecular structures that otherwise would eventually impair or destroy normal cell function. Osmotically equivalent concentrations of "compatible osmolytes," such as methylamines, polyols, and some amino acids or amino acid derivatives, do not perturb protein or nucleic acid structures (53). This process of cell adaptation to hypertonicity often involves changes in gene expression that result in an increase in either the synthesis or the accumulation of compatible osmolytes (6, 28).

In a number of cultured cells, exposure to hypertonic medium produces an early (2-6 h) increased activity of amino acid transport System A (8, 12, 14, 15, 25, 41, 42, 46, 49). After more prolonged (8-24 h) exposure of cells to hypertonicity, an increased activity of the betaine-GABA (BGT-1) transporter has been described in kidney-derived epithelial cells (17, 24, 41, 52) and in vascular endothelial cells obtained from calf pulmonary artery (25). Primary cultures of porcine chondrocytes respond in an exactly parallel fashion, although the induced betaine-GABA transporter in these cells may not be BGT-1 (15). On the other hand, rat liver sinusoidal endothelial cells (51), such as RAW 264.7 mouse macrophages (50), human monocytes, and macrophages (16), respond to hypertonic conditions with an early (6-12 h) upregulation of the BGT-1 transporter.

Although these responses of kidney-derived epithelial cells, and perhaps of chondrocytes, to hypertonicity are readily understandable in terms of their normal physiological functions, the similar behavior of the other cells is not so obviously explicable. Endothelial cells mediate and monitor the exchange of molecules between the plasma and the interstitial fluid, participating in homeostasis and playing a critical role in the inflammatory process (22). Because they are normally exposed to isotonic fluids containing rather constant nutrient supplies, however, it is not obvious why they should possess mechanisms that enable them to cope with hypertonicity. It is possible that certain pathological or trauma conditions might require such responses (1); under diabetic conditions, for example, endothelial cells are exposed to hyperglycemia, but normally isotonic volume changes are more likely (36). Despite this uncertainty, the somewhat different responses reported for two different kinds of endothelial cells, one similar to the kidney-derived cells and the other to monocytes and macrophages, are both impressively efficacious in coping with hypertonicity. The existence of these responses, which are quite costly in terms of metabolic input, suggest they have a physiological role.

Endothelial cells are a very heterogeneous class, with functional differences between those derived from arterial or venous origin, large or small vessels, and various different microvascular beds. There are also differences between endothelia from different species (2). More detailed study of their properties and behavior has been facilitated by recent improvements in culture techniques that provide endothelial cells from different origins able to maintain in vitro the differentiated properties of the original endothelium. We have examined in some detail the adaptive responses to a hypertonic environment of cultured endothelial cells derived from pig pulmonary arteries. The present work describes the adaptation from the initial cell shrinkage followed by changes in inorganic ion concentrations, the occurrence of an RVI, the induction of both System A and BGT-1 transport activities, the accumulation of their specific substrates, and the regulation of BGT-1 gene expression by accumulated betaine.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Chemicals. -[³²P]dCTP, -[³²P]ATP, 3-O-methyl-D-[1-³H]glucose, L-[4,5-³H]leucine, and -amino[2,3-³H]butyric acid were obtained from Amersham International, Amersham, Bucks, UK. 2-[1-¹⁴C]methylaminoisobutyric acid (MeAIB) was obtained from DuPont-New England Nuclear (Boston, MA). Choline oxidase, betaine, GABA, methylaminoisobutyric acid, collagenase type IA were bought from Sigma Chemical, Poole, Dorset, UK. [¹⁴C]betaine was prepared from [¹⁴C]choline by enzymic oxidation (41). A plasmid containing full-length BGT-1 cDNA (52) was kindly provided by Dr. Moo Kwon, Division of Nephrology, The Johns Hopkins University School of Medicine, Baltimore, MD, and a probe for 28S rRNA was obtained from Dr. Lorenza Tacchini (University of Milan, Italy). Single-stranded oligonucleotides containing the hypertonicity-responsive element Ton-E (upper stranded 5^'-TACTTGGTGGAAAAGTCCAG-3^') were obtained from Primm (Milan, Italy). Media, fetal calf serum, and antibiotics for culturing the cells were purchased from GIBCO (Grand Island, NY). Disposable plastics for laboratory use were obtained from Costar (Broadway, Cambridge, MA). Reagents of analytic grade were purchased from Sigma Chemical.

