Water permeability of Na+–K+–2Cl– cotransporters in mammalian epithelial cells
1 Nordic Centre for Water Imbalance Related Disorders, Department of Medical Physiology, The Panum Institute, University of Copenhagen, Denmark
2 Departamento de Neurobiología, Instituto Nacional de Psiquiatría, Mexico DF, Mexico
3 Department of Pharmacology and Toxicology, Wright State University, Dayton, OH, USA
Abstract
Water transport properties of the Na+–K+–2Cl– cotransporter (NKCC) were studied in cultures of pigmented epithelial cells (PE) from the ciliary body of the eye. Here, the membrane that faces upwards contains NKCCs and can be subjected to rapid changes in bathing solution composition and osmolarity. The anatomy of the cultured cell layer was investigated by light and electron microscopy. The transport rate of the cotransporter was determined from the bumetanide-sensitive component of 86Rb+ uptake, and volume changes were derived from quenching of the fluorescent dye calcein. The water permeability (Lp) of the membrane was halved by the specific inhibitor bumetanide. The bumetanide-sensitive component of the water transport exhibited apparent saturation at osmotic gradients higher than 200 mosmol l–1. Cell shrinkages produced by NaCl or KCl were smaller than those elicited by equi-osmolar applications of mannitol, indicating reflection coefficients for these salts close to zero. The activation energy of the bumetanide-sensitive component of the Lp was 21 kcal mol–1, which is four times higher than that of an aqueous pore. The data suggest that osmotic transport via the cotransporter involves conformational changes of the cotransporter and interaction with Na+, K+ and Cl–. Similar measurements were performed on immortalized cell cultures from the thick ascending limb of the loop of Henle (TALH). Given similar overall transport rates of bumetanide-sensitive 86Rb+, the NKCCs of this tissue did not contribute any bumetanide-sensitive Lp. This suggests that the cotransporters of the two tissues are either different isoforms or the same cotransporter but in two different transport modes.
, http://www.100md.com
Introduction
It is now generally accepted that cotransporters support passive water transport (King et al. 2004). The unit water permeability is significant; for example, for the Na+-coupled glutamate transporter EAAT1 it is about 10 times larger than that of aquaporin 0 (AQP0) and 10 times smaller than that of AQP1 (MacAulay et al. 2002). When the large number of cotransporters present per cell is taken into account, their contribution to the overall passive water permeability is physiologically relevant. Moreover, the water permeability of the cotransporters depends on their conformational state, which suggests that they play a role in the rapid regulation of cell water permeability (reviewed by MacAulay et al. 2004). It has been suggested that in addition to passive translocation, cotransporters transport water actively, with a strict stoichiometry, along with the non-aqueous substrates. The coupling ratio is high; for instance, the Na+-coupled glucose cotransporter transports 210 water molecules for each turnover of the protein. However, this active mode of water transport is currently under debate (Zeuthen et al. 2002; Gagnon et al. 2004).
, 百拇医药
The purpose of this paper is to test whether the Na+–K+–Cl– cotransporter (NKCC) has passive water permeability. The NKCC1 isoform of this cotransporter can be studied in cell cultures from the ciliary epithelium of the eye (Layne et al. 2001; Hochgesand et al. 2001). In the mammalian eye the ciliary epithelium is responsible for the secretion of aqueous humour, transferring solute and water from the blood in the ciliary stroma into the posterior chamber of the eye (Fig. 1A). The circulating clear aqueous humour nourishes the cornea and lens and maintains the structural integrity and optical properties of the globe. The ciliary epithelium consists of two cell layers, the pigmented layer (PE) and the non-pigmented layer (NPE). As in other secretory epithelia NKCC1 is located in the basal membrane across which water enters the epithelium from the blood (Dunn et al. 2001). In the ciliary epithelium water flows from the PE cells via gap junctions into the NPE cells from where it continues into the aqueous humour, probably via aquaporins AQP1 and 4 (Hamann et al. 1998; Hamann, 2002). When PE cells are separated from the NPE cells and grown in culture on cover glasses, the membrane facing upwards contains NKCC1 and can be subjected to rapid changes in bathing solution composition and osmolarity, Fig. 1B. There are no known aquaporins in the PE cells (Hamann et al. 1998).
, 百拇医药
A, the secretory epithelium of the ciliary body consists of pigmented (PE) and non-pigmented (NPE) cell layers. Water is transported from the interstitial solution (stroma or blood side) into the PE cells and via gap junctions into the NPE cells from where it enters the aqueous humour via aquaporins (AQP1 and 4). In vivo, the Na+–K+–Cl– cotransporter is located at the blood-facing membrane. B, in cultured PE cells, however, the membrane that contains the Na+–K+–Cl– cotransporter faces upwards (becomes apical) and its water transport properties can be derived from initial rates of changes in cell volume induced by abrupt changes in bathing solution osmolarity or composition. Cell volumes were monitored as quenching of the intracellular fluorophore calcein via an inverted microscope. The Na+–K+–Cl– cotransporter in the apical membrane of cultured cells from the thick ascending limb of the loop of Henle (TALH cells) was studied by the same set-up.
, 百拇医药
A comparative study was performed on cell cultures from the epithelial cells of the thick ascending limb of the loop of Henle of the kidney (TALH cells) (Scott et al. 1986). In vivo, this epithelium has a Na+–K+–Cl– cotransporter at its apical membrane (isoform NKCC2) (Plata et al. 2002), and is responsible for creating the interstitial hyperosmolarity. The epithelium is relatively water impermeable. In culture, TALH cells grow with their apical membrane facing upwards and can be subjected to experiments similar to the PE cells (Fig. 1B).
, 百拇医药
In order to compare the properties of PE and TALH cultures, we first determined the geometry and ultrastructure by light and electron microscopy. Second, we determined the ion transport rates of the NKCCs in the two cell types. With these factors accounted for, the present paper shows that the NKCC of the PE cells transports water while the NKCC of the TALH cells does not. The passive water transport of the NKCC from the PE differs from that of simple water channels (i.e. aquaporins) by having high activation energies and low reflection coefficients for NaCl and KCl, and by exhibiting apparent saturation with increasing osmotic gradients.
, 百拇医药
Methods
Cell cultures
Human PE cells were generated from frozen stocks of cells previously dissected from 5-month-old aborted fetuses as described (Von Brauchitsch & Crook, 1993). Cells were thawed and plated in Lab-Tek 8-well chambered cover glasses (Life Technologies, Denmark). In each well the glass had a square cross-section of 0.81 cm2. PE cells were grown to confluence in culture medium (M199) supplemented with 15% fetal calf serum (FCS), 1 ng ml–1 basic fibroblast growth factor (bFGF), 300 μg ml–1 glutamine, 50 μg ml–1 gentamicin, and 2.5 μg ml–1 fungizone. bFGF was from Calbiochem (CA, USA), and the remaining components were from Invitrogen (Denmark). Cultures were kept at 37°C in a 5% CO2 environment, and the medium was changed every second day. Upon confluence, cultures were maintained in M199 supplemented as above but with 10% FCS and without bFGF. Confluent layers of PE cells in their fifth to eighth passages were used for experiments. Cell cultures were kindly made available by Dr R. B. Crook (UCSF, CA, USA). All protocols followed the Declaration of Helsinki. Informed consent for the use of fetal tissue in research was obtained, and institutional human experimentation committee approval was obtained for the use of human eyes.
, http://www.100md.com
Rabbit TALH cells were generated from frozen stocks of cells kindly supplied by Dr R. K. H. Kinne (Scott et al. 1986). Cells were thawed and plated on the chambered cover glasses and grown to confluence in culture medium (Dulbecco's modified Eagle's medium, high glucose) supplemented with 1% non-essential amino acids, 5% fetal calf serum, 1% L-glutamine, 1% pyruvate, 0.1% -mercaptoethanol, 10–7 mol l–1 arginine-vasopressin and 10–7 mol l–1 thyrocalcitonin. Vasopressin and thyrocalcitonin were from Sigma (Denmark); other reagents came from Invitrogen (Denmark). Cultures were kept at 37°C and 7.5% CO2 and the medium was changed every second day. Confluent layers of TALH cells in sixth to tenth passage were used for experiments.
, 百拇医药
Light and electron microscopy
Light microscopy. Cultures of PE and TALH cells were fixed in 2% glutaraldehyde and 1% methylene blue in 0.1 mol l–1 sodium cacodylate buffer for 60 min. Subsequently, the cultures were rinsed in 0.9% NaCl and examined in an inverted microscope. Micrographs were taken with a Nikon Coolpix 4500 digital camera.
Electron microscopy. Cultures of PE and TALH cells were left overnight in a solution containing 1% formaldehyde and 1% glutaraldehyde in 0.1 mol l–1 sodium cacodylate buffer (pH 7.3). The fixed cultures attached firmly to the surface of the cover glasses. After buffer rinse, crossing lines in the cell layers were cut with the sharp tip of a fine needle and small flakes of cells were flushed away from the supporting glass surface. These tissue samples were further processed for electron microscopy as previously described (Hamann et al. 2000).
, http://www.100md.com
Measurement of cell height in living cells
PE cells in seventh passage or TALH cells in ninth passage were incubated for 40 min at room temperature in control solution containing 4 μmol l–1 calcein-AM, the membrane-permeant non-fluorescent acetoxymethyl ester of calcein. Inside the cells esterases cleave off the acetoxymethyl groups and produce the membrane-impermeant fluorescent dye calcein. Confocal optical sections were performed using a Noran Odyssey Confocal Microscope equipped with a x40, NA 1.3 oil immersion objective (Nikon). Single wavelength excitation laser light of 488 nm and a slit size of 10 μm were used. The epifluorescence emission was directed through a 515 LP FITC filter and 3D image reconstruction was used to calculate cell heights.
, 百拇医药
Isotopic flux measurements
86Rb+ was obtained from Amersham Biosciences Ltd (UK) as RbCl with a specific activity of 1.5 mCi ml–1. Test solutions were prepared with specific concentrations of about 5 μCi ml–1, which correlates with Rb+ concentrations between 3 and 10 μmol l–1. In order to normalize and compare experiments, the radioactive decay of 86Rb+ was corrected for as follows. For each experiment, the uptake of 86Rb+ in μCi (or Rb+ in mol) cm–2 of cell layer was divided by the 86Rb+ in μCi (or Rb+ in mol) cm–3 test solution (final units: cm). For example, a Rb+ uptake of 10–3 cm means that the cell layer has taken up 10 pmol cm–2, given a test solution concentration of 10 nmol cm–3. Thus, uptake per second is equivalent to permeability, PRb, in units of cm s–1.
, http://www.100md.com
86Rb+ uptake was initiated by applying 0.25 ml of radioactive test solution to the tissue culture well. After the given time, uptake was stopped by washing twice with 0.5 ml of ice-cold saline. Subsequently, the cells were dissolved in 0.5 ml of 5% SDS (sodium dodecyl sulphate) and transferred to 10 ml of distilled water in plastic vials and counted on a Hewlett Packard -counter. In experiments where the uptake of 86Rb+ was studied in the presence of ouabain and/or bumetanide, the inhibitors were applied 30 s prior to the 86Rb+. This would prevent short-term uptake of tracer before the inhibition was fully effective. Ouabain was always present at concentrations of 100 μmol l–1. Bumetanide was used in concentrations of 10–100 μmol l–1 unless otherwise indicated. The bumetanide-sensitive component of the 86Rb+ influx was taken as the uptake observed when ouabain was present minus the uptake in the presence of ouabain plus bumetanide (Hochgesand et al. 2001).
, 百拇医药
Cell water volume measurements
The set-up was similar to the one previously described for measurements of water permeabilities Lp (Hamann et al. 2002, 2003). Some experiments were performed in Mexico City (set-up 1), others in Copenhagen (set-up 2). There was no difference between the results obtained in either laboratory. Chambers with confluent layers of PE or TALH cells were mounted on the stage of a Nikon Diaphot 300 inverted microscope (set-up 1) or a Zeiss Axiovert 10 inverted microscope (set-up 2). Only the solution bathing the upper or apical surface of the cell layer was changed (Fig. 1B). The perfusion solutions were fed into the chamber by gravity (set-up 1) or via a Gilson Minipuls 3 peristaltic pump (Biolab, Denmark, set-up 2) and collected by aspiration. At maximal flow rates (230 μl s–1), the solution at the level of the cells was exchanged 90% in 5 s. Most measurements were performed on confluent cell layers. A few measurements were done at the centre of confluent sheets of cells (about 100 μm by 100 μm). In these cases, concentration changes at the abluminal membrane can be disregarded; according to conventional diffusion theory, solution changes at the luminal side cannot affect the concentration at the abluminal side within the experimental period of 10 s.
