Pflügers Arch - Eur J Physiol (2001) 442:771–781 DOI 10.1007/s004240100601

O R I G I N A L A RT I C L E

Andrew C. Hall · Peter G. Bush

The role of a swelling-activated taurine transport pathway in the regulation of articular chondrocyte volume

Received: 26 September 2000 / Revised: 5 February 2001 / Accepted: 12 April 2001 / Published online: 1 June 2001 © Springer-Verlag 2001

Abstract Swelling articular chondrocytes by reducing osmolarity stimulates a taurine transport pathway, which is implicated in regulatory volume decrease (RVD) in various cell types. The present study investigated factors controlling the activity of this pathway in chondrocytes, in particular (1) the effects of the acute (seconds) and chronic (hours) exposure of chondrocytes to anisotonic media, and (2) whether there is a role for metabolites from the arachidonic acid cascade in activating the taurine transport pathway. For in situ and isolated chondrocytes, the point at which swelling-activated [14C]taurine efflux was stimulated (the “set-point”) corresponded closely to the osmolarity of the incubation medium (180, 280 or 380 mosmol/l). However, the volume of chondrocytes isolated into these media and measured by confocal microscopy was not different (≅645 µm3). Activity of the swelling-activated taurine transport pathway was inhibited by REV5901 (an inhibitor of steps of the arachidonic acid cascade; K0.5 8±4 µM), NDGA (a general lipoxygenase inhibitor; K0.5 28±5 µM), or MK886 (an inhibitor of the 5-lipoxygenase-activating protein; 91% inhibition at 10 µM), but weakly by the more potent 5-lipoxygenase inhibitor REV5901 para (K0.5 350±100 µM). Addition of the leukotriene (LT) B4 or D4 receptor antagonists, CP-105,696 and L660,711 respectively, or of the leukotrienes LTB4, LTC4, LTD4 and LTE4 or lipoxins (hepoxylin A3 or B3) had no effect on the activity of the pathway in isotonic or hypotonic media. The role of the pathway in RVD was determined in isolated calceinloaded chondrocytes using fluorescence imaging. RVD was observed and inhibited by REV5901 (50 µM) and by NDGA (75 µM). The data show that despite chronic exposure of chondrocytes to anisotonic media, the cells maintain a pre-determined volume that is the “set-point” for the activation of the taurine transport pathway folA.C. Hall (✉) · P.G. Bush Department of Biomedical Sciences (Physiology), University Medical School, Hugh Robson Building, George Square, Edinburgh EH8 9XD, Scotland, UK e-mail: [email protected] Tel.: +44-131-6503263, Fax: +44-131-6506527

lowing acute hypotonic challenge. This pathway appears to play a role in chondrocyte RVD, but its activation does not involve metabolites of the arachidonic acid cascade. Keywords Cell swelling · Cell volume regulation · Chondrocyte · Organic osmolyte · Regulatory volume decrease · Volume-sensitive osmolyte transport pathway

Introduction The proteoglycans surrounding the chondrocytes of articular cartilage carry a high density of immobile negative charges, resulting in high concentrations of free cations (e.g. Na+, 250–350 mM) and low concentrations of free anions (e.g. Cl– 60–90 mM [10, 24, 26, 37]). The resulting interstitial osmolarity ranges over 350 to 450 mosmol/l, with the exact value depending on the local proteoglycan concentration, thus it is very different from synovial fluid/plasma (≅280 mosmol/l). Chondrocytes experience changes to their physico-chemical environment both physiologically (e.g. during static load leading to fluid expression) and pathophysiologically (i.e. in osteoarthrosis), where cartilage swelling is an early event resulting in decreased tissue osmolarity [25, 33, 34]. Cartilage resilience is reduced, and the cells are thus exposed to greater changes in osmolarity/ionic content during normal loading compared to healthy, resilient cartilage. Mechanical load, by its effects on chondrocyte metabolism, can influence the material properties of articular cartilage, principally by altering the proteoglycan concentration. In normal joints, load-bearing areas are thicker, have a higher proteoglycan concentration and are mechanically stronger than non-load-bearing regions of the same joint [36]. However, the signals from the complex and changing physico-chemical environment to which chondrocytes respond in normal cartilage and the cellular disturbances occurring in osteoarthrosis, where matrix catabolism exceeds anabolism, are not understood.

772

Matrix metabolism by isolated or in situ chondrocytes is sensitive to changes in interstitial composition and osmolarity [37]. Thus, the changes to matrix metabolism observed during loading and in early osteoarthrosis might result from perturbations to chondrocyte volume/intracellular composition. To clarify the relationship between extracellular environment and matrix metabolism, we have studied the response of chondrocytes to acute and chronic exposure to hypotonicity, focusing on the transport pathways involved in regulatory volume decrease (RVD; [11, 14, 15, 16, 35]). This is stimulated following hypotonicity, and leads to osmolyte loss and recovery of cell volume. Several systems are involved, but in various cell types, including chondrocytes, an osmolyte transport pathway with broad substrate specificity plays a significant role [9, 19]. Some of its characteristics are similar to those of other cell types (e.g. C6 glioma cells [35]; ascites tumour cells [20, 21, 22]). For Chinese hamster ovary cells, a key role for intracellular ionic strength (µi) has been proposed for the activation of the osmolyte pathway [5, 6], whereas in Ehrlich ascites tumour cells, the arachidonic acid cascade and leukotriene LTD4 release is important [14, 21]. Here, we have determined some of the factors involved in regulating the swelling-activated taurine transport pathway of chondrocytes within (i.e. in situ), and isolated from the matrix following a chronic (hours) and an acute (seconds) osmotic perturbation. We have incubated cartilage explants or chondrocytes in media of varying osmolarity (180, 280 or 380 mosmol/l; “chronic exposure”). The cells were then subjected to a hypotonic challenge (“acute exposure”) and the activity of the transport pathway determined by assessing [14C]taurine release. The point at which the pathway was stimulated (the “set-point”) was close to the osmolarity used for the chronic incubation. The volume of chondrocytes isolated at these osmolarities was the same, suggesting that they regulate their volume to a pre-determined value and that this controls the “set-point” of the taurine transport pathway. We have also attempted to clarify whether metabolites from the arachidonic acid cascade (leukotrienes, lipoxins) are involved in stimulating the swelling-activated taurine transport pathway. The data do not show a role for this system in activating the taurine transport pathway, and suggest that the inhibitory effects of the drugs studied are relatively non-specific.