Cell culture. Pulmonary arteries were obtained aseptically from healthy 3-wk-old pigs after they had been given an anesthetic. The vessels were submersed in 100 ml of Medium 199 containing penicillin (1,000 U/ml), streptomycin (500 µg/ml), and amphotericin B (2 µg/ml) (29). The fat and connective tissue surrounding blood vessels were removed, and the remaining tissue was minced with scissors. Pieces of tissue were then incubated for 2 h at 37°C in Medium 199 containing, in addition to the antibiotics listed above, 0.25% collagenase (type 1A) and 0.25% bovine serum albumin. The resultant cell suspension was rinsed in Medium 199 containing fetal calf serum (10% vol/vol) and filtered through a 20 µm pore size filter to remove all microaggregates containing unwanted cells. After centrifugation of the filtrate at 1,000 g for 20 min, the resulting pellets were resuspended in endothelial basal Medium 199 supplemented with fetal calf serum (10%), heparin (50 U/ml), hydrocortisone (1 µg/ml), epidermal growth factor (10 µg/ml), and endothelial cell growth supplement (50 µg/ml). The cells were seeded in 75-cm² flasks coated with sterile gelatin (2%). After cell attachment was ascertained by examining the flasks with an inverted microscope, the medium was removed and fresh medium was added to cell culture. Subsequently, the medium was changed again at 2-day intervals. The first confluence was obtained after ~10 days. Subpassages were made weekly by rinsing confluent monolayers with PBS and dispersing the sheet with a trypsin-EDTA solution. The cell suspension was inoculated into new flasks coated with 2% gelatin, and cells were used when confluent, usually within 3-5 days after plating. Growth medium was Dulbecco's modified Eagle's medium containing 100 U/ml penicillin, 100 µg/ml streptomycin, and supplemented with 2 mM glutamine, 2 mM sodium pyruvate, 20 mM HEPES (pH 7.4), 10% fetal calf serum, 50 U/ml heparin, and 50 µg/ml endothelial cell growth supplement. All cultures were kept in an incubator at 37°C in a water-saturated atmosphere of 5% CO₂ in air. Cells used in this study were all of early (up to the 10th) passages and were routinely checked for the expression of von Willebrand Factor by immunocytochemistry. Rabbit anti-human von Willebrand Factor (DAKO) was the purified immunoglobulin fraction of rabbit antiserum and cross-reacted with von Willebrand Factor from pig.

Culture media. The osmolalities of the culture media were checked with a vapor-pressure osmometer (Wescor). Normal growth medium was ~0.3 osmol/kgH₂O. Except where indicated (see Fig. 6A), hypertonic growth medium was made by addition of 200 mM sucrose to normal growth medium, giving a final osmolality of ~0.5 osmol/kgH₂O.

Cell counting and determination of cell survival. Cells were detached from the substratum with trypsin, and appropriate dilutions of the resulting suspension were counted in a hemocytometer, as described in detail previously (44). Cell survival was determined by cell viability and colony formation. After hyperosmotic treatment, the cells were removed from the plates with trypsin and counted, and their viability was determined by trypan blue exclusion (20). Colony formation by viable cells was then determined by seeding them at a density of 400 cells/9 cm² in dishes containing 2 ml of culture medium. After 5-6 days incubation, the cells were fixed with 95% ethanol, stained with 0.1% crystal violet, and counted.