, 百拇医药
The fluorescent dye calcein was used for measurement of changes in epithelial cell water volume (see below). The cells were incubated at room temperature in control solution containing 1–5 μmol l–1 calcein-AM. The latter is converted intracellularly to the membrane-impermeant fluorescent dye calcein, that accumulates intracellularly and reaches self-quenching concentrations, which in free solution are above 4 mmol l–1 (Hamann et al. 2002). After 40–60 min, the loading solution was washed out by control solution for 1 h before experiments were started.
, 百拇医药
In set-up 1, the microscope was attached to a DMX-1100 (SLM) monochromator, which provided the single wavelength excitation light of 497 nm (± 4 nm). Light pulses, lasting less than 300 ms, were delivered at a frequency of 0.5 Hz. A fast shutter was closed between the excitation pulses to reduce photobleaching and photodynamic damage. To further reduce exposure to UV light, a neutral density filter was used that transmitted only 5% of the light. In addition, an iris diaphragm confined the illuminated field to a light spot of 10–100 cells. A x40, NA 1.3 oil immersion Fluor objective (Nikon) was used for measurements. The epifluorescence emission was measured at 535 ± 12.5 nm, in defined regions, using a CCD camera (Hamamatsu C2400), and the output signals were digitally stored. Set-up 2 differed from set-up 1 in the following ways. The excitation light was directed through a 490 nm (± 5 nm) excitation filter, light pulses lasted less than 100 ms and were delivered at a frequency of between 0.3 and 1 Hz, the neutral density filter transmitted only 3% of the light, and the epifluorescence emission was measured between 503 and 530 nm using a photomultiplier tube (Zeiss).
, 百拇医药
The technique for measuring cell water volume changes using calcein self-quenching has been described in detail (Hamann et al. 2002). In brief, fluorescence signals are corrected for drift caused by volume-independent reduction in fluorescence intensity, and relative background caused by the fraction of intracellular calcein fluorescence, which is insensitive to changes in external osmolarity. This insensitive fraction presumably reflects calcein that is internally bound or compartmentalized (Alvarez-Leefmans et al. 1995; Crowe et al. 1995; Hamann et al. 2002; Solenov et al. 2004). In a given group of cells, the relative background was determined from calibrations obtained by applying hyposmotic and hyperosmotic solutions of known osmolarities. For subsequent experiments in the same group of cells, the fluorescence signals were corrected using the value of relative background obtained from the calibration. Drift-correction was performed for all individual experimental challenges. The change in cell water volume was calculated as
, 百拇医药
where Vt is the cell water volume at time t, Vt equals V0 at t = 0, Ft is the drift-corrected fluorescence at time t, F0 is the drift-corrected steady-state fluorescence before and after a given experimental challenge, and fb, the relative background, signifies the relative fluorescence insensitive to changes in external osmolarity. The influx of water per square centimetre of epithelium was calculated as
where 1/V0 x dVt/dt is the initial relative rate of change in cell water volume, h0 is the anatomical cell height at time zero, and is the ratio of cell water volume to the anatomical volume, which is taken as 1 (see Discussion).
, 百拇医药
The Lp was obtained according to eqn (3), from the initial rate of change in cell volume taken within 10–15 s, divided by the imposed osmotic gradient (),
under the assumption that the imposed osmotic gradient was instantaneous. However, the chamber was perfused at a rate of 12 ml min–1, and the solution at the level of the cells was exchanged 90% in 5 s (90 = 5 s; Hamann et al. 2003). This finite rate gives an upper limit to temporal resolution of the cellular volume changes. If the recorded cellular response has a 90 of 20 s, it can be calculated that the apparent Lp will be underestimated by a factor of about 1.1 compared to the true Lp that would have been determined if the solution change had been instantaneous. This was the case for PE cells at room temperature. If the recorded response had a 90 of 15 s the Lp would be underestimated by a factor of 1.25 (i.e. PE cells at 30°C); if the recorded response had a 90 of 10 s the Lp would have been underestimated by a factor of 1.8. The latter was the case for measurements of PE cells at 37°C: such data were not included.
, 百拇医药
Solutions
Control solution contained (in mmol l–1): 118 Na+, 6 K+, 1.2 Mg2+, 2.6 Ca2+, 118 Cl–, 1 PO42–, 25 Hepes, 5.5 glucose, and 29 mannitol. Anisosmotic solutions were prepared by mannitol addition or removal. In Cl–-free solutions equimolar amounts of Cl– were replaced with gluconate. In Na+-free solutions Na+ was replaced by N-methyl-D-glucamine, and K+ was replaced by mannitol in K+-free solutions. The salts were of analytical quality and the osmolarities were checked by means of a freezing-point depression osmometer. All solutions were titrated to pH 7.40. Calcein-AM was obtained from Molecular Probes (The Netherlands); other reagents were obtained from Sigma.
, 百拇医药
Statistics
Results are expressed as mean ± S.E.M.; comparisons were made using Student's unpaired t test, and a P value below 0.05 was used as a measure of statistical difference. Unless otherwise stated, the numbers given in parentheses correspond to the number of groups of cells tested. The number of cultures tested was, in most cases, larger than four. The changes in cell water volume were linear for at least the initial 10 s (4–11 data points depending on the frequency of data acquisition) and during this time, the initial rates of change in cell water volume were calculated by regression analysis.
, 百拇医药
Results
Anatomical properties
With light microscopy the PE cells appeared heterogeneous: locally the cultures were clearly multilayered, other fields appeared as partially confluent monolayers. The cells were large and elongated (15–20 μm x 40–50 μm) arranged in parallel (Fig. 2A). TALH cultures appeared as confluent cobblestone layers (Fig. 2B) of 10 μm x 15 μm cells. The position of nuclei and light lucent lines spaced at intervals corresponding to the dimensions of individual cells (Fig. 2B) indicating that the TALH cultures, at least in the focal plane, were two- or multilayered. At the ultrastructural level, both PE and TALH cells appeared moderately electron dense and rich in the usual organelles and microfilaments (Fig. 2, middle panels). Pigment granules at any developmental stage were not encountered in the cytoplasm of PE cells. Both cultures were organized as a transporting epithelium with luminal protrusions from the cells and with tight junctions sealing the luminal aspects of the intercellular clefts (Fig. 2, bottom panels). In the tissue samples chosen for electron microscopy both the PE and the TALH cell cultures were two- or three-layered. In addition, multiple flattened cellular processes occurred between and under the cells. The total height of the cell layers was between 5 and 10 μm. The cell height determined in the living cells by means of confocal microscopy was slightly larger than that estimated from electron microscopy. Accordingly, the height (h) of the PE cells was 7.8 ± 0.1 μm (n = 25) and that of the TALH cells was 6.7 ± 0.4 μm (n = 8).
, http://www.100md.com
A, PE cells. Top panel shows a light micrograph of cultured PE cells. The elongated cells are arranged in parallel with some degree of overlapping. Scale bar, 15 μm. Middle panel: electron micrograph of cultured PE cells, showing overlapping. Straight arrow marks microvilli-like processes from the luminal cell. Projections from neighbouring cells are indicated by arrowhead. Scale bar, 1 μm. Bottom panel: a tight junction between apposing cells in a culture of PE cells. Arrow marks membrane contacts between the luminal aspects of neighbouring cells. Scale bar, 0.1 μm. B, TALH cells. Top panel shows cultured TALH cells organized in a cobblestone layer. The lucent lines (arrows) possibly outline clefts between neighbouring cells in a deeper focal plane. Scale bar, 10 μm. Middle panel: electron micrograph of cultured TALH cells. Flattened cells and cellular processes are seen under the well-organized and specialized luminal cell layer. Arrow marks microvilli. Scale bar, 1 μm. Bottom panel: a junctional complex (arrow) at the luminal aspect of adjacent TALH cells. Scale bar, 0.1 μm.
, 百拇医药
Transport of 86Rb+
In order to evaluate and compare the rate of transport of the NKCC in PE and TALH cells, the rate of 86Rb+ uptake was monitored under various experimental conditions. 86Rb+ is known to replace K+ in cation–chloride cotransporters and the bumetanide-sensitive component of 86Rb+ uptake represents NKCC transport rates. This component was obtained as the 86Rb+ uptake in tissues treated with ouabain (100 μmol l–1) minus the 86Rb+ uptake in tissues treated with both ouabain (100 μmol l–1) and bumetanide (10–100 μmol l–1). Uptakes of 86Rb+ will be presented relative to the luminal concentration of 86Rb+ in permeability units (cm s–1). Most experiments were performed at both room temperature (20–24°C) and at 37°C in order to estimate Arrhenius activation energies (Ea). In general, for a given temperature, the 86Rb+ transport was found to be about 50% faster in TALH cells than in PE cells.
, http://www.100md.com
Uptake of 86Rb+ in tissues not treated with ion-transport inhibitors. In the absence of ouabain and bumetanide, the initial rate of 86Rb+ uptake in PE cells (37°C) was equivalent to a Rb+ permeability of (4.4 ± 0.3) x 10–6 cm s–1 (n = 4) and for TALH cells it was (6.9 ± 0.2) x 10–6 cm s–1 (n = 7), see Fig. 3A. The 86Rb+ uptake decreased mono-exponentially with time: for PE cells the time constant was 18 ± 0.3 min, and for TALH cells it was 64 ± 0.8 min. At 24°C the initial rates of isotope uptake were indistinguishable for the two cell types (about 2.0 x 10–6 cm s–1), but with longer uptake times a faster transport rate became evident for TALH cells.
, http://www.100md.com
The cell layers were bathed in control solution to which about 5 μCi ml–186Rb+ was added at time zero. A, uptake into cells not treated with inhibitors. Each point was determined from 4 or 8 tissues, except for PE at 37°C, where the 45 min and 90 min values are based on two single determinations. The data are fitted to mono-exponential curves; in all four cases the regression analysis gave R2 values larger than 0.99. B, the bumetanide-sensitive component of uptake was determined as the difference between uptakes determined in ouabain-treated tissues (100 μmol l–1; C), and in tissues treated with both ouabain and bumetanide (100 μmol l–1; D); data points were determined from between 4 and 8 experiments. S.E.M. values are not shown if smaller than points. Note the different scales in A compared to B, C and D. The ordinates can also be read as Rb+ uptake per square centimetre in units of 10 pmol cm–2 (see Methods).
, http://www.100md.com
The bumetanide-sensitive component of 86Rb+ uptake. As already mentioned, this component (Fig. 3B) was the difference between the uptake observed in ouabain-treated tissues (Fig. 3C) and those treated with both ouabain and bumetanide (Fig. 3D). The bumetanide-sensitive component was approximately linear during the first 10 min for PE cells and for the first 15 min for the TALH. The slope of these lines defines the bumetanide-sensitive 86Rb+ permeability (PRb,but). For PE cells PRb,but was (0.44 ± 0.09) x 10–6 cm s–1 at 24°C and (1.27 ± 0.07) x 10–6 cm s–1 at 37°C. For TALH cells PRb,but was (0.66 ± 0.07) x 10–6 cm s–1 at 24°C and (1.76 ± 0.09) x 10–6 cm s–1 at 37°C. In these experiments bumetanide was used at concentrations of 100 μmol l–1, but similar data were obtained with concentrations of 10 μmol l–1 (see Fig. 4).
, 百拇医药
A, PE cells were treated in paired experiments with either ouabain (control, C; n = 16) or with ouabain plus bumetanide. Bumetanide concentrations of 100, 10, 1 and 0.1 μmol l–1 were tested, n = 4 for each value. Bumetanide at 100 and 10 μmol l–1 inhibited maximally, about 50%, while 1 μmol l–1 inhibited significantly less than 10 μmol l–1 (P = 0.02). B, similar results were obtained with TALH cells, 1 μmol l–1 bumetanide inhibited significantly less than 10 μmol l–1 (P = 0.03). In C the effects of external Na+ or Cl– removal on 86Rb+ uptake by TALH cells are compared with those of bumetanide. TALH cells were treated with ouabain (control, C; n = 12) or in paired experiments with ouabain plus 100 μmol l–1 bumetanide, ouabain combined with Cl– removal (0 Cl–) or ouabain combined with Na+ removal (0 Na+). Cl– was replaced with gluconate and Na+ with choline ions. Uptakes lasted 10 min at 37°C.