Materials and methods Solutions and chemicals Dulbecco’s modified Eagle’s medium (DMEM) including HEPES (10 mM) with penicillin (50 units·ml–1) and streptomycin (25 µg·ml–1; pH 7.4; 280 mosmol/kg H2O) was used as the standard sterile medium for cartilage explant culture. For radiotracer flux and volume measurements, DMEM (without pen/strep) was used, with osmolarity varied using NaCl or buffered purified water (10 mM HEPES; pH 7.4) as required. REV5901 [α-pentyl3-(2-quinolinylmethoxy)benzenemethanol; L-656,323 using the

Merck nomenclature] and REV5901 para-isomer [α-pentyl-4(2-quinolinylmethoxy)benzenemethanol; L-655,238], inhibitors of steps of the arachidonic acid cascade [7], were purchased from Calbiochem-Novabiochem, Nottingham, UK. NDGA (nordihydroguaiaretic acid), MK886 {3-[1-p-chlorobenzyl-5-(isopropyl)-3-tbutyl-thioindol-2-yl]-2,2-dimethyl-propanoic acid Na}, the leukotrienes (LTB4, LTC4, LTD4, LTE4), the lipoxins (hepoxylin A3, hepoxylin B3) and bumetanide were obtained from Sigma, Poole, UK, or from Calbiochem-Novabiochem. L660,711 (or MK-571), (3-{3-[2-(7-chloro-2-quinolinyl)ethenyl]phenyl [(3-dimethyl amino-3-oxo propyl)thio]methyl}thio) propanoic acid was a gift from Dr. R.N. Young (Merck Frosst, Canada) and CP-105,696; (+)-1(3S,4R)-[3-(4-phenylbenzyl)-4-hydroxychroman-7-yl] cyclo pentane carboxylic acid was a gift from Dr. H.J. Showell (Pfizer, Groton). The osmolarities of media used for cell isolation and osmotic challenges were measured using a Wescor vapour pressure osmometer to ±2 mosmol/l. The pH of all solutions used was 7.4±0.1 at 37°C. Cartilage preparation and chondrocyte isolation Articular cartilage from load-bearing areas of the metacarpal-phalangeal joint was obtained from cows’ feet collected fresh from the local abattoir. For one experiment, macroscopically normal human cartilage with no evidence of surface fibrillation was removed (with ethical approval) from the load-bearing surface of the femoral condyle obtained from an above-knee amputation (male 55 years old). The techniques for chondrocyte isolation from bovine cartilage explants were as described elsewhere [12]. Chondrocyte viability was tested using Trypan blue exclusion [12] following isolation into media of differing osmolarity and was routinely >95% viable. [14C]Taurine (NEN, Hounslow, UK) was added to the chondrocyte isolation medium to give ≅1 kBq·ml–1. For efflux experiments from cartilage, explants were incubated (≅20 h, 37°C) with [14C]taurine (≅2 kBq·ml–1) in media with osmolarity above (380 mosmol/l), below (180 mosmol/l) or the same as that used for explant culture (280 mosmol/l). For efflux experiments from isolated chondrocytes, [14C]taurine (≅1 kBq·ml–1) was added to the isolation medium. The osmolarity of media used for cartilage culture, chondrocyte isolations and for subsequent incubations was varied by adding NaCl or buffered water as required. Cartilage was used for experiments within 2 days following removal from the joint and after incubation with the radio-label; cartilage/chondrocytes were used within 2 h. Radiotracer efflux experiments from cartilage and isolated chondrocytes The explants were washed (6× over 2 h) in isotope-free DMEM at the appropriate osmolarity after which the radioactivity of the culture medium was reduced to background. The cartilage pieces were then quickly placed in a multiwell plate with wells containing DMEM at a range of osmolarities equilibrated at 37°C. After 5 min of incubation, during which the medium was mixed, samples of medium were removed and placed in vials with scintillation fluid, and radioactivity determined. The total radioactivity at the start of the experiment after cartilage washing was measured in separate explants by digesting tissue in filtered papain solution (3 h at 65°C [10]). The radioactivity measured was taken as the initial (100%) value, and the tracer released into the medium over the 5-min period expressed as % loss in cpm·mg–1 cartilage. It was important to determine non-specific (i.e. extracellular) [14C]taurine binding to cartilage. Accordingly, explants were heat-treated (water at 100°C; 2 min) to kill chondrocytes, and incubated (280 mosmol/l) with radio-label for the same duration and concentration as for living cartilage. The explants were then washed and total radioactivity determined. Non-specific incorporation accounted for 8±3% (data are mean ±SEM from cartilage of three feet) of total radioactivity, and this value was unchanged in hypertonic or hypotonic solutions (data not shown). Cartilage weights

773 were expressed per mg wet weight by weighing the samples wet and when dried to constant weight (60°C for 18 h). Cartilage hydration was also studied as a function of medium osmolarity, but there was no significant change over the range of osmolarities investigated (180 to 380 mosmol/l; AC Hall, unpublished). For [14C]taurine efflux experiments from isolated chondrocytes, the labelled cells were quickly filtered and washed in DMEM of the appropriate osmolarity used for the isolation using standard techniques as described [12]. Flux experiments were performed at 37°C using microcentrifuge tubes containing the appropriate agents with a cell density of ≅106 cells/ml. Chondrocyte suspensions were pre-incubated with drugs at 37°C for 5 min (380 mosmol/l), then added to isotonic or hypotonic solutions in the presence/absence of drugs as shown (see figure legends), and the flux period commenced. For experiments studying leukotrienes or lipoxins, the agents were present in Eppendorf tubes with the anisotonic media, and the flux started by rapidly adding the cell suspension. At the end of the incubation period (5 min), the tubes were centrifuged (10,000 g, 10 s) and aliquots of supernatant taken for scintillation counting. Samples were taken to determine the total intracellular radioactivity at the beginning of the flux period. The initial (100%) value was determined by counting the total radioactivity in aliquots of cell suspension. Correction was also made for the small amount of extracellular radioactivity present at the start of the flux period. The data were expressed as % loss of total intracellular radiotracer over 5 min, and normalized against the values obtained at 380 mosmol/l for isolated chondrocytes and 280 mosmol/l (medium osmolarity) for cartilage (i.e. equivalent extracellular osmolarities for isolated and in situ chondrocytes). Control time course experiments established that [14C]taurine efflux from cartilage and isolated chondrocytes suspended in isotonic and hypotonic solutions followed first-order kinetics for at least 8 min at 37°C (data not shown). Calculation of extracellular osmolarity of in situ chondrocytes The immobile negative charges on the proteoglycans of cartilage markedly alter the extracellular (i.e. interstitial) osmolarity surrounding chondrocytes compared to the medium osmolarity. For one series of experiments (Fig. 1), therefore, it was important to estimate the extracellular osmolarity to which chondrocytes were exposed. The ion distributions within cartilage fluid, and hence the osmolarity, can be determined using Gibbs–Donnan equilibrium conditions and published data [24, 26, 37]. The effective fixed charge density (X) has been measured as 0.18 mEq/g H2O for bovine cartilage explants as used here, and this takes account of cartilage hydration, the fraction of intrafibrillar space and the collagen content [37]. The average molal partition coefficient of Na+ in DMEM was taken as 2 [24, 37]. It should be noted that these are mean values and the fixed charge density around some chondrocytes, particularly in the superficial zone of cartilage (where proteoglycan content is relatively low [26]), will be lower than this. Using these values, the average [Na+] around the chondrocyte (i.e. within the interstitial fluid or the extrafibrillar concentration, [Na+]o) can then be estimated from: [Na+]o=(K·Mf-F·C)m/(Mf-F·C)

(1)

where K is the molal partition coefficient (which for NaCl is 2), F is the fraction of water within the intrafibrillar space (1 g/g dry collagen), Mf is the cartilage hydration (2.72 g H2O/g dry wt), C is the collagen content (0.41 g collagen/g dry tissue) and m is the Na+ concentration in the tissue culture medium. The partition coefficients for the other components of DMEM (amino acids, glucose, HEPES, etc.) were taken to be 1, and for Cl– to be 0.6 [26], for explants of bovine articular cartilage in DMEM (280 mosmol/l) as used in this study, the calculated osmolarity of the extracellular compartment was ≅380 mosmol/l. The extracellular osmolarities were estimated in the other media using this method and the resulting values agreed with experimental values previously published [26, 37].