Determination of the rate of protein synthesis. The rate of protein synthesis was measured as the rate of incorporation of labeled leucine of constant specific activity (2.5 mCi/mmol, 2 µCi/ml) during a 30-min incubation of the cell monolayers in complete culture medium containing 0.8 mM leucine. This procedure has been described in detail elsewhere (43).

Determination of cell volume. Intracellular volumes were estimated by measurement of the equilibrium distribution of 3-O-methyl-D-[1-³H]glucose as described previously (49) using the method of Kletzien et al. (27).

Assay of Na⁺ and K⁺ content. The amounts of intracellular Na⁺ and K⁺ were estimated by the following procedure, based on that described by Dall'Asta et al. (13). Cell layers were quickly washed four times with ice-cold 0.1 M MgCl₂ and drained for a few minutes by inversion of the plates. The cells were then denatured by the addition of ethanol (0.5 ml/well) and allowed to dry before the cations were extracted with 10 mM CsCl (2 ml/well). The concentrations of Na⁺ and K⁺ ions in these extracts were measured with a Varian AA-275 atomic absorption spectrophotometer and standard solutions of NaCl and KCl containing 10 mM CsCl. The extracted cells were dissolved in 0.2 M NaOH, and samples were used for the assay of protein content as described below.

Ninhydrin-positive substances. The intracellular content of ninhydrin-positive substances (NPS) was measured by the method of Law and Turner (31) and expressed as nanomoles per milligram of protein.

Transport measurements. The rates of uptake of amino acids or betaine by endothelial cells were measured after the latter had been incubated in isotonic (0.3 osmol/kgH₂O) or hypertonic (0.5 osmol/kgH₂O) medium for the desired time. MeAIB was used as the characterizing substrate of amino acid transport System A in many types of mammalian cells (11, 18), including human endothelial cells derived from the umbilical vein (7). The cell monolayers were quickly washed with Earle's balanced salt solution containing 0.1% glucose and then incubated in this solution for 15 min at 37°C to diminish the cellular pool of amino acids. The cells were washed again and immediately incubated at 37°C for the desired time in the presence of labeled betaine, GABA, or MeAIB. When required, choline was then used in place of Na⁺ in the medium and acetate in place of Cl. The incubations were stopped by removal of the medium, and the cells were quickly washed three times with fresh cold medium. Trichloroacetic acid (5%, wt/vol) was added to denature the cells, and the radioactivity in samples of the acid extracts was measured by scintillation counting. Cell protein, precipitated by TCA, was dissolved in 0.2 N NaOH and its concentration was determined by a dye-fixation method (BioRad) using bovine serum albumin as standard (4).

Northern blotting. Total RNA was extracted from cultured cells using the Ultraspec RNA Isolation System from Biotecx (10). RNA samples (20 µg) were fractionated by electrophoresis through 1% agarose gels, and the separated bands were transferred to nylon filters. The quality and quantity of RNA blotted on the membranes were checked by ultraviolet absorption. Full-length BGT-1 cDNA (52) and 28S rRNA were nick-translated (Amersham kit no. 5000) with [-³²P]dCTP (3,000 Ci/mmol). Hybridization, washing, and autoradiography were carried out exactly as described previously (40).

Gel mobility-shift analysis. The gel retardation assay was conducted essentially as described by Mosser et al. (33) and Choi et al. (9) with the use of a double-stranded synthetic consensus oligonucleotide corresponding to nucleotides 69/50 in the 5^'-flanking region of the canine BGT-1 gene containing the hypertonicity-responsive element Ton-E (48). The entire procedure has been described in detail previously (39).