, http://www.100md.com
PRb,but could also be determined without using ouabain, i.e. as the difference between the uptakes obtained in control solutions and those obtained in control solutions with bumetanide added. At 37°C, PRb,but for PE cells was (1.14 ± 0.21) x 10–6 cm s–1 (n = 7), and for TALH cells (1.40 ± 0.33) x 10–6 cm s–1 (n = 4). These values are similar to those obtained using ouabain (as above), but the relative errors are larger since the values arise as the difference between larger numbers.
, 百拇医药
The temperature dependence of PRb,but was determined from uptakes obtained at temperatures between 20 and 37°C. Ea for the TALH cells was 18.9 ± 4.2 kcal mol–1 (n = 21) and for the PE cells 25.3 ± 4.6 kcal mol–1 (n = 19). The values were not significantly different.
Bumetanide sensitivity. NKCC has a bumetanide affinity of about 1 μmol l–1 (Russell, 2000). In accordance with this, it has been shown that 10 μmol l–1 bumetanide abolishes transport by the NKCC in PE cells (Crook et al. 2000). Here we confirm and extend these findings to the NKCC in TALH cells (Fig. 4A and B). Tissues were inhibited by ouabain plus various concentrations of bumetanide; tissues inhibited by ouabain only, served as controls. For both PE and TALH cells, 10 μmol l–1 of bumetanide gave maximal inhibition of the bumetanide-sensitive uptake of 86Rb+ (Fig. 4A and B), while 0.1–1 μmol l–1 gave half-maximal inhibitions.
, 百拇医药
Requirements for Cl– and Na+. The bumetanide-sensitive component of 86Rb+ uptake into PE cells requires Cl– and Na+ in the test solution (Hochgesand et al. 2001). To test this for TALH cells, we compared the uptake of 86Rb+ under three conditions. Cells were bathed (i) in control solution with ouabain and 100 μmol l–1 bumetanide, (ii) in Cl–-free bathing solution plus ouabain (Cl– replaced by gluconate ions), or (iii) in Na+-free bathing solution with ouabain (Na+ replaced by choline ions), all experiments at 37°C (Fig. 4C). Absence of Cl– and Na+ from the bathing solution inhibited the uptake of 86Rb+ to the same degree as bumetanide. We conclude that the bumetanide-sensitive uptake of 86Rb+ into TALH cells has an absolute requirement for Cl– and Na+ consistent with the notion that it represents uptake via NKCC.
, http://www.100md.com
Bumetanide-sensitive uptake of 86Rb+ during hyperosmotic challenges. In order to estimate the rate of cotransport during the measurements of passive water permeability (Lp), the 86Rb+ uptake was measured in control conditions and compared to situations where 100 mmol l–1 mannitol or 50 mmol l–1 NaCl had been added. Application of 50 mmol l–1 KCl was not attempted due to the inherently low specific activity in such experiments. All these experiments were performed at 37°C. For both kinds of tissues, uptakes obtained under hyperosmolar conditions were significantly larger than those obtained in the control situation. For PE cells, the uptake was (0.42 ± 0.07) x 10–3 cm (n = 8) under control conditions, (0.72 ± 0.06) x 10–3 cm (n = 8) when 100 mmol l–1 mannitol had been added, and (0.69 ± 0.07) x 10–3 cm (n = 8) when 50 mmol l–1 Na+ had been added; uptakes lasted 10 min. For TALH cells, the uptakes were (0.25 ± 0.06) x 10–3 cm (n = 8) under control conditions, (0.54 ± 0.08) x 10–3 cm (n = 8) when 100 mmol l–1 mannitol had been added, and (0.72 ± 0.03) x 10–3 cm (n = 8) when 50 mmol l–1 Na+ had been added; uptakes lasted 2.5 min.
, http://www.100md.com
Water transport properties in the PE cells
The finite rate of solution change in the chamber (90 = 5 s) gives an upper limit to the rate of cellular volume changes that can be resolved (see Methods). Consequently, we have limited our investigations of PE cells to temperatures below 30°C.
Passive water permeability, Lp. When the osmolarity of the bathing solution was changed by , cell volume, Vt (or cell height, ht), changed towards a new steady state (Fig. 5A). The initial rate of volume change (dVt/V0dt) taken within the first 10–15 s multiplied by the cell height at time zero, h0 (7.8 μm, see above), gave the passive water permeability (eqn (3)). For the smaller osmotic challenges, Lp was independent of the magnitude and direction of as shown by the proportionality between the volume changes, JH2O = Lp = h0 dVt/V0dt, and the osmotic challenge (Fig. 5C). Under control conditions, at 24°C, Lp was (1.51 ± 0.12) x 10–4 cm s–1 (osmol l–1)–1 (n = 89). Larger osmotic gradients, 200 and 400 mosmol l–1, were less efficient in inducing water transport; apparently the water transport saturates at higher osmotic gradients.
, 百拇医药
A, relative cell water volume changes Vt/V0 in response to hyperosmotic challenges of 50, 100, 200 and 400 mmol l–1 mannitol. B, same as A, but in the presence of 10 μmol l–1 bumetanide. All recordings from the same tissue. C, water fluxes, JH2O, derived from experiments as in A and B, as a function of the imposed osmotic gradient osm. Paired experiments at 24°C from 4 tissues.
Effects of bumetanide. Bumetanide reduced Lp to about half, (0.81 ± 0.08) x 10–4 cm s–1 (osmol l–1)–1 (n = 8), and the Lp of the bumetanide-treated tissues was the same at small and large osmotic gradients (Fig. 5B and C). In most experiments bumetanide was used at concentrations of 10 μmol l–1, but concentrations of 20–100 μmol l–1 gave similar results. Bumetanide itself did not change the cell volume or cell height when added to the control isosmotic solution (data not shown), so the reduction in Lp was a result of a low value of dVt/V0dt.
, http://www.100md.com
Effects of temperature. The bumetanide-sensitive component of the Lp was large and significant at higher temperatures. In the present group of measurements it was 40% of the total Lp at 30°C, 24% at 22.9°C, but was insignificant at 15.6°C (Fig. 6). Under control conditions, the total Lp had an Ea of 9.3 ± 1.0 kcal mol–1 while Ea was reduced to 2.9 ± 1.0 kcal mol–1 in the presence of bumetanide. This shows that the bumetanide-sensitive component of the Lp has a high Ea. In the range 22.9–30.0°C it is calculated as 21 kcal mol–1. The data of the lower temperature range suggest an even higher Ea, but the scatter of the data at 15.6°C excludes an exact determination.
, 百拇医药
The Lp of PE cells were measured at 30.0, 22.9 and 15.6°C without and with bumetanide present (20 μmol l–1). Data are presented as ln(Lp) versus reciprocal absolute temperature. Lp values with and without bumetanide were significantly different at the higher temperatures but not at 15.6°C. Numbers of observations are shown in parentheses.
Reflection coefficients for NaCl (NaCl) and KCl (KCl). Lp was determined from additions of 100 mmol l–1 mannitol, while NaCl x Lp was determined from addition of 50 mmol l–1 NaCl and KCl x Lp by the addition of 50 mmol l–1 KCl. The values were obtained by division, both NaCl and KCl were close to 0.5 (data in Fig. 7). In the presence of bumetanide (20 μmol l–1) Lp decreased to about half (compare Fig. 5), while NaCl and KCl increased to 1. If we assume that the water transport across the membrane is shared about equally between the bumetanide-insensitive pathway in parallel with the bumetanide-sensitive pathway, it follows from irreversible thermodynamic theory of parallel pathways in composite membranes (Kedem & Katchalsky, 1963), that the reflection coefficients of the bumetanide-sensitive pathways for NaCl and KCl are both close to zero, while those of the bumetanide-insensitive pathway are close to one.
, http://www.100md.com
Upper panel, cell shrinkage induced by mannitol compared to shrinkage induced by isosmolar concentrations of NaCl, recordings from the same cells. Lower left panel, summary of Lp in control measured by mannitol (man) without and with bumetanide (+b). Lower right panel, reflection coefficient measured with 50 mmol l–1 NaCl or KCl without and with bumetanide. Recordings at 24°C. Numbers of tests in parentheses, error bars are S.E.M. as in Fig. 9.
Water transport properties in the TALH cells
, http://www.100md.com
Passive water permeability (Lp), effects of temperature and bumetanide. When the osmolarity of the bathing solution was changed by , cell volume changed towards a new steady state; an example is shown in Fig. 9. The initial rate of volume change taken within the first 10–15 s multiplied by the cell height at time zero, h0 (6.7 μm, see above), gave the passive water permeability Lp (eqn (3)). Lp was independent of the magnitude and direction of as shown by the proportionality between the volume changes, JH2O, and the osmotic challenge (Fig. 8).
, http://www.100md.com
Upper panel, cell shrinkage induced by 50 mmol l–1 NaCl, 50 mmol l–1 KCl and 100 mmol l–1 mannitol. Recordings obtained from the same group of cells. Lower left panel, Lp measured by mannitol (man) without and with bumetanide (+b). Lower right panel, reflection coefficient measured with 50 mmol l–1 NaCl or KCl without and with bumetanide. Recordings at 37°C. Numbers of tests in parentheses, error bars are S.E.M.
Water fluxes, JH2O, as a function of imposed osmotic gradient osm with and without bumetanide. Paired experiments from 4 tissues. Recordings at 24°C.
, 百拇医药
Under control conditions at 37°C, Lp was (0.87 ± 0.07) x 10–4 cm s–1 (osmol l–1)–1 (n = 5). From paired experiments at 24°C and 37°C, Ea was calculated as 11.1 ± 2.1 kcal mol (n = 5). Bumetanide (10–100 μmol l–1) had no significant effect on the Lp either at 24°C (Fig. 8) or at 37°C (Fig. 9). Lp showed no dependence on in the range 0–200 mosmol l–1 (Fig. 8).
Reflection coefficients for NaCl (NaCl) and KCl (KCl). There was no significant difference between the rate of cell shrinkage obtained by abrupt additions of 100 mmol l–1 mannitol, 50 mmol l–1 NaCl, or 50 mmol l–1 KCl (Fig. 9, upper panel). Accordingly, NaCl and KCl of the TALH cell membranes were both close to 1. The presence of bumetanide (20 μmol l–1) did not affect the rate of cell shrinkage whether induced by mannitol, NaCl or KCl. It follows that NaCl and KCl are also close to 1 in the presence of bumetanide.
, 百拇医药
Discussion
The present investigation shows that the apical membranes of the PE cell culture and those of the TALH cell culture both contain a Na+–K+–Cl– cotransporter. 86Rb+ transport in both cell types was inhibited by bumetanide at concentrations of 10 μmol l–1 and by the removal of Cl– and Na+ ions from the bathing solution. The parameters obtained from the PE and TALH cultures could be compared directly: the bumetanide-sensitive rates of cation transport were about the same in the two tissues (Fig. 3B); both had relatively flat apical membranes defined by tight junctions; and neither of the cell types showed any unusual obstacles to intracellular diffusion, for example there were no pigment granules in the PE cells. The cotransporters of the two cell types responded differently, however, when challenged by an osmotic gradient; the cotransporter of the PE cells had a passive, bumetanide-sensitive, water permeability while that of the TALH cells did not.
, 百拇医药
Properties of water transport in the Na+–K+–Cl– cotransporter of the PE cells
The Na+–K+–Cl– cotransporter in the PE cell cultures transported water when exposed to a transmembrane osmotic gradient. It is possible to give a rough estimate of the unit water permeability per cotransporter. Given an activation energy of about 20 kcal mol–1 (Fig. 6) it follows that the protein turns over at a rate of about 100 s–1. The bumetanide-sensitive rate of K+ transport was about 1 nmol cm–2 min–1 (Fig. 3). From these numbers we calculate a density of 1011 cotransporters cm–2. The cotransporters contributed an Lp of about 0.8 x 10–4 cm s–1 (osmol l–1)–1 or 4.4 x 10–3 cm s–1. Accordingly, the Lp per molecule is 4 x 10–14 cm3 s–1, of the same order as that of aquaporins (for a review, see MacAulay et al. 2004). Apart from this, the properties of this water transport differed from that of an aqueous pore: (i) it showed an apparent saturation at higher osmotic gradients (Fig. 5); (ii) it was abolished by low doses of bumetanide (Fig. 5); (iii) it had high activation energy; and (iv) the reflection coefficients for NaCl and KCl were close to zero. These properties differ from those of passive, channel-mediated osmotic water transport, which does not saturate at increasing osmotic gradients, displays activation energies of 5–7 kcal mol–1 (Meinild et al. 1998) and has reflection coefficients for NaCl and KCl close to 1, underlining the fact that most water channels are impermeable to NaCl and KCl.