Fig. 1 Stimulation of taurine efflux from cartilage explants exposed to anisotonic solutions following culture at various osmolarities. Bovine articular cartilage explants were cultured at 180 (▲ ▲ ), 280 (● ● ) or 380 (■ ■ ) mosmol/l DMEM in the presence of [14C]taurine for 20 h (see Materials and methods). The explants were then rinsed in label-free DMEM of the same osmolarity, and quickly placed in vials containing a range of osmolarities (varied using NaCl or buffered water as required), and the release of label determined over 5 min. The broken lines indicate the data corresponding to the incubation osmolarities at 180, 280 and 380 mosmol/l. The upper horizontal axis shows calculated extracellular (i.e. interstitial) osmolarities determined as described (see Materials and methods). Results are means ±SEM for triplicate determinations for at least four separate experiments at each condition Measurement of isolated chondrocyte volume Chondrocyte volume was determined by confocal scanning laser microscopy on cells lightly attached to glass coverslips (60 min, 37°C) in the presence of calcein-AM (5 µM, 30 min, 37°C; Molecular Probes, Eugene Ore., USA). An upright microscope with water-immersion objective (63×) was used, with optical sections being taken in 1 µm z-axis increments. Data were collected and volume computations performed using Voxelshop Pro (Bitplane, Zurich, Switzerland) on a Silicon Graphics O2 workstation. Volume calibrations were carried out using fluorescent beads (Polysciences, Warrington, Pa., USA) of known diameter (10 and 25 µm). For control experiments to test the technique, chondrocytes were incubated at 20°C in DMEM (280 mosmol/l) in the presence of REV5901 and bumetanide (at 75 and 50 µM respectively) for 30 min. This was to block any volume regulatory behaviour that might occur. The cells were then exposed to 180, 280 or 380 mosmol/l (with inhibitors present) and cell volume measured. The data showed that the technique could accurately record changes to cell volume (see Table 1). Measurement of chondrocyte volume regulation by fluorescence imaging Isolated chondrocytes were suspended in DMEM (380 mosmol/l) at ≅0.2×106 cells·ml–1. Aliquots (1 ml) were pipetted onto clean coverslips (22 mm diameter No. 0; Merck) supported by a stain-

774 Table 1 The volume of bovine articular chondrocytes isolated from cartilage at different osmolarities. Chondrocytes were isolated into DMEM of 280 mosmol/l and their volume determined by confocal microscopy (see Materials and methods). Aliquots of cells were also exposed to 180 or 380 mosmol/l DMEM for 2.5 min containing REV5901 and bumetanide (at 75 and 50 µM respectively) to block volume regulatory behaviour (see Materials and methods). In parallel, chondrocytes were also isolated into DMEM of 180, 280 or 380 mosmol/l over 20 h, and their volumes measured. There was a significant change from the resting volume (measured in 280 mosmol/l) for cells exposed to an acute challenge of 180 or 380 mosmol/l (P<0.001), whereas there was no significant difference between the volumes of chondrocytes isolated into 180, 280 or 380 mosmol/l (P>0.05). Results (means ±SEM) were from at least 28 cells at each condition, with data being obtained from at least separate four experiments Medium osmolarity (mosmol/l) 180 280 380

Chondrocyte volume (µm3) 2.5 min

20 h

728±29** 630±32 541±24**

653±39 637±22 645±25

less steel ring, and placed in a humidified atmosphere (37°C). Thirty minutes before starting volume measurements, the fluorescent dye calcein-AM was added. An aliquot of the stock solution (5 mM in DMSO) was added to DMEM and then to the chondrocyte suspension, giving a final concentration of 5 µM and DMSO concentration of ≅0.1%. The coverslips were then gently rinsed with DMEM to remove unattached cells, cell clumps and unincorporated dye, and placed on the temperature-controlled stage of an inverted microscope (Nikon Diaphot, model TMD). The cells were initially viewed using a 40× CF fluor objective (numerical aperture=1.3) under transmitted light to select an area of sufficiently high density of isolated chondrocytes. Only single, isolated chondrocytes were investigated. The cells were then perfused with DMEM (380 mosmol/l, 37°C; 4 ml/min) and the fluorescent dye excited (EX=495 nm) and emission (EM=535 nm) observed using ImageMaster Pro™ software (Photon Technology International, New Jersey). The photometric output from images was determined by defining regions of interest around single chondrocytes. Images were collected at 0.33 Hz and background subtraction performed using values obtained from unloaded chondrocytes. The times of addition of the fluorescent dye were staggered to avoid the chondrocytes being incubated with dye for more than 30 min before experimentation. The principle of the technique was that changes to the intracellular fluorophore concentration resulting from concentration/dilution of the cell contents, arising from water loss/gain respectively, would be reflected in changes in fluorescence intensity [1]. Thus, as the cell water content changed as a result of exposure of cells to anisotonic media and volume regulatory changes, the fluorescence intensity would vary inversely. Since the total fluorescence of the whole cell would not change during alterations to cell volume, it was necessary to record from a relatively small region of the loaded cell. This was achieved using an objective lens with high numerical aperture, and focusing at a plane where the fluorescence intensity was near maximal, which was close to the axial centre of the cell [1]. Data presentation Unless otherwise indicated, data were from separate experiments (mean ±SEM) on cartilage/chondrocytes obtained from at least four feet of different animals. The means of the experimental groups were compared using Student’s unpaired t-tests, and oneway analysis of variance (ANOVA). Least-squares linear regression analysis was performed on independent and dependent vari-

ables and slopes compared by unpaired t-tests. For all tests, the SigmaStat programme (Jandel Scientific) was used, with P<0.05 accepted as a significant difference.