Statistical analyses. Results are expressed as mean values ± SE for the indicated number of independent measurements. The significance of differences between the mean values recorded for different experimental conditions was calculated by the Student's t-test, and P values are indicated where appropriate in the figures and their legends.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Effects of hypertonicity on cell protein synthesis, growth, and survival. Hypertonic treatment did not affect the ability of endothelial cells to form colonies but did cause a decrease (~50%) in the rate of cell growth during the first 24 h of incubation in 0.5 osmol/kgH₂O medium. Subsequently, however, the rate of cell proliferation increased, whereas that of cells in isotonic medium slowed, so that after another 24 h cell density under hypertonic conditions was only a little lower than under isotonic conditions (Fig. 1A). Total protein synthesis was inhibited >40% during the first 30 min of exposure of the cells to hypertonic medium. Thereafter recovery was rapid, so that after 2 h the inhibition was <20% and by 4 h there was no significant difference between cells from hypertonic and isotonic conditions (Fig. 1B). Taken together, these data show that endothelial cells survive the hypertonic challenge and, after a transient period of inhibition, resume their usual growth rate in this adverse environment.

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Effect of hypertonicity on cell proliferation and protein synthesis. A: 48 h after seeding and culture in isotonic (0.3 osmol/kgH2O) medium, some cells were transferred to hypertonic (0.5 osmol/kgH2O) medium and culture in both media was continued for another 48 h. Cell growth was estimated by measuring the protein content per well every 24 h. B: the rate of protein synthesis was measured as the incorporation of labeled L-leucine into endothelial cell proteins during a 30-min pulse after the cells had been incubated for the indicated times in isotonic (0.3 osmol/kgH2O) or hypertonic (0.5 osmol/kgH2O) culture medium. Osmolality was adjusted by addition of sucrose. Each value is the mean (±SE) of 6 measurements. Differences between isotonic and hypertonic conditions: **P < 0.01; *P < 0.05; NS, not significant.

Cell volume, cation, and NPS content. The volume of the endothelial cells was monitored during their incubation in 0.5 osmol/kgH₂O medium. To determine cell volume changes at early times of the hypertonic treatment, we measured the concentration of labeled methylglucose in cells that had been preincubated with the tracer before treatment until it had reached equilibrium distribution. With the use of these conditions, cell shrinkage was observed during the first few minutes of incubation at 0.5 osmol/kgH₂O. This rapid shrinkage was followed by an RVI, and the cells regained the volume of control cells within 1 h of treatment (Fig. 2).

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Effect of hypertonicity on cell volume. Cells were preincubated in isotonic (0.3 osmol/kgH2O) medium with labeled 3-O-methyl-D-glucose for 30 min (to obtain equilibrium distribution) and then incubated in the continuous presence of the tracer in hypertonic (0.5 osmol/kgH2O) medium. Sucrose was used as the extra osmolyte. At the indicated times, the distribution of the tracer was measured and the cell volume was calculated as described in the MATERIALS AND METHODS. Each value is the mean (±SE) of 6 measurements. (* P < 0.05; ** P < 0.01, compared with the value at time zero.)

View larger version (36K): [in this window] [in a new window]   Fig. 3.   Effect of hypertonicity on the monovalent cation content. Endothelial cells were incubated in isotonic (0.3 osmol/kgH2O) or hypertonic (0.5 osmol/kgH2O) medium, the increased osmolality being obtained by addition of sucrose. After the indicated times, both the volumes of the cells and their contents of Na+ and K+ were measured, as described in MATERIALS AND METHODS. A: Na+ and K+ content expressed as nmol/mg of cell protein. B: intracellular Na+ and K+ concentrations. The values are the means (±SE) of 6 measurements. (*P < 0.05; **P < 0.01, compared with values at time zero.)

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Effect of hypertonicity on ninhydrin-positive solutes (NPS) content. Endothelial cells were incubated in isotonic (0.3 osmol/kgH2O) or hypertonic (0.5 osmol/kgH2O) medium and at the indicated times their NPS content was assayed as described in MATERIALS AND METHODS. Sucrose was used as the extra osmolyte. Values given are the means (±SE) of 6-12 measurements.