, 百拇医药
The high activation energy for the water transport and the apparent saturation observed at high osmotic gradients suggest that the water permeability is coupled to conformational changes in the protein. The activation energy for the water transport was similar to that determined for ion transport in the NKCC1 in the human red cell of 18 kcal mol–1 (Ellory & Hall, 1988), which indicates that water and ion transport are coupled processes. The low reflection coefficients for NaCl and KCl are another indication that the salts and water interact in the protein, in agreement with findings for the K+–Cl– cotransporter in the amphibian choroid plexus (Zeuthen, 1991a,b, 1994). In contrast to this, the Na+–K+–Cl– cotransporter residing in the TALH cells did not exhibit any water transport properties. The water transport of the apical membrane was not affected by bumetanide and showed no sign of saturation (Fig. 8). It had activation energies of about 11 kcal mol–1, which is similar to that of lipid bilayers. It should be emphasized that the comparison between the two tissues is based on practically equal rates of bumetanide-sensitive cation transport (Fig. 3B), but not on the number of cotransporters actually expressed.
, 百拇医药
Intracellular osmolarities and unstirred layers
To what extent are the data obfuscated by the changes in intracellular ion concentrations induced by the experiments themselvesWhen the cells are exposed to osmotic gradients implemented by NaCl or KCl there is an inward transport of salt by the Na+–K+–Cl– cotransporter. This would result in an increase in osmolarity at the inner aspects of the apical membrane (unstirred layers) and to an increase in the osmolarity of the whole cell. As a consequence, the resulting transmembrane osmotic difference would diminish and the calculated reflection coefficients for NaCl and KCl would be underestimated. However, a comparison of the transport rates and an assessment of intracellular diffusion coefficients in the PE cells and the TALH cells render such effects unlikely. According to these considerations, the TALH cells should be more prone to unstirred layer effects than the PE cells; yet, the values of the TALH cell membranes equalled one.
, 百拇医药
First, the measurements of in PE cells were done at 24°C while those of TALH cells were done at 37°C (Figs 7 and 9). At these temperatures the transport rate of 86Rb+ is four times higher in TALH cells than in PE cells (Fig. 3B) while the Lp of the TALH cells is only 25% lower than that of the PE cells. Taking these results together, the fact that the values of the TALH cell membranes are close to one does not reflect smaller transport rates.
Second, the overall increase in intracellular osmolarity that would result from the increased influx of ions is minute. Assume that the 86Rb+ permeability of PE cells increases to 10–6 cm s–1 under the hyperosmotic challenge. With an external K+ concentration of 2 mmol l–1 and assuming that K+ and Rb+ are treated similarly by the cotransporter, it can be calculated that the total influx, Jion, of Na+, K+ and Cl– amounts to 8 x 10–12 mol cm–2 s–1. Given a cell height of 7 μm and an intracellular free volume of 30% it follows that the intracellular osmolarity increases by a rate of no more than 5 μmol l–1 s–1. During a typical recording of the initial rates of cell shrinkage, lasting no more than 20 s, the increase would be about 0.1 mmol l–1 which is small compared to the imposed gradient of 25–400 mosmol l–1.
, http://www.100md.com
Third, the intracellular concentration gradient that could result from the increased influx of ions is also minute. With the numbers above, and an intracellular diffusion coefficient of 0.5 x 10–5 cm2 s–1 the total concentration gradient required to drive the flux through the cytoplasm would only be about 1.5 μmol l–1 (for calculation see Zeuthen, 1991b). It would require an effective reduction of the rate of diffusion in the cytoplasm by four orders of magnitude to produce significant unstirred layers.
, 百拇医药
Finally, the anatomical data do not suggest that diffusion should be slower in the PE cells than in the TALH cells. From the microscopic investigation of the PE and TALH cells (Fig. 2A and B) it appears that both cell cultures have fairly flat apical membranes connected by tight junctions; there was no folding in the apical membranes that could support significant standing gradients. Furthermore, there were no signs of dense packing of organelles, not even pigment granules in the PE cells. It is reasonable, therefore, to assume a diffusion coefficient in the cytoplasm of both cell types both for ions and small organic molecules of 0.5 x 10–5 cm2 s–1 which is about one-half to one-third of that found in free solutions (Zeuthen et al. 2002). It should be noted that cell nuclei offer no larger resistance to diffusion than cytoplasm (Palmer & Civan, 1977). The basolateral aspects of the cell layers overlap each other which show that cells extend underneath neighbouring cells. This tendency is more pronounced in the TALH cells than in the PE cells which partly results from the fact that the TALH cells are smaller than the PE cells (compare Fig. 2A and B). For reasons of cell geometry, therefore, diffusion through the TALH cell layer should be slower than through the PE cell layer. Thus, intracellular unstirred layer effects, if present, should be more pronounced in TALH cells than in PE cells.
, 百拇医药
General conclusions
The Na+–K+–Cl– cotransporter exists in two isoforms (reviewed by Russel, 2000). The isoform NKCC1 is found in epithelia responsible for secretion of salt and water, while NKCC2 is restricted to the thick ascending limb of the loop of Henle, which transports salt but not water. The Na+–K+–Cl– cotransporters investigated in the cultures of PE cells and TALH cells were expressed at levels where the two tissues had similar rates of K+ (Rb+) transport. Yet, they behaved differently with regard to water transport. As discussed above, it is unlikely that this should arise as a trivial result of the geometries of the two cell cultures. The finding suggests at least two possibilities. (i) The PE cell cultures express a NKCC1 isoform and the TALH cell cultures express the NKCC2 isoform. Compared to its K+ transport capacity, the NKCC1 has a large capacity for water transport while the NKCC2 has a small or no capacity for water transport. (ii) Both the PE and the TALH cultures express the same isoform, say NKCC1, but the cotransporter is in different functional modes in the two cell types. In the PE cells the mode allows for an interaction between salt and water, in the TALH cells the cotransporter is in a mode where there is no interaction. To answer these questions, it is necessary to determine the unit number and isoform type of the cotransporters active in the two types of tissues.
, http://www.100md.com
Footnotes
F. J. Alvarez-Leefmans and T. Zeuthen contributed equally to this work.
References
Alvarez-Leefmans FJ, Altamirano J & Crowe WE (1995). Use of ion-selective microelectrodes and fluorescent probes to measure cell volume. Meth Neurosci 27, 361–391.
Crook RB, Takahashi K, Mead A, Dunn JJ & Sears ML (2000). The role of Na-K-Cl cotransport in blood-to-aqueous chloride fluxes across rabbit ciliary epithelium. Invest Ophthalmol Vis Sci 41, 2574–2583.
, 百拇医药
Crowe WE, Altamirano J, Huerto L & Alvarez-Leefmans FJ (1995). Volume changes in single N1E-115 neuroblastoma cells measured with a fluorescent probe. Neuroscience 69, 283–296.
Dunn JJ, Lytle C & Crook RB (2001). Immunolocalization of the Na-K-Cl cotransporter in bovine ciliary epithelium. Invest Ophthalmol Vis Sci 42, 343–353.
Ellory JC & Hall AC (1988). Human red cell volume regulation in hypotonic media. Comp Biochem Physiol A 90, 533–537.
, 百拇医药
Gagnon MP, Bissonnette P, Deslandes LM, Wallendorff B & Lapointe JY (2004). Glucose accumulation can account for the initial water flux triggered by Na+/glucose cotransport. Biophys J 86, 125–133.
Hamann S (2002). Molecular mechanisms of water transport in the eye. Int Rev Cytol 215, 395–431.
Hamann S, Kiilgaard JF, la Cour M, Prause JU & Zeuthen T (2003). Cotransport of H+, lactate, and H2O in porcine retinal pigment epithelial cells. Exp Eye Res 76, 493–504.
, http://www.100md.com
Hamann S, Kiilgaard JF, Litman T, Alvarez-Leefmans FJ, Winther BR & Zeuthen T (2002). Measurement of cell volume changes by fluorescence self-quenching. J Fluoresc 12, 139–145.
Hamann S, la Cour M, Lui GM, Bundgaard M & Zeuthen T (2000). Transport of protons and lactate in cultured human fetal retinal pigment epithelial cells. Pflugers Arch 440, 84–92.
Hamann S, Zeuthen T, la Cour M, Nagelhus EA, Ottersen OP, Agre P & Nielsen S (1998). Aquaporins in complex tissues: distribution of aquaporins 1–5 in human and rat eye. Am J Physiol 274, C1332–C1345.
, 百拇医药
Hochgesand DH, Dunn JJ & Crook RB (2001). Catecholaminergic regulation of Na-K-Cl cotransport in pigmented ciliary epithelium: differences between PE and NPE. Exp Eye Res 72, 1–12.
Kedem O & Katchalsky A (1963). Permeability of composite membranes. Trans Faraday Soc 59, 1941–1953.
King LS, Kozono D & Agre P (2004). From structure to disease: the evolving tale of aquaporin biology. Nat Rev Mol Cell Biol 5, 687–698.
, 百拇医药
Layne J, Yip S & Crook RB (2001). Down-regulation of Na-K-Cl cotransport by protein kinase C is mediated by protein phosphatase 1 in pigmented ciliary epithelial cells. Exp Eye Res 72, 371–379.
MacAulay N, Gether U, Klaeke DA & Zeuthen T (2002). Passive water and urea permeability of a human Na+–glutamate cotransporter expressed in Xenopus oocytes. J Physiol 542, 817–828.
MacAulay N, Hamann S & Zeuthen T (2004). Water transport in the brain: Role of cotransporters. Neuroscience 129, 1029–1042.
, 百拇医药
Meinild AK, Klaerke DA & Zeuthen T (1998). Bidirectional water fluxes and specificity for small hydrophilic molecules in aquaporins 0–5. J Biol Chem 273, 32446–32451.
Palmer LG & Civan M (1977). Distribution of Na+, K+ and Cl– between nucleus and cytoplasm in Chironimus salivary gland cells. J Membr Biol 33, 41–61.
Plata C, Meade P, Vazquez N, Hebert SC & Gamba G (2002). Functional properties of the apical Na-K-2Cl cotransporter isoforms. J Biol Chem 277, 11004–11012.
, 百拇医药
Russell JM (2000). Sodium-potassium-chloride cotransport. Physiol Rev 80, 211–276.
Scott DM, Zierold K & Kinne R (1986). Development of differentiated characteristics in cultured kidney (thick ascending loop of Henle) cells. Exp Cell Res 162, 521–529.
Solenov E, Watanabe H, Manley GT & Verkman AS (2004). Sevenfold-reduced osmotic water permeability in primary astrocyte cultures from AQP-4-deficient mice, measured by a fluorescence quenching method. Am J Physiol Cell Physiol 286, C426–C432.
, 百拇医药
Von Brauchitsch DK & Crook RB (1993). Protein kinase C regulation of a Na+, K+, Cl– cotransporter in fetal human pigmented ciliary epithelial cells. Exp Eye Res 57, 699–708.
Zeuthen T (1991a). Water perrmeability of ventricular membrane in choroid plexus epithelium from Necturus maculosus. J Physiol 444, 133–151.
Zeuthen T (1991b). Secondary active transport of water across ventricular cell membrane of choroid plexus epithelium of Necturus maculosus. J Physiol 444, 153–173.
, 百拇医药
Zeuthen T (1994). Cotransport of K+, Cl– and H2O by membrane proteins from choroid plexus epithelium of Necturus maculosus. J Physiol 478, 203–219.