Results Stimulation of 14C-taurine efflux from bovine articular cartilage explants cultured chronically at various osmolarities by acute osmotic challenge Figure 1 shows the increase in taurine efflux from chondrocytes in situ following hypotonic challenge, after preincubation of cartilage explants in media of 180, 280 or 380 mosmol/l. The data indicate that the osmolarities at which efflux was stimulated significantly (the “set point”) were close to those of the solutions used during the pre-incubation of the cartilage with the radio-label. Thus, significant increases in taurine efflux were observed at 160, 260 and 340 mosmol/l for cartilage incubated in 180, 280 and 380 mosmol/l respectively (Fig. 1). Also shown are the calculated extracellular osmolarities (see Materials and methods), which, because of matrix proteoglycans, raise the osmolarity surrounding in situ chondrocytes. Increasing the osmolarities above those used for cartilage pre-incubation appeared to have a small but not significant inhibitory effect on the rate of tracer efflux, except for cartilage incubated in 180 mosmol/l, and exposed to 260 mosmol/l, where there was significant inhibition (Fig. 1). Stimulation of [14C]taurine efflux from isolated chondrocytes isolated at various osmolarities by an acute reduction in osmolarity Chondrocytes were isolated from cartilage explants incubated at 180, 280 or 380 mosmol/l, with [14C]taurine present (see Materials and methods). The efflux of tracer was then determined with medium osmolarity varying above and below the osmolarity in which the cells were isolated (Fig. 2). As with cartilage explants (Fig. 1) for all conditions, reducing osmolarities below those used for cell isolation, significantly increased efflux (Fig. 2). Significant increases in taurine efflux were observed at 170, 250 and 360 mosmol/l for chondrocytes incubated in 180, 280 and 380 mosmol/l respectively (Fig. 2). The increase in taurine efflux was relatively sensitive to a reduction in osmotic pressure. A decrease in osmolarity from 380 to 360 mosmol/l (a reduction of ≅5%) significantly stimulated taurine efflux from 2.3±0.4 to 3.5±0.2% loss of taurine over 5 min (data are mean ±SEM, for n=3), an absolute increase of ≅34%. Increasing osmolarity above that used for chondrocyte isolation had no effect on the rate of tracer efflux. Thus, the results with isolated chondrocytes paralleled those obtained with cartilage (Fig. 2) since the stimulation of efflux was observed as osmolarity was reduced below that used for cartilage or chondrocyte incubation.

775

380 to 320 mosmol/l raised taurine efflux from 2.0±0.4 to 7.6±0.8% efflux over 5 min (by about 3.8×, data from Fig. 2). Thus, the data suggest that the magnitude of the increase in taurine efflux was similar and that there was no marked difference in the sensitivity of the taurine transport pathway between in situ and isolated chondrocytes exposed to similar osmotic challenges. The effect of an even longer incubation of cartilage in anisotonic media on the subsequent response of the taurine transport pathway was also studied. Cartilage was incubated for 7 days in 180, 280 or 380 mosmol/l DMEM. The chondrocytes were then isolated at these osmolarities in the presence of [14C]taurine, and efflux measured as a function of extracellular osmolarity to determine the “set-point” of swelling-stimulated taurine efflux (see Materials and methods). The osmolarity at which taurine efflux was enhanced was almost exactly (±5 mosmol/l) the same as the osmolarity used for chondrocyte isolation (data not shown). Fig. 2 The increase of taurine efflux from chondrocytes exposed to anisotonic solutions following isolation at various osmolarities. Bovine articular chondrocytes were isolated at 180 (▲ ▲ ), 280 (● ●) or 380 (■ ■ ) mosmol/l DMEM in the presence of [14C]taurine as described (see Materials and methods). Cells were then washed free of extracellular label at the appropriate osmolarity, and quickly suspended in DMEM of varying osmolarity, and the release of [14C]taurine determined. The broken lines indicate the data corresponding to the osmolarities of the isolating media of 180, 280 and 380 mosmol/l. Results are means ±SEM for triplicate determinations on at least four experiments at each condition

The slopes of the relationships between medium osmolarity and the significant increases in taurine efflux above the “set-point” osmolarities were compared for cells isolated in media of different osmolarity (Fig. 2). The values were as follows: 180 mosmol/l, –0.511± 0.102 (178–150 mosmol/l; r=–0.962); 280 mosmol/l, –0.256±0.040 (238–188 mosmol/l, r=–0.979); and 380 mosmol/l, –0.138±0.020 (323–260 mosmol/l, r=–0.983) [data are slopes (% loss of taurine over 5 min/mosmol)] for cells isolated at the osmolarities indicated (Fig. 2). These were significantly different from each other, suggesting that the sensitivity of the taurine transport pathway increased for a given change in osmotic challenge as the isolation osmolarity was reduced. There were no significant differences in the comparable slopes for the data obtained for cartilage explants (Fig. 1) and this appeared to be because of greater variation in the data and fewer experimental points (data not shown). It was also of interest to observe the response of the taurine transport pathway of isolated and in situ chondrocytes at comparable extracellular osmolarities. Thus for chondrocytes in situ at a calculated extracellular osmolarity of 380 mosmol/l (i.e. medium osmolarity of 280 mosmol/l, Fig. 1), a reduction in osmolarity to 320 mosmol/l (medium osmolarity ≅200 mosmol/l) increased taurine efflux over 5 min from 3.0±0.5 to 12.5±2.2% (by ≅4.2×, data from Fig. 1). For isolated chondrocytes, a reduction in medium osmolarity from

Volume of chondrocytes isolated at different osmolarities In parallel experiments to those studying [14C]taurine release as a function of isolation osmolarity, the volume of chondrocytes was measured using confocal microscopy. It was important first of all to confirm that the technique could accurately measure resting volumes, and those following changes in medium osmolarity. This was tested by isolating chondrocytes at 280 mosmol/l, then exposing them to osmotic challenges of 180 or 380 mosmol/l, which after 2.5 min significantly increased or decreased cell volume respectively (Table 1). It should be noted that these experiments were performed at 21°C in the presence of REV5901 and bumetanide (75 µM and 50 µM respectively) to limit volume-regulatory behaviour [3]. These measurements confirmed that the technique could detect changes in chondrocyte volume. However, when the volume of chondrocytes isolated over 20 h at 180 and 280 (both hypotonic solutions relative to in situ osmolarity) or 380 mosmol/l (isotonic) were compared, there was no significant difference (Table 1). This suggests that, despite the large differences in osmolarity and hence volume experienced by chondrocytes during their release from the matrix, chondrocyte volume eventually stabilizes at a similar value (≅645 µm3). Hypotonically stimulated taurine efflux from cartilage explants and its inhibition by REV5901 It was important to check that swelling-activated taurine release from chondrocytes of cartilage explants (Fig. 1) could also be blocked pharmacologically (see Table 2). Reducing osmolarity from 280 mosmol/l to 180 mosmol/l increased the rate of [14C]taurine loss from cartilage explants whereas raising osmolarity to 380 mosmol/l had no significant effect. The taurine release from