Induction of transport activities. Continuous incubation of endothelial cells in 0.5 osmol/kgH₂O medium resulted in a marked but temporary increase in the ability of the cells to take up MeAIB, taken as a measure of the activity of amino acid transport System A. The increased uptake of MeAIB (measured as initial rate of transport) was detected as early as 3 h after the beginning of the treatment, reached a maximum after 6 h, and then slowly declined toward isotonic control values (Fig. 5). This hypertonic treatment also resulted in a progressive increase in the transport activity of BGT-1, assayed by the measurement of the initial rates of uptake of GABA. The timing of this increase, however, was quite different, being just detectable at 6 h it then followed a sigmoid activation curve to reach a maximum at ~24 h (Fig. 5). These inductions of transport activity appear to depend only on the change in osmolarity of the culture medium, rather than on any specific solute or ionic effect, because similar responses were observed when extra NaCl was used instead of extra sucrose to make the culture medium hypertonic (Fig. 6A).

View larger version (25K): [in this window] [in a new window]   Fig. 5.   Induction of transport activities. The initial rates of uptake of 2-[1-14C]methylaminoisobutyric acid (MeAIB) and GABA (extracellular concentrations 0.1 mM) by endothelial cells were measured after the cells had been incubated in either isotonic (0.3 osmol/kgH2O) or hypertonic (0.5 osmol/kgH2O) medium for the indicated times. Sucrose was used as the extra osmolyte. Values given are the means (±SE) of 3 measurements.

View larger version (21K): [in this window] [in a new window]   Fig. 6.   Comparison of osmolytes and dose response of hypertonic treatment. Endothelial cells were incubated for 6 or 24 h in media of the indicated osmolality, and then their initial rates of uptake of MeAIB (after the 6-h incubation) and GABA (after the 24-h incubation) were measured as described in MATERIALS AND METHODS. A: isotonic solutions were ~0.3 osmol/kgH2O, and hypertonic solutions were ~0.5 osmol/kgH2O, osmolality being adjusted with either sucrose of NaCl, as indicated. Values are the means (±SE) of 6 measurements. (**Difference between sucrose and NaCl is significant, P < 0.01.) B: final osmolalities were obtained by addition of sucrose to complete culture medium. Values are the means (±SE) of 3 measurements. Note that the y-axis scale is logarithmic.

View larger version (28K): [in this window] [in a new window]   Fig. 7.   Density-dependent changes of amino acid transport. Endothelial cells were seeded at different densities and cultivated under isotonic (0.3 osmol/kgH2O) conditions for 2 days. Then they were incubated in isotonic (0.3 osmol/kgH2O) or hypertonic (0.5 osmol/kgH2O) media, and their initial rates of uptake of MeAIB and GABA were measured after 6 and 24 h, respectively, as described in MATERIALS AND METHODS. Sucrose was used as the extra osmolyte. Cell densities (µg cell protein/cm2) were measured at the same time. The lines shown are linear regressions, but note that the y-axis scale is logarithmic.

Characteristics of induced GABA transport. The dependence of GABA transport on inorganic ions was checked for the characteristics expected of BGT-1 activity. Figure 8A shows clearly that the induced GABA transport activity was not only Na⁺ dependent but also anion dependent, acetate supporting only ~25% of the uptake given with Cl as the main anion. Much of the smaller influx of GABA into the isotonic control cells was similarly dependent on Na⁺ and Cl. Similar results were obtained when gluconate was used to replace chloride. Mean values (±SE, n = 6) for GABA uptake (nmol · mg protein¹ · 5 min¹) from media containing mainly NaCl, sodium acetate, or sodium gluconate, respectively, were 1.10 ± 0.03, 0.53 ± 0.01, and 0.53 ± 0.01 under isotonic conditions and 7.88 ± 0.26, 2.02 ± 0.08, and 2.25 ± 0.005 under hypertonic conditions. Figure 8B shows that the rate of betaine uptake similarly increased after incubation of the cells for 24 h in hypertonic medium. In addition, betaine influx was almost completely inhibited by the addition of 10 mM GABA, but not significantly affected by 10 mM MeAIB. All these observations support the notion that GABA influx reflects BGT-1 activity and indicate that betaine is taken up almost completely via system BGT-1 in cells exposed to hypertonic conditions for 24 h.