Zeuthen T, Zeuthen E & Klaerke DA (2002). Mobility of ions, sugar, and water in the cytoplasm of Xenopus oocytes expressing Na+-coupled sugar transporters (SGLT1). J Physiol 542, 71–87., http://www.100md.com(Steffen Hamann, José Jaim)
2 Departamento de Neurobiología, Instituto Nacional de Psiquiatría, Mexico DF, Mexico
3 Department of Pharmacology and Toxicology, Wright State University, Dayton, OH, USA
Abstract
Water transport properties of the Na+–K+–2Cl– cotransporter (NKCC) were studied in cultures of pigmented epithelial cells (PE) from the ciliary body of the eye. Here, the membrane that faces upwards contains NKCCs and can be subjected to rapid changes in bathing solution composition and osmolarity. The anatomy of the cultured cell layer was investigated by light and electron microscopy. The transport rate of the cotransporter was determined from the bumetanide-sensitive component of 86Rb+ uptake, and volume changes were derived from quenching of the fluorescent dye calcein. The water permeability (Lp) of the membrane was halved by the specific inhibitor bumetanide. The bumetanide-sensitive component of the water transport exhibited apparent saturation at osmotic gradients higher than 200 mosmol l–1. Cell shrinkages produced by NaCl or KCl were smaller than those elicited by equi-osmolar applications of mannitol, indicating reflection coefficients for these salts close to zero. The activation energy of the bumetanide-sensitive component of the Lp was 21 kcal mol–1, which is four times higher than that of an aqueous pore. The data suggest that osmotic transport via the cotransporter involves conformational changes of the cotransporter and interaction with Na+, K+ and Cl–. Similar measurements were performed on immortalized cell cultures from the thick ascending limb of the loop of Henle (TALH). Given similar overall transport rates of bumetanide-sensitive 86Rb+, the NKCCs of this tissue did not contribute any bumetanide-sensitive Lp. This suggests that the cotransporters of the two tissues are either different isoforms or the same cotransporter but in two different transport modes.
, http://www.100md.com
Introduction
It is now generally accepted that cotransporters support passive water transport (King et al. 2004). The unit water permeability is significant; for example, for the Na+-coupled glutamate transporter EAAT1 it is about 10 times larger than that of aquaporin 0 (AQP0) and 10 times smaller than that of AQP1 (MacAulay et al. 2002). When the large number of cotransporters present per cell is taken into account, their contribution to the overall passive water permeability is physiologically relevant. Moreover, the water permeability of the cotransporters depends on their conformational state, which suggests that they play a role in the rapid regulation of cell water permeability (reviewed by MacAulay et al. 2004). It has been suggested that in addition to passive translocation, cotransporters transport water actively, with a strict stoichiometry, along with the non-aqueous substrates. The coupling ratio is high; for instance, the Na+-coupled glucose cotransporter transports 210 water molecules for each turnover of the protein. However, this active mode of water transport is currently under debate (Zeuthen et al. 2002; Gagnon et al. 2004).
, 百拇医药
The purpose of this paper is to test whether the Na+–K+–Cl– cotransporter (NKCC) has passive water permeability. The NKCC1 isoform of this cotransporter can be studied in cell cultures from the ciliary epithelium of the eye (Layne et al. 2001; Hochgesand et al. 2001). In the mammalian eye the ciliary epithelium is responsible for the secretion of aqueous humour, transferring solute and water from the blood in the ciliary stroma into the posterior chamber of the eye (Fig. 1A). The circulating clear aqueous humour nourishes the cornea and lens and maintains the structural integrity and optical properties of the globe. The ciliary epithelium consists of two cell layers, the pigmented layer (PE) and the non-pigmented layer (NPE). As in other secretory epithelia NKCC1 is located in the basal membrane across which water enters the epithelium from the blood (Dunn et al. 2001). In the ciliary epithelium water flows from the PE cells via gap junctions into the NPE cells from where it continues into the aqueous humour, probably via aquaporins AQP1 and 4 (Hamann et al. 1998; Hamann, 2002). When PE cells are separated from the NPE cells and grown in culture on cover glasses, the membrane facing upwards contains NKCC1 and can be subjected to rapid changes in bathing solution composition and osmolarity, Fig. 1B. There are no known aquaporins in the PE cells (Hamann et al. 1998).
, 百拇医药
A, the secretory epithelium of the ciliary body consists of pigmented (PE) and non-pigmented (NPE) cell layers. Water is transported from the interstitial solution (stroma or blood side) into the PE cells and via gap junctions into the NPE cells from where it enters the aqueous humour via aquaporins (AQP1 and 4). In vivo, the Na+–K+–Cl– cotransporter is located at the blood-facing membrane. B, in cultured PE cells, however, the membrane that contains the Na+–K+–Cl– cotransporter faces upwards (becomes apical) and its water transport properties can be derived from initial rates of changes in cell volume induced by abrupt changes in bathing solution osmolarity or composition. Cell volumes were monitored as quenching of the intracellular fluorophore calcein via an inverted microscope. The Na+–K+–Cl– cotransporter in the apical membrane of cultured cells from the thick ascending limb of the loop of Henle (TALH cells) was studied by the same set-up.
, 百拇医药
A comparative study was performed on cell cultures from the epithelial cells of the thick ascending limb of the loop of Henle of the kidney (TALH cells) (Scott et al. 1986). In vivo, this epithelium has a Na+–K+–Cl– cotransporter at its apical membrane (isoform NKCC2) (Plata et al. 2002), and is responsible for creating the interstitial hyperosmolarity. The epithelium is relatively water impermeable. In culture, TALH cells grow with their apical membrane facing upwards and can be subjected to experiments similar to the PE cells (Fig. 1B).
, 百拇医药
In order to compare the properties of PE and TALH cultures, we first determined the geometry and ultrastructure by light and electron microscopy. Second, we determined the ion transport rates of the NKCCs in the two cell types. With these factors accounted for, the present paper shows that the NKCC of the PE cells transports water while the NKCC of the TALH cells does not. The passive water transport of the NKCC from the PE differs from that of simple water channels (i.e. aquaporins) by having high activation energies and low reflection coefficients for NaCl and KCl, and by exhibiting apparent saturation with increasing osmotic gradients.
, 百拇医药
Methods
Cell cultures
Human PE cells were generated from frozen stocks of cells previously dissected from 5-month-old aborted fetuses as described (Von Brauchitsch & Crook, 1993). Cells were thawed and plated in Lab-Tek 8-well chambered cover glasses (Life Technologies, Denmark). In each well the glass had a square cross-section of 0.81 cm2. PE cells were grown to confluence in culture medium (M199) supplemented with 15% fetal calf serum (FCS), 1 ng ml–1 basic fibroblast growth factor (bFGF), 300 μg ml–1 glutamine, 50 μg ml–1 gentamicin, and 2.5 μg ml–1 fungizone. bFGF was from Calbiochem (CA, USA), and the remaining components were from Invitrogen (Denmark). Cultures were kept at 37°C in a 5% CO2 environment, and the medium was changed every second day. Upon confluence, cultures were maintained in M199 supplemented as above but with 10% FCS and without bFGF. Confluent layers of PE cells in their fifth to eighth passages were used for experiments. Cell cultures were kindly made available by Dr R. B. Crook (UCSF, CA, USA). All protocols followed the Declaration of Helsinki. Informed consent for the use of fetal tissue in research was obtained, and institutional human experimentation committee approval was obtained for the use of human eyes.
, http://www.100md.com
Rabbit TALH cells were generated from frozen stocks of cells kindly supplied by Dr R. K. H. Kinne (Scott et al. 1986). Cells were thawed and plated on the chambered cover glasses and grown to confluence in culture medium (Dulbecco's modified Eagle's medium, high glucose) supplemented with 1% non-essential amino acids, 5% fetal calf serum, 1% L-glutamine, 1% pyruvate, 0.1% -mercaptoethanol, 10–7 mol l–1 arginine-vasopressin and 10–7 mol l–1 thyrocalcitonin. Vasopressin and thyrocalcitonin were from Sigma (Denmark); other reagents came from Invitrogen (Denmark). Cultures were kept at 37°C and 7.5% CO2 and the medium was changed every second day. Confluent layers of TALH cells in sixth to tenth passage were used for experiments.
, 百拇医药
Light and electron microscopy
Light microscopy. Cultures of PE and TALH cells were fixed in 2% glutaraldehyde and 1% methylene blue in 0.1 mol l–1 sodium cacodylate buffer for 60 min. Subsequently, the cultures were rinsed in 0.9% NaCl and examined in an inverted microscope. Micrographs were taken with a Nikon Coolpix 4500 digital camera.
Electron microscopy. Cultures of PE and TALH cells were left overnight in a solution containing 1% formaldehyde and 1% glutaraldehyde in 0.1 mol l–1 sodium cacodylate buffer (pH 7.3). The fixed cultures attached firmly to the surface of the cover glasses. After buffer rinse, crossing lines in the cell layers were cut with the sharp tip of a fine needle and small flakes of cells were flushed away from the supporting glass surface. These tissue samples were further processed for electron microscopy as previously described (Hamann et al. 2000).
, http://www.100md.com
Measurement of cell height in living cells
PE cells in seventh passage or TALH cells in ninth passage were incubated for 40 min at room temperature in control solution containing 4 μmol l–1 calcein-AM, the membrane-permeant non-fluorescent acetoxymethyl ester of calcein. Inside the cells esterases cleave off the acetoxymethyl groups and produce the membrane-impermeant fluorescent dye calcein. Confocal optical sections were performed using a Noran Odyssey Confocal Microscope equipped with a x40, NA 1.3 oil immersion objective (Nikon). Single wavelength excitation laser light of 488 nm and a slit size of 10 μm were used. The epifluorescence emission was directed through a 515 LP FITC filter and 3D image reconstruction was used to calculate cell heights.
, 百拇医药
Isotopic flux measurements
86Rb+ was obtained from Amersham Biosciences Ltd (UK) as RbCl with a specific activity of 1.5 mCi ml–1. Test solutions were prepared with specific concentrations of about 5 μCi ml–1, which correlates with Rb+ concentrations between 3 and 10 μmol l–1. In order to normalize and compare experiments, the radioactive decay of 86Rb+ was corrected for as follows. For each experiment, the uptake of 86Rb+ in μCi (or Rb+ in mol) cm–2 of cell layer was divided by the 86Rb+ in μCi (or Rb+ in mol) cm–3 test solution (final units: cm). For example, a Rb+ uptake of 10–3 cm means that the cell layer has taken up 10 pmol cm–2, given a test solution concentration of 10 nmol cm–3. Thus, uptake per second is equivalent to permeability, PRb, in units of cm s–1.
, http://www.100md.com
86Rb+ uptake was initiated by applying 0.25 ml of radioactive test solution to the tissue culture well. After the given time, uptake was stopped by washing twice with 0.5 ml of ice-cold saline. Subsequently, the cells were dissolved in 0.5 ml of 5% SDS (sodium dodecyl sulphate) and transferred to 10 ml of distilled water in plastic vials and counted on a Hewlett Packard -counter. In experiments where the uptake of 86Rb+ was studied in the presence of ouabain and/or bumetanide, the inhibitors were applied 30 s prior to the 86Rb+. This would prevent short-term uptake of tracer before the inhibition was fully effective. Ouabain was always present at concentrations of 100 μmol l–1. Bumetanide was used in concentrations of 10–100 μmol l–1 unless otherwise indicated. The bumetanide-sensitive component of the 86Rb+ influx was taken as the uptake observed when ouabain was present minus the uptake in the presence of ouabain plus bumetanide (Hochgesand et al. 2001).
, 百拇医药
Cell water volume measurements
The set-up was similar to the one previously described for measurements of water permeabilities Lp (Hamann et al. 2002, 2003). Some experiments were performed in Mexico City (set-up 1), others in Copenhagen (set-up 2). There was no difference between the results obtained in either laboratory. Chambers with confluent layers of PE or TALH cells were mounted on the stage of a Nikon Diaphot 300 inverted microscope (set-up 1) or a Zeiss Axiovert 10 inverted microscope (set-up 2). Only the solution bathing the upper or apical surface of the cell layer was changed (Fig. 1B). The perfusion solutions were fed into the chamber by gravity (set-up 1) or via a Gilson Minipuls 3 peristaltic pump (Biolab, Denmark, set-up 2) and collected by aspiration. At maximal flow rates (230 μl s–1), the solution at the level of the cells was exchanged 90% in 5 s. Most measurements were performed on confluent cell layers. A few measurements were done at the centre of confluent sheets of cells (about 100 μm by 100 μm). In these cases, concentration changes at the abluminal membrane can be disregarded; according to conventional diffusion theory, solution changes at the luminal side cannot affect the concentration at the abluminal side within the experimental period of 10 s.