776 Table 2 Inhibition of swelling-activated taurine efflux from cartilage explants by REV5901. Bovine cartilage was incubated for 20 h in 280 mosmol/l DMEM with [14C]taurine as described (see Materials and methods). Some of the explants were then incubated for a further 2 h with REV5901 (or equivalent volume of DMSO) at the concentrations indicated. The cartilage pieces were then washed in tracer-free medium, and transferred to the solutions indicated, and taurine release measured (as a % loss over 5 min) and expressed per mg cartilage (see Materials and methods for details). Tests for significance were performed by comparing values for taurine release under the conditions shown, with that measured in 280 mosmol/l. Data (mean ±SEM) are from four experiments Condition

Taurine release (%) (cpm/mg)

380 mosmol/l 280 mosmol/l 180 mosmol/l 180 mosmol/l + REV5901(500 µM)

1.8±0.6 (N/S) 3.1±0.5 22.2±3.0** 2.2±0.5 (N/S)

explants pre-treated with REV5901 and then subjected to osmotic challenge was totally blocked at 500 µM. A higher concentration of the drug was used (see below) to compensate for any reduced permeation of REV5901 into the matrix. In a single experiment, reducing medium osmolarity (280 to 180 mosmol/l) also stimulated the release of [14C]taurine from normal human articular cartilage explants (Fig. 3). The extent of the increase (by ≅5×) was similar to that for bovine cartilage (≅7×; Table 2). This suggests that chondrocytes in situ within normal bovine and human cartilage are osmotically sensitive, and respond to a reduction in osmolarity by the activation of the taurine transport pathway. Effects of various agents on the chondrocyte taurine transport pathway Previous work suggested that metabolites from the arachidonic acid cascade play a key role in mediating volume regulation in various cell types, including ascites tumour cells (LTD4 [14]) and platelets (hepoxylin A3 [23]). However, in some cell types (e.g. C6 glioma cells), lipoxygenase products of arachidonic acid metabolism are probably not involved, and inhibitory effects of drugs (e.g. ketoconazole) on osmolyte efflux could arise from the relatively non-specific block of the pathway, rather than by inhibition of secondary messenger pathways [27, 35]. To clarify whether metabolites from the arachidonic acid cascade were involved in the activation of the chondrocyte taurine transport pathway, agents were tested for their ability to block or stimulate the pathway (Table 3 and see below). The LTD4 receptor antagonist and weak inhibitor of 5-lipoxygenase, REV5901, was found to be a relatively effective inhibitor of swelling-stimulated taurine release from isolated chondrocytes (Table 3). The time course of inhibition of the swelling-activated taurine transport pathway by REV5901 was tested by comparing the inhibition measured when the drug was (1) included in the

Fig. 3 The increase in taurine efflux from human articular cartilage subjected to osmotic challenge. Explants of human cartilage were incubated in DMEM 280 mosmol/l in the presence of [14C]taurine. The explants were then rinsed in label-free DMEM (280 mosmol/l), and then quickly placed in vials containing a range of osmolarities and release of [14C]taurine determined over 5 min (see Materials and methods). Results are means ±SD for sextuplicate determinations on one experiment

solution used to deliver the acute hypotonic stress to [14C]taurine-labelled chondrocytes, or (2) present (at the same concentration, 50 µM) for 5 min during the preincubation in isotonic solution (380 mosmol/l) before the hypotonic challenge. The control swelling-activated flux was 4.7±0.4 and 13.2±0.5% taurine loss over 5 min in 380 or 280 mosmol/l respectively, giving a volume-activated flux of 8.5±0.6% (data are % taurine efflux over 5 min; means ±SEM for n=3 experiments). Pre-treatment of chondrocytes with REV5901 reduced the swellingactivated flux to 7.9±0.2% of the maximum value, an inhibition of ≅93%, whereas without drug pre-treatment the flux was reduced to 7.6±0.4%, an inhibition of ≅89% (data are % taurine efflux over 5 min; means (±SEM for n=4 experiments). In other words, it did not matter if the REV5901 was present during the pre-incubation prior to osmotic challenge, or if the drug was present in the hypotonic solution itself, since in both cases the inhibition was essentially complete, and identical. This result suggested that the inhibition by REV5901 was more rapid than chondrocyte swelling and activation of the taurine transport pathway. The para-isomer of REV5901, which is a more potent inhibitor of 5-lipoxygenase than REV5901 (by ≅20× [7]), was a poor inhibitor of the pathway (Table 3, Fig. 4). However, the general lipoxygenase inhibitor NDGA (at 100 µM) almost completely blocked the pathway, and separate dose–response experiments gave a K0.5 of 28±5 µM (mean ±SEM, n=3). Concentrations above 100 µM increased non-specific radiotracer flux, probably as a result of membrane damage, and gave an apparent

777

Fig. 4 Differential effect of REV5901 and REV5901 para-isomer on swelling-stimulated [14C]taurine efflux from isolated bovine articular chondrocytes. Volume-activated radio-label loss was measured in the presence or absence of the drugs at the concentrations shown, and estimates of K0.5 made from each experiment. Data (mean ±SD) were from one experiment with pooled values given in Table 2

Experiments studying the effects of various leukotrienes (LTB4, LTC4, LTD4, LTE4; 4 µM each) and lipoxins (hepoxylin A3 or B3; 20 µM each) on [14C]taurine efflux showed no significant stimulation (P>0.05) of basal efflux in isotonic DMEM (380 mosmol/l), nor any change to the swelling-stimulated efflux (% loss over 5 min at 300 mosmol/l – 380 mosmol/l; data not shown). It has been reported that there is a complex dose–response relationship for the effect of LTD4 such that low doses stimulate the taurine transport pathway in ascites tumour cells, whereas higher doses inhibit it [14]. However, varying the LTD4 concentration (0.05–4 µM) did not stimulate taurine efflux (not shown). Experiments were also performed on chondrocytes pre-treated for 5 min with NDGA (50 µM) or pimozide (10 µM) to block any endogenous production of leukotrienes [21]. There was no change in % taurine efflux over 5 min in isotonic (380 mosmol/l) or hypotonic (280 mosmol/l) DMEM following the addition of LTD4 (0.4 or 4 µM; not shown). Response of chondrocytes to hypotonicity measured by fluorescence imaging Relationship between fluorescence intensity, medium osmolarity and cell volume

increase in the swelling-activated component. Interestingly, MK886 (10 µM), a more potent inhibitor of 5-lipoxygenase than REV5901 and which binds to the 5-lipoxygenase-activating protein [7], reduced the volume-sensitive flux (Table 3). In contrast, specific antagonists of the LTB4 or LTD4 receptors, CP-105,696 and L660,711 respectively, tested at 20 µM had no effect on swelling-stimulated taurine efflux, strongly suggesting that these receptors are not involved in the signal transduction process for the activation of the taurine transport pathway (Table 3).