View larger version (22K): [in this window] [in a new window]   Fig. 8.   Characterization of betaine-GABA transporter (BGT-1) transport activity: a discrimination analysis. Endothelial cells were incubated for 24 h in either isotonic (0.3 osmol/kgH2O) or hypertonic (0.5 osmol/kgH2O) medium, sucrose being used as the extra osmolyte. A: the initial rate of GABA uptake (from 0.1 mM) was then measured in the presence of NaCl (control), choline chloride, or CH3COONa as the main salt. B: the effects of MeAIB and GABA (final concentrations 10 mM) on the initial rate of uptake of betaine (from 0.1 mM) were measured. Values given are the means (±SE) of 3 determinations. (*P < 0.05; **P < 0.01, compared with control values.)

Betaine uptake during hypertonic incubation. Endothelial cells incubated in media containing 0.1 mM betaine accumulated this compatible osmolyte so that its intracellular concentration was higher than that outside under both isotonic and hypertonic conditions (Fig. 9). The extent of accumulation, however, was much greater under hypertonic conditions, the cellular concentration reaching ~20-25 mM after 24 h.

View larger version (19K): [in this window] [in a new window]   Fig. 9.   Betaine uptake. Endothelial cells were incubated in isotonic (0.3 osmol/kgH2O) or hypertonic (0.5 osmol/kgH2O) medium containing 0.1 mM labeled betaine, and its accumulation by the cells was measured at the times indicated. Sucrose was used as the extra osmolyte. Values are the means (±SE) of 3 measurements.

Effects of actinomycin D and cycloheximide. The requirement of transcription and translation for the hypertonic induction of the transport activities was explored with the use of specific inhibitors of both gene expression steps: actinomycin D (80 nM) and cycloheximide (35 µM), which inhibit RNA or protein synthesis, respectively, by ~90% in cultured animal cells (49). As shown in Fig. 10, each inhibitor markedly decreased the induction of uptake of both MeAIB and GABA, thus showing the requirement of gene expression for the induction of both System A and BGT-1 activities. (The apparent stimulatory effect of both inhibitors on the control value of GABA uptake may be explained in terms of the lower cell density reached by the endothelial cultures in the presence of actinomycin D or cycloheximide during the 16 h incubation; see Fig. 7.)

View larger version (27K): [in this window] [in a new window]   Fig. 10.   Effects of actinomycin D and cycloheximide on hyperosmotic induction of System A and BGT-1. Endothelial cells were incubated for 6 or 24 h in either isotonic (0.3 osmol/kgH2O) or hypertonic (0.5 osmol/kgH2O) medium in the absence or presence of 80 nM actinomycin D (A) and 35 µM cycloheximide (B). Sucrose was used as the extra osmolyte. Initial rates of uptake of MeAIB or GABA were then measured as described in MATERIALS AND METHODS. Values given are the means (±SE) of 3 determinations.

Induced transcription of transporter mRNA. The mRNA coding for the BGT-1 transporter has been isolated from Madin-Darby canine kidney (MDCK) cells and the gene cloned (52). We tested the endothelial cells to see if the induced GABA transport activity was accompanied by increased gene expression detectable by a BGT-1 probe. Northern blotting with full-length BGT-1 cDNA readily revealed the presence of BGT-1 mRNA in both MDCK cells and the endothelial cells exposed to hypertonic (0.5 osmol/kgH₂O) conditions for 20 h, but not in control cells incubated in isotonic (0.3 osmol/kgH₂O) conditions (Fig. 11). This result suggests, therefore, that the increase in transport of GABA and betaine observed in endothelial cells cultured in hypertonic conditions is due to an increased expression of a gene identical, or very similar, to the BGT-1 gene of the MDCK cells. It also indicates that no marked difference in the sequence of this gene occurs in the cells of these different tissues (kidney/endothelium) and species (canine/porcine).