, 百拇医药
The fluorescent dye calcein was used for measurement of changes in epithelial cell water volume (see below). The cells were incubated at room temperature in control solution containing 1–5 μmol l–1 calcein-AM. The latter is converted intracellularly to the membrane-impermeant fluorescent dye calcein, that accumulates intracellularly and reaches self-quenching concentrations, which in free solution are above 4 mmol l–1 (Hamann et al. 2002). After 40–60 min, the loading solution was washed out by control solution for 1 h before experiments were started.
, 百拇医药
In set-up 1, the microscope was attached to a DMX-1100 (SLM) monochromator, which provided the single wavelength excitation light of 497 nm (± 4 nm). Light pulses, lasting less than 300 ms, were delivered at a frequency of 0.5 Hz. A fast shutter was closed between the excitation pulses to reduce photobleaching and photodynamic damage. To further reduce exposure to UV light, a neutral density filter was used that transmitted only 5% of the light. In addition, an iris diaphragm confined the illuminated field to a light spot of 10–100 cells. A x40, NA 1.3 oil immersion Fluor objective (Nikon) was used for measurements. The epifluorescence emission was measured at 535 ± 12.5 nm, in defined regions, using a CCD camera (Hamamatsu C2400), and the output signals were digitally stored. Set-up 2 differed from set-up 1 in the following ways. The excitation light was directed through a 490 nm (± 5 nm) excitation filter, light pulses lasted less than 100 ms and were delivered at a frequency of between 0.3 and 1 Hz, the neutral density filter transmitted only 3% of the light, and the epifluorescence emission was measured between 503 and 530 nm using a photomultiplier tube (Zeiss).
, 百拇医药
The technique for measuring cell water volume changes using calcein self-quenching has been described in detail (Hamann et al. 2002). In brief, fluorescence signals are corrected for drift caused by volume-independent reduction in fluorescence intensity, and relative background caused by the fraction of intracellular calcein fluorescence, which is insensitive to changes in external osmolarity. This insensitive fraction presumably reflects calcein that is internally bound or compartmentalized (Alvarez-Leefmans et al. 1995; Crowe et al. 1995; Hamann et al. 2002; Solenov et al. 2004). In a given group of cells, the relative background was determined from calibrations obtained by applying hyposmotic and hyperosmotic solutions of known osmolarities. For subsequent experiments in the same group of cells, the fluorescence signals were corrected using the value of relative background obtained from the calibration. Drift-correction was performed for all individual experimental challenges. The change in cell water volume was calculated as
, 百拇医药
where Vt is the cell water volume at time t, Vt equals V0 at t = 0, Ft is the drift-corrected fluorescence at time t, F0 is the drift-corrected steady-state fluorescence before and after a given experimental challenge, and fb, the relative background, signifies the relative fluorescence insensitive to changes in external osmolarity. The influx of water per square centimetre of epithelium was calculated as
where 1/V0 x dVt/dt is the initial relative rate of change in cell water volume, h0 is the anatomical cell height at time zero, and is the ratio of cell water volume to the anatomical volume, which is taken as 1 (see Discussion).
, 百拇医药
The Lp was obtained according to eqn (3), from the initial rate of change in cell volume taken within 10–15 s, divided by the imposed osmotic gradient (),
under the assumption that the imposed osmotic gradient was instantaneous. However, the chamber was perfused at a rate of 12 ml min–1, and the solution at the level of the cells was exchanged 90% in 5 s (90 = 5 s; Hamann et al. 2003). This finite rate gives an upper limit to temporal resolution of the cellular volume changes. If the recorded cellular response has a 90 of 20 s, it can be calculated that the apparent Lp will be underestimated by a factor of about 1.1 compared to the true Lp that would have been determined if the solution change had been instantaneous. This was the case for PE cells at room temperature. If the recorded response had a 90 of 15 s the Lp would be underestimated by a factor of 1.25 (i.e. PE cells at 30°C); if the recorded response had a 90 of 10 s the Lp would have been underestimated by a factor of 1.8. The latter was the case for measurements of PE cells at 37°C: such data were not included.
, 百拇医药
Solutions
Control solution contained (in mmol l–1): 118 Na+, 6 K+, 1.2 Mg2+, 2.6 Ca2+, 118 Cl–, 1 PO42–, 25 Hepes, 5.5 glucose, and 29 mannitol. Anisosmotic solutions were prepared by mannitol addition or removal. In Cl–-free solutions equimolar amounts of Cl– were replaced with gluconate. In Na+-free solutions Na+ was replaced by N-methyl-D-glucamine, and K+ was replaced by mannitol in K+-free solutions. The salts were of analytical quality and the osmolarities were checked by means of a freezing-point depression osmometer. All solutions were titrated to pH 7.40. Calcein-AM was obtained from Molecular Probes (The Netherlands); other reagents were obtained from Sigma.
, 百拇医药
Statistics
Results are expressed as mean ± S.E.M.; comparisons were made using Student's unpaired t test, and a P value below 0.05 was used as a measure of statistical difference. Unless otherwise stated, the numbers given in parentheses correspond to the number of groups of cells tested. The number of cultures tested was, in most cases, larger than four. The changes in cell water volume were linear for at least the initial 10 s (4–11 data points depending on the frequency of data acquisition) and during this time, the initial rates of change in cell water volume were calculated by regression analysis.
, 百拇医药
Results
Anatomical properties
With light microscopy the PE cells appeared heterogeneous: locally the cultures were clearly multilayered, other fields appeared as partially confluent monolayers. The cells were large and elongated (15–20 μm x 40–50 μm) arranged in parallel (Fig. 2A). TALH cultures appeared as confluent cobblestone layers (Fig. 2B) of 10 μm x 15 μm cells. The position of nuclei and light lucent lines spaced at intervals corresponding to the dimensions of individual cells (Fig. 2B) indicating that the TALH cultures, at least in the focal plane, were two- or multilayered. At the ultrastructural level, both PE and TALH cells appeared moderately electron dense and rich in the usual organelles and microfilaments (Fig. 2, middle panels). Pigment granules at any developmental stage were not encountered in the cytoplasm of PE cells. Both cultures were organized as a transporting epithelium with luminal protrusions from the cells and with tight junctions sealing the luminal aspects of the intercellular clefts (Fig. 2, bottom panels). In the tissue samples chosen for electron microscopy both the PE and the TALH cell cultures were two- or three-layered. In addition, multiple flattened cellular processes occurred between and under the cells. The total height of the cell layers was between 5 and 10 μm. The cell height determined in the living cells by means of confocal microscopy was slightly larger than that estimated from electron microscopy. Accordingly, the height (h) of the PE cells was 7.8 ± 0.1 μm (n = 25) and that of the TALH cells was 6.7 ± 0.4 μm (n = 8).
, http://www.100md.com
A, PE cells. Top panel shows a light micrograph of cultured PE cells. The elongated cells are arranged in parallel with some degree of overlapping. Scale bar, 15 μm. Middle panel: electron micrograph of cultured PE cells, showing overlapping. Straight arrow marks microvilli-like processes from the luminal cell. Projections from neighbouring cells are indicated by arrowhead. Scale bar, 1 μm. Bottom panel: a tight junction between apposing cells in a culture of PE cells. Arrow marks membrane contacts between the luminal aspects of neighbouring cells. Scale bar, 0.1 μm. B, TALH cells. Top panel shows cultured TALH cells organized in a cobblestone layer. The lucent lines (arrows) possibly outline clefts between neighbouring cells in a deeper focal plane. Scale bar, 10 μm. Middle panel: electron micrograph of cultured TALH cells. Flattened cells and cellular processes are seen under the well-organized and specialized luminal cell layer. Arrow marks microvilli. Scale bar, 1 μm. Bottom panel: a junctional complex (arrow) at the luminal aspect of adjacent TALH cells. Scale bar, 0.1 μm.
, 百拇医药
Transport of 86Rb+
In order to evaluate and compare the rate of transport of the NKCC in PE and TALH cells, the rate of 86Rb+ uptake was monitored under various experimental conditions. 86Rb+ is known to replace K+ in cation–chloride cotransporters and the bumetanide-sensitive component of 86Rb+ uptake represents NKCC transport rates. This component was obtained as the 86Rb+ uptake in tissues treated with ouabain (100 μmol l–1) minus the 86Rb+ uptake in tissues treated with both ouabain (100 μmol l–1) and bumetanide (10–100 μmol l–1). Uptakes of 86Rb+ will be presented relative to the luminal concentration of 86Rb+ in permeability units (cm s–1). Most experiments were performed at both room temperature (20–24°C) and at 37°C in order to estimate Arrhenius activation energies (Ea). In general, for a given temperature, the 86Rb+ transport was found to be about 50% faster in TALH cells than in PE cells.
, http://www.100md.com
Uptake of 86Rb+ in tissues not treated with ion-transport inhibitors. In the absence of ouabain and bumetanide, the initial rate of 86Rb+ uptake in PE cells (37°C) was equivalent to a Rb+ permeability of (4.4 ± 0.3) x 10–6 cm s–1 (n = 4) and for TALH cells it was (6.9 ± 0.2) x 10–6 cm s–1 (n = 7), see Fig. 3A. The 86Rb+ uptake decreased mono-exponentially with time: for PE cells the time constant was 18 ± 0.3 min, and for TALH cells it was 64 ± 0.8 min. At 24°C the initial rates of isotope uptake were indistinguishable for the two cell types (about 2.0 x 10–6 cm s–1), but with longer uptake times a faster transport rate became evident for TALH cells.
, http://www.100md.com
The cell layers were bathed in control solution to which about 5 μCi ml–186Rb+ was added at time zero. A, uptake into cells not treated with inhibitors. Each point was determined from 4 or 8 tissues, except for PE at 37°C, where the 45 min and 90 min values are based on two single determinations. The data are fitted to mono-exponential curves; in all four cases the regression analysis gave R2 values larger than 0.99. B, the bumetanide-sensitive component of uptake was determined as the difference between uptakes determined in ouabain-treated tissues (100 μmol l–1; C), and in tissues treated with both ouabain and bumetanide (100 μmol l–1; D); data points were determined from between 4 and 8 experiments. S.E.M. values are not shown if smaller than points. Note the different scales in A compared to B, C and D. The ordinates can also be read as Rb+ uptake per square centimetre in units of 10 pmol cm–2 (see Methods).
, http://www.100md.com
The bumetanide-sensitive component of 86Rb+ uptake. As already mentioned, this component (Fig. 3B) was the difference between the uptake observed in ouabain-treated tissues (Fig. 3C) and those treated with both ouabain and bumetanide (Fig. 3D). The bumetanide-sensitive component was approximately linear during the first 10 min for PE cells and for the first 15 min for the TALH. The slope of these lines defines the bumetanide-sensitive 86Rb+ permeability (PRb,but). For PE cells PRb,but was (0.44 ± 0.09) x 10–6 cm s–1 at 24°C and (1.27 ± 0.07) x 10–6 cm s–1 at 37°C. For TALH cells PRb,but was (0.66 ± 0.07) x 10–6 cm s–1 at 24°C and (1.76 ± 0.09) x 10–6 cm s–1 at 37°C. In these experiments bumetanide was used at concentrations of 100 μmol l–1, but similar data were obtained with concentrations of 10 μmol l–1 (see Fig. 4).
, 百拇医药
A, PE cells were treated in paired experiments with either ouabain (control, C; n = 16) or with ouabain plus bumetanide. Bumetanide concentrations of 100, 10, 1 and 0.1 μmol l–1 were tested, n = 4 for each value. Bumetanide at 100 and 10 μmol l–1 inhibited maximally, about 50%, while 1 μmol l–1 inhibited significantly less than 10 μmol l–1 (P = 0.02). B, similar results were obtained with TALH cells, 1 μmol l–1 bumetanide inhibited significantly less than 10 μmol l–1 (P = 0.03). In C the effects of external Na+ or Cl– removal on 86Rb+ uptake by TALH cells are compared with those of bumetanide. TALH cells were treated with ouabain (control, C; n = 12) or in paired experiments with ouabain plus 100 μmol l–1 bumetanide, ouabain combined with Cl– removal (0 Cl–) or ouabain combined with Na+ removal (0 Na+). Cl– was replaced with gluconate and Na+ with choline ions. Uptakes lasted 10 min at 37°C.