It was important first of all to confirm that the fluorescence imaging technique could qualitatively record the volume and volume-regulatory capacity of chondrocytes when perfused with various anisotonic media in the presence/absence of inhibitors. Control experiments were performed using calcein-loaded chondrocytes, perfusing them with DMEM at 380 mosmol/l (i.e. control) medium and measuring fluorescence (fo). The osmolarity was then varied over 220 to 480 mosmol/l, and fluorescence intensity (ft) measured, and expressed as a ratio of the control fluorescence (ft/fo). For these measurements to be

Table 3 A summary of the drugs studied on the swelling-activated chondrocyte taurine transport pathway, some of their possible sites of action and their effect. Cells were incubated for up to 10 min in isotonic DMEM (380 mosmol/l) before delivery of hypotonic shock (280 mosmol/l) and initiation of the efflux experiment. The control (380 mosmol/l) taurine efflux was 3.3±0.4% (% taurine efflux over 5 min) and was raised to 9.2±0.8% at 280 mosmol/l, an

increase of ≅2.8× and a swelling-stimulated efflux (i.e. flux at 280 mosmol/l – flux at 380 mosmol/l) of 5.9±0.6% (% taurine efflux over 5 min; n=12). The swelling-activated flux was then measured in the presence of the drugs as shown at both osmolarities and the value expressed as a percentage of the control flux. Data are given as means ±SEM for n separate experiments. (References are given in square brackets)

Drug

Proposed site(s) of action

Effect

REV5901

Competitive antagonist of LTD4 receptor, also inhibits 5-lipoxygenase [30]

Inhibitory (K0.5 8±4 µM; n=4) 93±5% inhibition at 100 µM (P<0.001; n=3)

REV5901 para-isomer

Inhibits 5-lipoxygenase [7]

Inhibitory (K0.5 350±100 µM; n=4) 1 5±8% inhibition at 100 µM (P<0.05; n=3)

NDGA

Inhibitor of 5-, 12- and 15-lipoxygenase [20, 21, 22]

Inhibitory (K0.5 28±5 µM; n=4) 83±5% inhibition at 100 µM (P<0.001; n=3)

MK886

Inhibitor of 5-lipoxygenase-activating protein [7]

91±3% inhibition at 10 µM (P<0.001; n=3)

L660,711 (MK-571)

Specific LTD4 receptor antagonist [18]

2±5% inhibition at 20 µM (n=3; N/S)

CP-105,696

Specific LTB4 receptor antagonist [8]

6±5% inhibition at 20 µM (n=3; N/S)

778

accurate, it was necessary to inhibit any volume-regulatory capacity. For RVD, which occurred in hypotonic solutions, inhibition was achieved by adding NDGA or REV5901 (75 µM), and for any RVI that might occur,

bumetanide (50 µM) was present. Over 220–480 mosmol/l, the data were well-described by a linear relationship in the form y=mx+c as ft/fo=8.92×10-4·medium osmolarity + 0.6588 with standard errors on the slope and intercepts of 4.96×10–5 and 0.017 respectively. The regression coefficient (r) was 0.995±0.008 (data were obtained from five separate experiments on at least 20 cells/osmolarity). It should be remembered that it is possible that only a fraction of the intracellular dye is osmotically sensitive, and in these experiments it was estimated to be ≅40% (data not shown). In other cell types, it has been estimated that ≅35% of the intracellular calcein behaves as if it was sensitive to changes in cell volume [1]. Volume regulation by isolated chondrocytes Subjecting isolated chondrocytes to osmotic challenge (380 to 220 mosmol/l) caused a rapid fall in fluorescence intensity (Fig. 5A). The rate of fall was influenced by the perfusion rate, and, although a rapid rate was desirable (to limit any RVD that might occur during this period), it caused chondrocyte detachment if it was excessive. In the absence of inhibitors, the fluorescence of individual cells returned towards initial values (i.e. those in 380 mosmol/l), although the rate of recovery varied between cells (Fig. 5A). Pooled data (Table 4) showed that, for most cells, recovery was complete within 6 min after hypotonic challenge, with an estimated t1/2 of 101±19 s (mean ±SEM, for n=6 experiments on 49 cells; range 40–200 s). In the presence of REV5901 (or NDGA) in the perfusion medium, the reduction in fluorescence in hypotonic solution was not significantly different compared to the absence of the drug (Table 4). The recovery in fluorescence resulting from RVD was, however, abolished by the drugs, such that after 6 min there was no difference compared to that following chondrocyte swelling after 1 min (Fig. 5B; Table 4). This, taken with the inhibitory effects of REV5901 and NDGA on volume-sensitive taurine efflux (Table 3), suggests that

Fig. 5A, B An example of an experiment showing the change in fluorescence intensity of isolated articular chondrocytes following a reduction in medium osmolarity. Fluorescence was recorded as described (see Materials and methods): A in six cells under control conditions (380 mosmol/l) and B in eight cells in the presence of REV5901 (50 µM, 380 mosmol/l). The value obtained was termed the initial fluorescence (fo). After about 1 min, the perfusion solution was changed to 220 mosmol/l DMEM (in the presence/absence of REV5901) which produced a corresponding rapid fall in fluorescence. The point at which fluorescence was at a minimum (≅1 min) corresponded to the maximum volume of the cells during hypotonic challenge and this value was designated ft. The ratio of this value to the initial value (ft/fo) represented the maximum initial cell swelling. The fluorescence was then determined after a standard time period (ft; 6 min after reduced medium osmolarity) and then expressed as a ratio of fo. Note the recovery of fluorescence in the control which was almost completely abolished when the chondrocytes were pre-treated with REV5901. (Pooled data are given in Table 4)

779 Table 4 Inhibition of regulatory volume decrease (RVD) in isolated articular chondrocytes by nordihydroguaiaretic acid (NDGA) or [α-pentyl-3-(2-quinolinylmethoxy)benzenemethanol; L-656,323 (REV5901). The fluorescence of single chondrocytes when perfused with 380 mosmol/l DMEM was recorded in the presence/absence of the drugs NDGA or REV5901 at the concentrations indicated using techniques as described in Materials and methods. The perfusion solution was then switched to 220 mosmol/l DMEM (including drugs at the same concentrations) resulting in a rapid fall in fluorescence as shown by the reduction in the ft/fo value below unity. The ratio was then determined at 6 min following hypotonic challenge (see Fig. 5 for details). Results shown are from n experiments on chondrocytes obtained from the cartilage from different animals, with the number of cells studied also given Condition

ft/fo 1 min

Control (n=6; 186) 0.892±0.013 NDGA 0.886±0.008 (75 µM, n=4; 103) REV5901 0.857±0.010 (50 µM), n=4; 59)

Significance level 6 min 1.018±0.027** 0.894±0.009

P<0.001 N/S

0.896±0.008

N/S

RVD in chondrocytes is mediated in part by the swelling-activated taurine transport pathway.