View larger version (62K): [in this window] [in a new window]   Fig. 11.   Expression of mRNA for BGT-1. Endothelial cells (SPAE) were incubated for 20 h in 0.3 osmol/kgH2O (C) medium, 0.5 osmol/kgH2O (T) medium, or 0.5 osmol/kgH2O medium containing 5 mM betaine (T+Bet). Madine-Darby canine kidney (MDCK) cells were similarly incubated in 0.5 osmol/kgH2O (T) medium for 20 h. Sucrose was used as the extra osmolyte in the test solutions. Then total cellular RNA was extracted and analyzed for the detection of BGT-1 mRNA by Northern blotting, as described in MATERIALS AND METHODS. 28S rRNA was used for standardization.

1

1

View larger version (33K): [in this window] [in a new window]   Fig. 12.   Effect of betaine on hypertonic induction of System A and BGT-1 activities. Endothelial cells were incubated in isotonic (0.3 osmol/kgH2O) or hypertonic (0.5 osmol/kgH2O) medium in the absence and in the presence of 5 mM betaine. Sucrose was used to adjust the osmolality of the hypertonic medium. Initial rates of uptake of MeAIB or GABA by the cells were measured after 6 or 24 h incubation, respectively, as described in MATERIALS AND METHODS. Values given are the means (±SE) of 4 measurements. (*P < 0.05; **P < 0.01, compared with values in the absence of betaine.)

1

1

1

1

Induction of Ton-E binding. Takenaka et al. (48) identified a regulatory element (which they named Ton-E) in the promoter region of the BGT-1 gene in MDCK cells that mediates the transcriptional stimulation in response to hypertonicity. We used the gel mobility-shift assay to see if the increased expression of the BGT-1 gene in the endothelial cells is related to activation of Ton-E. As shown in Fig. 13, extracts from the endothelial cells that had been exposed to hypertonic (0.5 osmol/kgH₂O) conditions for 16 h mimicked those from similarly treated MDCK cells in producing a band shift. Competition with an excess of unlabeled Ton-E oligonucleotide indicated that the observed binding was specific and extracts from control cells incubated under isotonic (0.3 osmol/kgH₂O) conditions did not produce a band shift.

View larger version (56K): [in this window] [in a new window]   Fig. 13.   Gel mobility-shift analysis of Ton-E. Endothelial cells (SPAE) and MDCK cells were cultured in isotonic (0.3 osmol/kgH2O, control) or hypertonic (0.5 osmol/kgH2O, test) media for 16 h and then cells extracts (20 µg) were incubated with radiolabeled, double-stranded oligonucleotide corresponding to nucleotides 69/50 in the 5'-flanking region of the canine BGT-1 gene (Ton-E). In the competition experiment, the cell extracts were mixed with a 100-fold molar excess of unlabeled double-stranded Ton-E before the addition of labeled Ton-E. The arrow indicates the position of the specific complex.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The most notable feature of these observations is the remarkable ability of the endothelial cells to adapt to hypertonic conditions: they appear to be as good, if not better, than MDCK cells in this respect. As has been pointed out by O'Neill (36), mammalian cells do need volume-regulatory mechanisms for a number of reasons and control of cell volume appears to be particularly important for the barrier function of endothelial cells (23). However, isotonic volume change, resulting from a change in the cellular content of osmotically active molecules, is much more likely than anisotonic change due to a change in extracellular osmolarity (36). There seems to be no evidence for an exposure of endothelial cells to hypertonicity parallel to that normally experienced by some kidney cells. Yet the response of these porcine endothelial cells to hypertonic stress is remarkably similar to that of MDCK cells, particularly with regard to the induction of System A and BGT-1 activities. In this respect they resemble endothelial cells from calf pulmonary artery (25) rather than rat liver sinusoidal endothelial cells (51). The latter, however, together with RAW 264.7 mouse macrophages (50), human monocytes, and macrophages (16), also adapt extremely well to hypertonicity and respond unusually quickly in terms of the induction of BGT-1 after only 6-12 h of exposure to hypertonic conditions. As far as we know, there are no reports about the induction of the amino acid transport System A in these latter cells. This would be interesting to test, because in the endothelial cells from pulmonary arteries, as in most other cells examined, it is the induction of System A that reaches a peak after ~6 h exposure to hypertonicity.