, http://www.100md.com
PRb,but could also be determined without using ouabain, i.e. as the difference between the uptakes obtained in control solutions and those obtained in control solutions with bumetanide added. At 37°C, PRb,but for PE cells was (1.14 ± 0.21) x 10–6 cm s–1 (n = 7), and for TALH cells (1.40 ± 0.33) x 10–6 cm s–1 (n = 4). These values are similar to those obtained using ouabain (as above), but the relative errors are larger since the values arise as the difference between larger numbers.
, 百拇医药
The temperature dependence of PRb,but was determined from uptakes obtained at temperatures between 20 and 37°C. Ea for the TALH cells was 18.9 ± 4.2 kcal mol–1 (n = 21) and for the PE cells 25.3 ± 4.6 kcal mol–1 (n = 19). The values were not significantly different.
Bumetanide sensitivity. NKCC has a bumetanide affinity of about 1 μmol l–1 (Russell, 2000). In accordance with this, it has been shown that 10 μmol l–1 bumetanide abolishes transport by the NKCC in PE cells (Crook et al. 2000). Here we confirm and extend these findings to the NKCC in TALH cells (Fig. 4A and B). Tissues were inhibited by ouabain plus various concentrations of bumetanide; tissues inhibited by ouabain only, served as controls. For both PE and TALH cells, 10 μmol l–1 of bumetanide gave maximal inhibition of the bumetanide-sensitive uptake of 86Rb+ (Fig. 4A and B), while 0.1–1 μmol l–1 gave half-maximal inhibitions.
, 百拇医药
Requirements for Cl– and Na+. The bumetanide-sensitive component of 86Rb+ uptake into PE cells requires Cl– and Na+ in the test solution (Hochgesand et al. 2001). To test this for TALH cells, we compared the uptake of 86Rb+ under three conditions. Cells were bathed (i) in control solution with ouabain and 100 μmol l–1 bumetanide, (ii) in Cl–-free bathing solution plus ouabain (Cl– replaced by gluconate ions), or (iii) in Na+-free bathing solution with ouabain (Na+ replaced by choline ions), all experiments at 37°C (Fig. 4C). Absence of Cl– and Na+ from the bathing solution inhibited the uptake of 86Rb+ to the same degree as bumetanide. We conclude that the bumetanide-sensitive uptake of 86Rb+ into TALH cells has an absolute requirement for Cl– and Na+ consistent with the notion that it represents uptake via NKCC.
, http://www.100md.com
Bumetanide-sensitive uptake of 86Rb+ during hyperosmotic challenges. In order to estimate the rate of cotransport during the measurements of passive water permeability (Lp), the 86Rb+ uptake was measured in control conditions and compared to situations where 100 mmol l–1 mannitol or 50 mmol l–1 NaCl had been added. Application of 50 mmol l–1 KCl was not attempted due to the inherently low specific activity in such experiments. All these experiments were performed at 37°C. For both kinds of tissues, uptakes obtained under hyperosmolar conditions were significantly larger than those obtained in the control situation. For PE cells, the uptake was (0.42 ± 0.07) x 10–3 cm (n = 8) under control conditions, (0.72 ± 0.06) x 10–3 cm (n = 8) when 100 mmol l–1 mannitol had been added, and (0.69 ± 0.07) x 10–3 cm (n = 8) when 50 mmol l–1 Na+ had been added; uptakes lasted 10 min. For TALH cells, the uptakes were (0.25 ± 0.06) x 10–3 cm (n = 8) under control conditions, (0.54 ± 0.08) x 10–3 cm (n = 8) when 100 mmol l–1 mannitol had been added, and (0.72 ± 0.03) x 10–3 cm (n = 8) when 50 mmol l–1 Na+ had been added; uptakes lasted 2.5 min.
, http://www.100md.com
Water transport properties in the PE cells
The finite rate of solution change in the chamber (90 = 5 s) gives an upper limit to the rate of cellular volume changes that can be resolved (see Methods). Consequently, we have limited our investigations of PE cells to temperatures below 30°C.
Passive water permeability, Lp. When the osmolarity of the bathing solution was changed by , cell volume, Vt (or cell height, ht), changed towards a new steady state (Fig. 5A). The initial rate of volume change (dVt/V0dt) taken within the first 10–15 s multiplied by the cell height at time zero, h0 (7.8 μm, see above), gave the passive water permeability (eqn (3)). For the smaller osmotic challenges, Lp was independent of the magnitude and direction of as shown by the proportionality between the volume changes, JH2O = Lp = h0 dVt/V0dt, and the osmotic challenge (Fig. 5C). Under control conditions, at 24°C, Lp was (1.51 ± 0.12) x 10–4 cm s–1 (osmol l–1)–1 (n = 89). Larger osmotic gradients, 200 and 400 mosmol l–1, were less efficient in inducing water transport; apparently the water transport saturates at higher osmotic gradients.
, 百拇医药
A, relative cell water volume changes Vt/V0 in response to hyperosmotic challenges of 50, 100, 200 and 400 mmol l–1 mannitol. B, same as A, but in the presence of 10 μmol l–1 bumetanide. All recordings from the same tissue. C, water fluxes, JH2O, derived from experiments as in A and B, as a function of the imposed osmotic gradient osm. Paired experiments at 24°C from 4 tissues.
Effects of bumetanide. Bumetanide reduced Lp to about half, (0.81 ± 0.08) x 10–4 cm s–1 (osmol l–1)–1 (n = 8), and the Lp of the bumetanide-treated tissues was the same at small and large osmotic gradients (Fig. 5B and C). In most experiments bumetanide was used at concentrations of 10 μmol l–1, but concentrations of 20–100 μmol l–1 gave similar results. Bumetanide itself did not change the cell volume or cell height when added to the control isosmotic solution (data not shown), so the reduction in Lp was a result of a low value of dVt/V0dt.
, http://www.100md.com
Effects of temperature. The bumetanide-sensitive component of the Lp was large and significant at higher temperatures. In the present group of measurements it was 40% of the total Lp at 30°C, 24% at 22.9°C, but was insignificant at 15.6°C (Fig. 6). Under control conditions, the total Lp had an Ea of 9.3 ± 1.0 kcal mol–1 while Ea was reduced to 2.9 ± 1.0 kcal mol–1 in the presence of bumetanide. This shows that the bumetanide-sensitive component of the Lp has a high Ea. In the range 22.9–30.0°C it is calculated as 21 kcal mol–1. The data of the lower temperature range suggest an even higher Ea, but the scatter of the data at 15.6°C excludes an exact determination.
, 百拇医药
The Lp of PE cells were measured at 30.0, 22.9 and 15.6°C without and with bumetanide present (20 μmol l–1). Data are presented as ln(Lp) versus reciprocal absolute temperature. Lp values with and without bumetanide were significantly different at the higher temperatures but not at 15.6°C. Numbers of observations are shown in parentheses.
Reflection coefficients for NaCl (NaCl) and KCl (KCl). Lp was determined from additions of 100 mmol l–1 mannitol, while NaCl x Lp was determined from addition of 50 mmol l–1 NaCl and KCl x Lp by the addition of 50 mmol l–1 KCl. The values were obtained by division, both NaCl and KCl were close to 0.5 (data in Fig. 7). In the presence of bumetanide (20 μmol l–1) Lp decreased to about half (compare Fig. 5), while NaCl and KCl increased to 1. If we assume that the water transport across the membrane is shared about equally between the bumetanide-insensitive pathway in parallel with the bumetanide-sensitive pathway, it follows from irreversible thermodynamic theory of parallel pathways in composite membranes (Kedem & Katchalsky, 1963), that the reflection coefficients of the bumetanide-sensitive pathways for NaCl and KCl are both close to zero, while those of the bumetanide-insensitive pathway are close to one.
, http://www.100md.com
Upper panel, cell shrinkage induced by mannitol compared to shrinkage induced by isosmolar concentrations of NaCl, recordings from the same cells. Lower left panel, summary of Lp in control measured by mannitol (man) without and with bumetanide (+b). Lower right panel, reflection coefficient measured with 50 mmol l–1 NaCl or KCl without and with bumetanide. Recordings at 24°C. Numbers of tests in parentheses, error bars are S.E.M. as in Fig. 9.
Water transport properties in the TALH cells
, http://www.100md.com
Passive water permeability (Lp), effects of temperature and bumetanide. When the osmolarity of the bathing solution was changed by , cell volume changed towards a new steady state; an example is shown in Fig. 9. The initial rate of volume change taken within the first 10–15 s multiplied by the cell height at time zero, h0 (6.7 μm, see above), gave the passive water permeability Lp (eqn (3)). Lp was independent of the magnitude and direction of as shown by the proportionality between the volume changes, JH2O, and the osmotic challenge (Fig. 8).
, http://www.100md.com
Upper panel, cell shrinkage induced by 50 mmol l–1 NaCl, 50 mmol l–1 KCl and 100 mmol l–1 mannitol. Recordings obtained from the same group of cells. Lower left panel, Lp measured by mannitol (man) without and with bumetanide (+b). Lower right panel, reflection coefficient measured with 50 mmol l–1 NaCl or KCl without and with bumetanide. Recordings at 37°C. Numbers of tests in parentheses, error bars are S.E.M.
Water fluxes, JH2O, as a function of imposed osmotic gradient osm with and without bumetanide. Paired experiments from 4 tissues. Recordings at 24°C.
, 百拇医药
Under control conditions at 37°C, Lp was (0.87 ± 0.07) x 10–4 cm s–1 (osmol l–1)–1 (n = 5). From paired experiments at 24°C and 37°C, Ea was calculated as 11.1 ± 2.1 kcal mol (n = 5). Bumetanide (10–100 μmol l–1) had no significant effect on the Lp either at 24°C (Fig. 8) or at 37°C (Fig. 9). Lp showed no dependence on in the range 0–200 mosmol l–1 (Fig. 8).
Reflection coefficients for NaCl (NaCl) and KCl (KCl). There was no significant difference between the rate of cell shrinkage obtained by abrupt additions of 100 mmol l–1 mannitol, 50 mmol l–1 NaCl, or 50 mmol l–1 KCl (Fig. 9, upper panel). Accordingly, NaCl and KCl of the TALH cell membranes were both close to 1. The presence of bumetanide (20 μmol l–1) did not affect the rate of cell shrinkage whether induced by mannitol, NaCl or KCl. It follows that NaCl and KCl are also close to 1 in the presence of bumetanide.
, 百拇医药
Discussion
The present investigation shows that the apical membranes of the PE cell culture and those of the TALH cell culture both contain a Na+–K+–Cl– cotransporter. 86Rb+ transport in both cell types was inhibited by bumetanide at concentrations of 10 μmol l–1 and by the removal of Cl– and Na+ ions from the bathing solution. The parameters obtained from the PE and TALH cultures could be compared directly: the bumetanide-sensitive rates of cation transport were about the same in the two tissues (Fig. 3B); both had relatively flat apical membranes defined by tight junctions; and neither of the cell types showed any unusual obstacles to intracellular diffusion, for example there were no pigment granules in the PE cells. The cotransporters of the two cell types responded differently, however, when challenged by an osmotic gradient; the cotransporter of the PE cells had a passive, bumetanide-sensitive, water permeability while that of the TALH cells did not.
, 百拇医药
Properties of water transport in the Na+–K+–Cl– cotransporter of the PE cells
The Na+–K+–Cl– cotransporter in the PE cell cultures transported water when exposed to a transmembrane osmotic gradient. It is possible to give a rough estimate of the unit water permeability per cotransporter. Given an activation energy of about 20 kcal mol–1 (Fig. 6) it follows that the protein turns over at a rate of about 100 s–1. The bumetanide-sensitive rate of K+ transport was about 1 nmol cm–2 min–1 (Fig. 3). From these numbers we calculate a density of 1011 cotransporters cm–2. The cotransporters contributed an Lp of about 0.8 x 10–4 cm s–1 (osmol l–1)–1 or 4.4 x 10–3 cm s–1. Accordingly, the Lp per molecule is 4 x 10–14 cm3 s–1, of the same order as that of aquaporins (for a review, see MacAulay et al. 2004). Apart from this, the properties of this water transport differed from that of an aqueous pore: (i) it showed an apparent saturation at higher osmotic gradients (Fig. 5); (ii) it was abolished by low doses of bumetanide (Fig. 5); (iii) it had high activation energy; and (iv) the reflection coefficients for NaCl and KCl were close to zero. These properties differ from those of passive, channel-mediated osmotic water transport, which does not saturate at increasing osmotic gradients, displays activation energies of 5–7 kcal mol–1 (Meinild et al. 1998) and has reflection coefficients for NaCl and KCl close to 1, underlining the fact that most water channels are impermeable to NaCl and KCl.