Discussion Despite the chronic anisotonic media used for chondrocyte isolation (180, 280 or 380 mosmol/l), cell volume was identical, suggesting that the volume-regulatory response maintained a “set-point” volume for activation of the taurine transport pathway. Lowering the osmolarity from the “set-point” by hypotonic challenge increased cell volume and stimulated the taurine transport pathway of both in situ and isolated chondrocytes. This pathway appeared to play a role in RVD, but its activation did not involve metabolites from the arachidonic acid cascade. Chondrocyte volume changed rapidly following acute exposure to hyper/hypotonic media; however, after ≅20 h, cell volume was identical (Table 1). Thus, the increase in volume following cell isolation into the anisotonic solutions (180, 280 mosmol/l) stimulated RVD, restoring volume towards normal. The cells isolated into 380 mosmol/l should have similar volumes as those of in situ chondrocytes since this was close to the in situ osmolarity [37]. However, the volume of bovine chondrocytes within the mid-zone of cartilage incubated in 280 mosmol/l measured using almost identical techniques to those used here gave a smaller volume (≅550 µm3; [4]) than that of isolated chondrocytes (≅645 µm3; Table 1). It is possible that changes in the chondrocytes’ physico-chemical environment (e.g. altered cation and/or anion composition) occurring following their liberation from the matrix [37] might be involved. This is important since matrix synthesis rates by freshly isolated chondrocytes are only ≅10% those of chondrocytes in situ on a per cell basis [37] and, thus,

there are substantial changes to chondrocyte behaviour following isolation. Selected pharmacological agents were used to determine whether, as in other cell types [20, 21, 22], metabolites from the arachidonic acid cascade might be involved in activating the taurine transport pathway. The inhibition observed with REV5901, MK886 and NDGA (Table 3) suggested a role for lipoxygenases. However, the addition of compounds involved in activating RVD in other cell types (i.e. leukotrienes and lipoxins; [14, 23]) directly to chondrocytes had no effect on the taurine transport pathway either in isotonic or hypotonic media (see Results). This makes it unlikely that transport pathway inhibition occurred by blocking the arachidonic acid cascade, and suggests that relatively non-specific actions were responsible. Additionally, the inhibitors CP105,696 (an antagonist that blocks the binding of LTB4 to LTB4 receptors; K0.5≅4 nM [8]) or L-660,711 (a potent and selective competitive inhibitor of leukotriene D4 binding to LTD4 receptors; K0.5 ≅2 nM [18]) had no effect on the taurine pathway. Interestingly, REV5901, the competitive LTD4 receptor antagonist and weak 5-lipoxygenase inhibitor [30], was ≅40× more potent than the REV5901 para-isomer at inhibiting the pathway, whereas the latter drug was ≅20× more effective at blocking 5-lipoxygenase (K0.5≅100 nM, [7]; Table 3, Fig. 4). We do not have an explanation for this differential effect. It might be thought that 5-lipoxygenase, or products from this pathway are not involved; however, caution should be exercised when interpreting results from these less specific drugs because of the possibility of multiple actions, as reported by others. For example, LTD4 does not relieve the inhibition of swelling-activated taurine efflux from NDGA-treated ascites tumour cells probably because, in addition to its inhibitory effects on the arachidonic acid cascade, NDGA blocks the LTD4 receptor [21]. Other agents could have similarly complex effects, for example NPPB the widely used Cl– channel blocker (e.g. [13, 19, 20]) inhibits cyclooxygenase (K0.5 ≅8 µM [2]). The signal transduction process for the chondrocyte taurine transport pathway is unknown, but might have similarities to that suggested for related swelling-activated pathways of other cell types [14, 15]. The characteristics of the system in chondrocytes and the responses of similar transport pathways in other cell types suggest that they are related [13, 35]. For swelling-activated osmolyte transport by C6 glioma and CHO cells [6, 35], amino acid and KCl efflux from trout erythrocytes [28] and for volume-sensitive Cl– currents in bovine endothelial cells [31], a key role for cytoplasmic ionic strength (µi) has been postulated. It is thought that µi controls both the volume “set-point” and rate of swelling-induced activation of volume-sensitive organic osmolyte/anion channel (VSOAC), and in C6 cells greater increases in cell volume are required to stimulate VSOAC with raised intracellular ionic strength [6]. This parallels data here (Fig. 2), where the slope of the relationship between medium osmolarity and taurine efflux was less in chon-

780

drocytes incubated at raised osmolarity. However, support for this is limited since chondrocyte ionic strength was not measured here. Interestingly, activation does not appear to result from a swelling-induced reduction in ionic strength [5]. This is because VSOAC can be activated without osmotically driven water influx by forcing fluid into the cytoplasm through a patch pipette without changing µi [5]. Thus, although swelling-induced changes to cytoplasmic ionic strength may modulate VSOAC, other signalling events, possibly involving changes to [Cl–]i, the cytoskeleton, and/or phospholipase A2, [15, 32, 35] must be involved. Although the volume of chondrocytes isolated into 180, 280 or 380 mosmol/l media are the same (Table 1), matrix synthesis rates are very different [37]. Thus, when osmolarity was altered over 280–480 mosmol/l and matrix synthesis measured using 35SO4 incorporation over 2 h, the maximum rate was ≅380 mosmol/l, i.e. close to the in situ osmolarity [37]. However, when synthesis was determined over 2 h following 24 h of incubation in 280 mosmol/l, the optimal rate was then at 280 mosmol/l. The present results suggest that cell volume would be the same following isolation into 280 or 380 mosmol/l (Table 1), indicating that factors other than chondrocyte volume regulate matrix synthesis. In other cell types, metabolism is sensitive to changes in cell cation (particularly K+, Na+, H+, Ca2+) concentrations. For example, in oocytes, protein synthesis increases four- to fivefold for only a 20–30% change in [K+]i [17]. It is known that matrix metabolism by chondrocytes is controlled by feedback from the matrix [29]. Changes to the interstitial ionic/osmotic composition around chondrocytes in vivo arising from static load, or in the early stages of osteoarthrosis, might lead to variations in cell composition, and hence this could mediate the changes to matrix metabolism by chondrocytes. Acknowledgements This work was supported by the Arthritis Research Campaign (H0559) and the Wellcome Trust (045925/ Z/95/A). We thank Dr. R.N. Young (Merck Frosst, Canada) for the gift of the L660,711 and Dr. H.J. Showell (Pfizer, Groton) for the gift of the CP-105,696. We also thank Mr. A. Armitage and Ms. R. Kerr for help with some of the early experiments, and Mr. C. Adams for providing human cartilage.