Regulatory volume increase. These results show that K⁺ influx plays a major role in the RVI after the initial hypertonic cell shrinkage. The early changes in cell content and concentration of K⁺ are consistent with influx of Na⁺ and K⁺ via Na⁺-K⁺-Cl cotransport being the primary response to hypertonicity in these cells. The consequent increase of cellular Na⁺ concentration would increase the rate of Na⁺-K⁺-ATPase activity, expelling Na⁺ in return for more K⁺, so that the net result would be a gain of K⁺, as observed. This interpretation is consistent with other reports of the involvement of Na⁺-K⁺-Cl cotransport in the regulation of the volume of other endothelial cells (26, 34, 35, 37). With sucrose as the extra osmolyte, however, it is not sure whether the net driving force for this cotransport is inward. From the figures we have, the cellular Cl concentration in the shrunken cells would have to be less than ~35 mM to make ([Na⁺][K⁺][Cl]²)_out > ([Na⁺][K⁺][Cl]²)_in and involvement of Na⁺-K⁺-Cl cotransport likely. The other possibility is an increased Na⁺ influx via the Na⁺/H⁺ exchanger, similarly coupled with increased sodium-pump activity (36). Further experiments, particularly with the use of appropriate inhibitors, will be required to check these possibilities.

Induction of transport activities. The primary response to hypertonicity in endothelial cells derived from saphenous veins appears to be the induction of System A and the subsequent accumulation of amino acids, particularly glutamine and proline (12). Our results on the accumulation of NPS, which occurs in parallel with induction of System A activity, indicates that the accumulation of amino acids as compatible osmolytes is a secondary response, with maximum accumulation coinciding with the gradual loss of excess K⁺ (Figs. 3 and 4). Cell volume had already been restored at that stage, so that the increased intracellular accumulation of neutral amino acids presumably enables the excess salt content to be reduced, although the exact mechanism by which K⁺ leaves the cells is uncertain. The cellular concentration of NPS obtained, however, was only ~70 mM, insufficient to account for the extra external osmolarity of 0.2 osM, so some other solute must also have been accumulated. Hypertonic induction of BGT-1 occurred concomitantly with the downregulation of System A, and cells exposed to hypertonic medium containing 100 µM betaine did accumulate it, but only to 20-30 mM (Fig. 10). Hence, although any betaine present in the usual experimental media is likely to have been used as a compatible osmolyte, there would not have been enough to make up the osmotic imbalance noted above. The other likely contributions are from a similarly induced uptake of other compatible osmolytes, such as taurine or myo-inositol, or an induced synthesis of sorbitol (51).

Regulation. Regulatory sequences similar to Ton-E have been discovered in the promoter of the aldose reductase gene of several species (21) and it seems likely that other similar regulatory elements exist in connection with many cellular responses to increased osmolarity. There is some evidence that a regulatory protein is involved in the osmotic induction of System A activity (46); studies with inhibitors indicated that p38 MAP kinase is involved in the induction of BGT-1 in MDCK cells (47) as well as human monocytes and macrophages (16). Otherwise, little is known at present about the cellular mechanisms that specifically induce System A and BGT-1 activities in cells exposed to hypertonicity. The way in which added betaine (or GABA) inhibits the induction of the transport activity is also unclear, although it presumably involves specific binding of the betaine to some intracellular regulatory molecule.