, 百拇医药
The high activation energy for the water transport and the apparent saturation observed at high osmotic gradients suggest that the water permeability is coupled to conformational changes in the protein. The activation energy for the water transport was similar to that determined for ion transport in the NKCC1 in the human red cell of 18 kcal mol–1 (Ellory & Hall, 1988), which indicates that water and ion transport are coupled processes. The low reflection coefficients for NaCl and KCl are another indication that the salts and water interact in the protein, in agreement with findings for the K+–Cl– cotransporter in the amphibian choroid plexus (Zeuthen, 1991a,b, 1994). In contrast to this, the Na+–K+–Cl– cotransporter residing in the TALH cells did not exhibit any water transport properties. The water transport of the apical membrane was not affected by bumetanide and showed no sign of saturation (Fig. 8). It had activation energies of about 11 kcal mol–1, which is similar to that of lipid bilayers. It should be emphasized that the comparison between the two tissues is based on practically equal rates of bumetanide-sensitive cation transport (Fig. 3B), but not on the number of cotransporters actually expressed.
, 百拇医药
Intracellular osmolarities and unstirred layers
To what extent are the data obfuscated by the changes in intracellular ion concentrations induced by the experiments themselvesWhen the cells are exposed to osmotic gradients implemented by NaCl or KCl there is an inward transport of salt by the Na+–K+–Cl– cotransporter. This would result in an increase in osmolarity at the inner aspects of the apical membrane (unstirred layers) and to an increase in the osmolarity of the whole cell. As a consequence, the resulting transmembrane osmotic difference would diminish and the calculated reflection coefficients for NaCl and KCl would be underestimated. However, a comparison of the transport rates and an assessment of intracellular diffusion coefficients in the PE cells and the TALH cells render such effects unlikely. According to these considerations, the TALH cells should be more prone to unstirred layer effects than the PE cells; yet, the values of the TALH cell membranes equalled one.
, 百拇医药
First, the measurements of in PE cells were done at 24°C while those of TALH cells were done at 37°C (Figs 7 and 9). At these temperatures the transport rate of 86Rb+ is four times higher in TALH cells than in PE cells (Fig. 3B) while the Lp of the TALH cells is only 25% lower than that of the PE cells. Taking these results together, the fact that the values of the TALH cell membranes are close to one does not reflect smaller transport rates.
Second, the overall increase in intracellular osmolarity that would result from the increased influx of ions is minute. Assume that the 86Rb+ permeability of PE cells increases to 10–6 cm s–1 under the hyperosmotic challenge. With an external K+ concentration of 2 mmol l–1 and assuming that K+ and Rb+ are treated similarly by the cotransporter, it can be calculated that the total influx, Jion, of Na+, K+ and Cl– amounts to 8 x 10–12 mol cm–2 s–1. Given a cell height of 7 μm and an intracellular free volume of 30% it follows that the intracellular osmolarity increases by a rate of no more than 5 μmol l–1 s–1. During a typical recording of the initial rates of cell shrinkage, lasting no more than 20 s, the increase would be about 0.1 mmol l–1 which is small compared to the imposed gradient of 25–400 mosmol l–1.
, http://www.100md.com
Third, the intracellular concentration gradient that could result from the increased influx of ions is also minute. With the numbers above, and an intracellular diffusion coefficient of 0.5 x 10–5 cm2 s–1 the total concentration gradient required to drive the flux through the cytoplasm would only be about 1.5 μmol l–1 (for calculation see Zeuthen, 1991b). It would require an effective reduction of the rate of diffusion in the cytoplasm by four orders of magnitude to produce significant unstirred layers.
, 百拇医药
Finally, the anatomical data do not suggest that diffusion should be slower in the PE cells than in the TALH cells. From the microscopic investigation of the PE and TALH cells (Fig. 2A and B) it appears that both cell cultures have fairly flat apical membranes connected by tight junctions; there was no folding in the apical membranes that could support significant standing gradients. Furthermore, there were no signs of dense packing of organelles, not even pigment granules in the PE cells. It is reasonable, therefore, to assume a diffusion coefficient in the cytoplasm of both cell types both for ions and small organic molecules of 0.5 x 10–5 cm2 s–1 which is about one-half to one-third of that found in free solutions (Zeuthen et al. 2002). It should be noted that cell nuclei offer no larger resistance to diffusion than cytoplasm (Palmer & Civan, 1977). The basolateral aspects of the cell layers overlap each other which show that cells extend underneath neighbouring cells. This tendency is more pronounced in the TALH cells than in the PE cells which partly results from the fact that the TALH cells are smaller than the PE cells (compare Fig. 2A and B). For reasons of cell geometry, therefore, diffusion through the TALH cell layer should be slower than through the PE cell layer. Thus, intracellular unstirred layer effects, if present, should be more pronounced in TALH cells than in PE cells.
, 百拇医药
General conclusions
The Na+–K+–Cl– cotransporter exists in two isoforms (reviewed by Russel, 2000). The isoform NKCC1 is found in epithelia responsible for secretion of salt and water, while NKCC2 is restricted to the thick ascending limb of the loop of Henle, which transports salt but not water. The Na+–K+–Cl– cotransporters investigated in the cultures of PE cells and TALH cells were expressed at levels where the two tissues had similar rates of K+ (Rb+) transport. Yet, they behaved differently with regard to water transport. As discussed above, it is unlikely that this should arise as a trivial result of the geometries of the two cell cultures. The finding suggests at least two possibilities. (i) The PE cell cultures express a NKCC1 isoform and the TALH cell cultures express the NKCC2 isoform. Compared to its K+ transport capacity, the NKCC1 has a large capacity for water transport while the NKCC2 has a small or no capacity for water transport. (ii) Both the PE and the TALH cultures express the same isoform, say NKCC1, but the cotransporter is in different functional modes in the two cell types. In the PE cells the mode allows for an interaction between salt and water, in the TALH cells the cotransporter is in a mode where there is no interaction. To answer these questions, it is necessary to determine the unit number and isoform type of the cotransporters active in the two types of tissues.
, http://www.100md.com
Footnotes
F. J. Alvarez-Leefmans and T. Zeuthen contributed equally to this work.
References
Alvarez-Leefmans FJ, Altamirano J & Crowe WE (1995). Use of ion-selective microelectrodes and fluorescent probes to measure cell volume. Meth Neurosci 27, 361–391.
Crook RB, Takahashi K, Mead A, Dunn JJ & Sears ML (2000). The role of Na-K-Cl cotransport in blood-to-aqueous chloride fluxes across rabbit ciliary epithelium. Invest Ophthalmol Vis Sci 41, 2574–2583.
, 百拇医药
Crowe WE, Altamirano J, Huerto L & Alvarez-Leefmans FJ (1995). Volume changes in single N1E-115 neuroblastoma cells measured with a fluorescent probe. Neuroscience 69, 283–296.
Dunn JJ, Lytle C & Crook RB (2001). Immunolocalization of the Na-K-Cl cotransporter in bovine ciliary epithelium. Invest Ophthalmol Vis Sci 42, 343–353.
Ellory JC & Hall AC (1988). Human red cell volume regulation in hypotonic media. Comp Biochem Physiol A 90, 533–537.
, 百拇医药
Gagnon MP, Bissonnette P, Deslandes LM, Wallendorff B & Lapointe JY (2004). Glucose accumulation can account for the initial water flux triggered by Na+/glucose cotransport. Biophys J 86, 125–133.
Hamann S (2002). Molecular mechanisms of water transport in the eye. Int Rev Cytol 215, 395–431.
Hamann S, Kiilgaard JF, la Cour M, Prause JU & Zeuthen T (2003). Cotransport of H+, lactate, and H2O in porcine retinal pigment epithelial cells. Exp Eye Res 76, 493–504.
, http://www.100md.com
Hamann S, Kiilgaard JF, Litman T, Alvarez-Leefmans FJ, Winther BR & Zeuthen T (2002). Measurement of cell volume changes by fluorescence self-quenching. J Fluoresc 12, 139–145.
Hamann S, la Cour M, Lui GM, Bundgaard M & Zeuthen T (2000). Transport of protons and lactate in cultured human fetal retinal pigment epithelial cells. Pflugers Arch 440, 84–92.
Hamann S, Zeuthen T, la Cour M, Nagelhus EA, Ottersen OP, Agre P & Nielsen S (1998). Aquaporins in complex tissues: distribution of aquaporins 1–5 in human and rat eye. Am J Physiol 274, C1332–C1345.
, 百拇医药
Hochgesand DH, Dunn JJ & Crook RB (2001). Catecholaminergic regulation of Na-K-Cl cotransport in pigmented ciliary epithelium: differences between PE and NPE. Exp Eye Res 72, 1–12.
Kedem O & Katchalsky A (1963). Permeability of composite membranes. Trans Faraday Soc 59, 1941–1953.
King LS, Kozono D & Agre P (2004). From structure to disease: the evolving tale of aquaporin biology. Nat Rev Mol Cell Biol 5, 687–698.
, 百拇医药
Layne J, Yip S & Crook RB (2001). Down-regulation of Na-K-Cl cotransport by protein kinase C is mediated by protein phosphatase 1 in pigmented ciliary epithelial cells. Exp Eye Res 72, 371–379.
MacAulay N, Gether U, Klaeke DA & Zeuthen T (2002). Passive water and urea permeability of a human Na+–glutamate cotransporter expressed in Xenopus oocytes. J Physiol 542, 817–828.
MacAulay N, Hamann S & Zeuthen T (2004). Water transport in the brain: Role of cotransporters. Neuroscience 129, 1029–1042.
, 百拇医药
Meinild AK, Klaerke DA & Zeuthen T (1998). Bidirectional water fluxes and specificity for small hydrophilic molecules in aquaporins 0–5. J Biol Chem 273, 32446–32451.
Palmer LG & Civan M (1977). Distribution of Na+, K+ and Cl– between nucleus and cytoplasm in Chironimus salivary gland cells. J Membr Biol 33, 41–61.
Plata C, Meade P, Vazquez N, Hebert SC & Gamba G (2002). Functional properties of the apical Na-K-2Cl cotransporter isoforms. J Biol Chem 277, 11004–11012.
, 百拇医药
Russell JM (2000). Sodium-potassium-chloride cotransport. Physiol Rev 80, 211–276.
Scott DM, Zierold K & Kinne R (1986). Development of differentiated characteristics in cultured kidney (thick ascending loop of Henle) cells. Exp Cell Res 162, 521–529.
Solenov E, Watanabe H, Manley GT & Verkman AS (2004). Sevenfold-reduced osmotic water permeability in primary astrocyte cultures from AQP-4-deficient mice, measured by a fluorescence quenching method. Am J Physiol Cell Physiol 286, C426–C432.
, 百拇医药
Von Brauchitsch DK & Crook RB (1993). Protein kinase C regulation of a Na+, K+, Cl– cotransporter in fetal human pigmented ciliary epithelial cells. Exp Eye Res 57, 699–708.
Zeuthen T (1991a). Water perrmeability of ventricular membrane in choroid plexus epithelium from Necturus maculosus. J Physiol 444, 133–151.
Zeuthen T (1991b). Secondary active transport of water across ventricular cell membrane of choroid plexus epithelium of Necturus maculosus. J Physiol 444, 153–173.
, 百拇医药
Zeuthen T (1994). Cotransport of K+, Cl– and H2O by membrane proteins from choroid plexus epithelium of Necturus maculosus. J Physiol 478, 203–219.
Zeuthen T, Zeuthen E & Klaerke DA (2002). Mobility of ions, sugar, and water in the cytoplasm of Xenopus oocytes expressing Na+-coupled sugar transporters (SGLT1). J Physiol 542, 71–87., http://www.100md.com(Steffen Hamann, José Jaim)