References 1. Alvarez-Leefmans FJ, Altamirano J, Crowe WE (1995) Use of ion selective microelectrodes and fluorescent probes to measure cell volume. Methods Neurosci 27:361–391 2. Breuer W, Skorecki KL (1989) Inhibition of prostaglandin E2 synthesis by a blocker of epithelial chloride channels. Biochem Biophys Res Commun 163:398–405 3. Bush PG, Hall AC (2001) Volume regulation by isolated and in situ articular chondrocytes. J Cell Physiol 187:304–314 4. Bush PG, Gardner DL, Adams C, Walker C, Aigner T, Hall AC (1999) Volume and morphology of articular chondrocytes studied within normal and degenerate cartilage. Trans Orthop Res Soc 24:128 5. Cannon CL, Basavappa S, Strange K (1998) Intracellular ionic strength regulates the volume sensitivity of a swelling-activated anion channel. Amer J Physiol 275:C416–C422

6. Emma F, McManus M, Strange K (1997) Intracellular electrolytes regulate the volume set point of the organic osmolyte/ anion channel VSOAC. Am J Physiol 272:C1766–C1775 7. Evans JF, Leveille C, Mancini JA, Prasit P, Therien M, Zamboni R, Gauthier JY, Fortin R, Charleson P, Macintyre DE, Luell S, Bach TJ, Meurer, R, Guay J, Vickers PJ, Rouzer CA, Gillard JW, Miller DK (1991) 5-Lipoxygenase-activating protein is the target of a quinoline class of leukotriene synthesis inhibitors. Mol Pharmacol 40:22–27 8. Griffiths RJ, Pettipher ER, Koch K, Farrell CA, Breslow R, Conklyn MJ, Smith MA, Hackman BC, Wimberly DJ, Milici AJ, Scampoli DN, Cheng JB, Lippar JS, Pazoles CJ, Doherty NS, Melvin LS, Reiter LA, Biggars MS, Falkner FC, Mitchell DY, Liston TE, Showell HJ (1995) Leukotriene B4 plays a critical role in the progression of collagen-induced arthritis. Proc Natl Acad Sci 92:517–521 9. Hall AC (1995) Volume-sensitive taurine transport in bovine articular chondrocytes. J Physiol (Lond) 484:755–766 10. Hall AC (1998) Physiology of cartilage. In Hughes SPF, McCarthy ID (eds) Basic science for orthopaedic trainees. Saunders, London, pp. 45–69 11. Hall AC, Horwitz ER, Wilkins RJ (1996) The cellular physiology of articular cartilage. Exp Physiol 81:535–545 12. Hall AC, Starks I, Shoults CL, Rashidbigi S (1996) Pathways for K+ transport across the bovine articular chondrocyte membrane and their sensitivity to cell volume. Am J Physiol 270:C1300–C1310 13. Hall JA, Kirk J, Potts JR, Rae C, Kirk K (1996) Anion channel blockers inhibit swelling-activated anion, cation and non-electrolyte transport in HeLa cells. Am J Physiol 271:C579–C588 14. Hoffmann EK, Dunham PB (1995) Membrane mechanisms and intracellular signalling in cell volume regulation. Int Rev Cytol 161:173–260 15. Hoffmann EK, Mills JW (1999) Membrane events involved in volume regulation. Curr Topics Membr Trans 48:123–196 16. Hoffmann EK, Pedersen SF (1998) Sensors and signal transduction in the activation of cell volume regulatory ion transport systems. In Lang F (ed) Cell volume regulation. Karger, Basel, pp 50–78 17. Horowitz S, Lau Y-T (1988) A function that relates protein synthetic rates to potassium activity in vivo J Cell Physiol 135:425–434 18. Jones TR, Zamboni R, Belley M, Champion E, Charette L, Ford-Hutchinson AW, Frenette R, Gauthier J-Y, Leger S, Masson P, McFarlane CS, Piechuta H, Rokach J, Williams H, Young RN (1989) Pharmacology of L-660,711 (MK-571): a novel potent and selective leukotriene D4 receptor antagonist. Can J Physiol Pharmacol 67:17–28 19. Kirk K, Ellory JC, Young JD (1992) Transport of organic substrates via a volume-activated channel. J Biol Chem 267: 23475–23478 20. Lambert IH (1993) Eicosanoids and cell volume regulation. In: Strange K (ed) Cellular and molecular physiology of cell volume regulation. CRC, Boca Raton, Fla., pp 273–292 21. Lambert IH, Hoffmann EK (1993) Regulation of taurine transport in Ehrlich ascites tumour cells. J Membr Biol 131:67–79 22. Lambert IH, Hoffmann EK (1994) Cell swelling activates separate taurine and chloride channels in Ehrlich mouse ascites tumour cells. J Membr Biol 142:289–298 23. Margalit A, Sofer Y, Grossman S Reynaud D, Pace-Asciak CR, Livne AA (1993) Hepoxilin A3 is the endogenous lipid mediator opposing hypotonic swelling in intact human platelets. Proc Natl Acad Sci USA 90:2589–2592 24. Maroudas A (1980) Physical chemistry of articular cartilage and the intervertebral disc. In: Sokoloff L (ed) The joints and synovial fluid, Vol. II. Academic, New York, pp 239–291 25. Maroudas A (1990) Different ways of expressing concentration of cartilage constituents. In: Maroudas A, Kuettner K (eds) Methods in cartilage research. Academic, London, pp 211–219 26. Maroudas A, Evans H (1972) A study of ionic equilibria in cartilage. Connective Tissue Res 1:69–79

781 27. McManus M, Serhan C, Jackson P, Strange K (1994) Ketoconazole blocks organic osmolyte efflux independently of its effect on arachidonic acid conversion. Am J Physiol 267:C266– C271 28. Motais R, Guizouarn H, Garcia-Romeu F (1991) Red cell volume regulation: the pivotal role of ionic strength in controlling swelling-dependent transport processes. Biochim Biophys Acta 1075:169–180 29. Muir H (1981) Chemistry of the ground substance in joint cartilage. In: Sokoloff L (ed) The joints and synovial fluid, Vol. II. Academic, New York, pp 27–94 30. Musser JH, Chakraborty VR, Sciortino S, Gordon RJ, Khandwala A, Neiss ES, Pruss TP, Inwegen RV, Weinryb I, Coutts SM (1987) Substituted arylmethyl phenyl ethers 1. A novel series of 5-lipoxygenase inhibitors and leukotriene antagonists. J Med Chem 30:96–104 31. Nilius B, Prenen J, Voets T, Eggermont J, Droogmans G (1998) Activation of volume-regulated chloride currents by reduction of intracellular ionic strength in bovine endothelial cells. J Physiol (Lond) 506:353–361

32. Okada Y (1997) Volume expansion-sensing outward-rectifier Cl– channels: fresh start to the molecular identity and volume sensor. Am J Physiol 273:C755–C789 33. Schneiderman R, Keret D, Maroudas A (1986) Effects of mechanical and osmotic pressure on the rate of glycosaminoglycan synthesis in the human adult femoral head. J Orthop Res 4:393–409 34. Stockwell RA (1991) Cartilage failure in osteoarthritis: relevance of normal structure and function. A review. Clin Anat 4:161–191 35. Strange K, Emma F, Jackson PS (1996) Cellular and molecular physiology of volume-sensitive anion channels. Am J Physiol 270:C711–C730 36. Urban JPG (1994) The chondrocyte: a cell under pressure. Br J Rheumatol 33:901–908 37. Urban JPG, Hall AC, Gehl KA (1993) Regulation of matrix synthesis rates by the ionic and osmotic environment of articular chondrocytes. J Cell Physiol 154:262–270