ABSTRACT
The mechanism that underlies transcellular Ca2+ reabsorption in the kidney of the euryhaline teleost Oreochromis mossambicus was studied. Preparations of membrane vesicles made from the kidneys of freshwater-and seawater-adapted fish were more than sevenfold enriched in the basolateral plasma membrane marker Na+/K+-ATPase. Significant recovery of NADH– cytochrome c reductase enzyme activity and of oxalate-stimulated Ca2+ pump activities in the membrane preparations indicated that the membrane fraction was of endoplasmic reticular origin. Indeed, thapsigargin specifically inhibited Ca2+ pump activity that could be attributed to oxalate-permeable endoplasmic reticular fragments. Kinetic analysis of thapsigargin-insensitive
Ca2+ pump activity indicated the existence of a homogeneous, high-affinity, ATP-driven Ca2+ pump. No Na+-driven Ca2+ transport mechanism could be demonstrated. Plasma membrane Ca2+ pump activity was 56 % lower in preparations from seawater-adapted fish than in preparations from freshwater-adapted fish, suggesting a physiological role for this Ca2+ pump activity in renal Ca2+ handling by euryhaline species, with an involvement in the regulation of Ca2+ reabsorption.
Introduction
In euryhaline teleosts the kidney maintains Ca2+ homeostasis over a wide range of ambient Ca2+ concentrations. In sea water (SW) these fish are confronted with a constant Ca2+ influx from a hypercalcic environment and also with water efflux. Excess Ca2+ is actively excreted via the kidney, whilst urine flow is reduced. As a consequence, SW-adapted euryhaline teleost fish can produce urine with Ca2+ concentrations well above plasma levels (Björnsson and Nilsson, 1985; Schmidt-Nielsen and Renfro, 1975; Elger et al. 1987). In fresh water (FW), however, euryhaline fish need to minimize urinary Ca2+ loss, by reabsorbing filtered Ca2+. Thus, the tubular epithelium of the nephron mediates both Ca2+ secretion (in SW) and reabsorption (in FW). It follows that transfer of the fish from FW to SW (or vice versa) must have a pronounced effect on the properties of Ca2+ transport via the tubular epithelium.
In mammals, Ca2+ reabsorption is mainly mediated through a paracellular route and is driven by the transepithelial potential (lumen positive) created by Na+ and Cl- movement over the tubular epithelium; only in the distal nephron are hormonally controlled transcellular Ca2+ reabsorption mechanisms predominant (Friedman and Gesek, 1993). For freshwater fish, similar Na+-and Cl--dependent transepithelial potentials have been reported (Nishimura and Imai, 1982). This would suggest that, as in mammals, Ca2+ reabsorption in fish follows a paracellular route and is driven by the transepithelial potential. However, the principal function of the kidneys of freshwater fish is the excretion of excess water. Reabsorption of water must, therefore, be minimized and this will limit solute-linked, paracellular reabsorption of electrolytes.
The few studies published that have examined the mechanism of renal Ca2+ handling in fish seem to imply a dependence on ATP-regulated or ATP-driven mechanisms for secretion as well as for absorption of Ca2+. Renfro et al. (1982) proposed that in sea water winter flounder, Pseudopleuronectes americanus, intracellular ATP levels influenced Ca2+ secretion, whereas Na+ movement was not directly involved. More recently, in the kidney of Gillichthys mirabilis, a high-affinity Ca2+-ATPase was identified, with a presumed function in Ca2+ reabsorption (Doneen, 1993).
In the euryhaline teleost Oreochromis mossambicus (hereafter called tilapia), Ca2+ transport in both gill and intestinal epithelium is mediated through Ca2+-ATPase activity and Na+/Ca2+ exchange (Flik et al. 1985, 1990). The object of this study was to investigate the possible involvement of these transport mechanisms in Ca2+ handling in the kidneys of tilapia. Assuming that Ca2+ reabsorption is strongly enhanced in freshwater environments, the kidneys of euryhaline fish offer a model to study the contribution of Ca2+ transport mechanisms to this process.
Materials and methods
Tilapia, Oreochromis mossambicus (Peters), of both sexes were obtained from laboratory stock. Freshwater-adapted fish were kept in Nijmegen tap water ([Ca2+]=0.7 mmol l-1). Artificial sea water was prepared by adding Wimex sea salt (Wiegandt GMBH and Co., Krefeld, Germany) to tap water until a final concentration of 1.022 g l-1 was reached ([Ca2+]=10 mmol l-1). Fish were adapted to sea water over a 3-day period by the gradual infusion of sea water into tanks that were initially filled with fresh water. Once full-strength sea water was obtained, the water was constantly filtered and one-third of the volume was replaced weekly. Fish were kept for at least 3 weeks at full-strength salinity before use. The water temperature was 25 °C and the photoperiod was 12 h:12 h light:dark. Fish were fed Trouvit fish pellets (Trouw and Co., Putten, The Netherlands) at daily rations of 1.5 % of the fish total mass; the calcium content of the food was 0.34 mol kg-1.
Membrane isolation
Fish were killed by spinal transection and weighed (FW: 180±43 g, mean mass ± S.D., N=13; SW mean mass: 200±31 g, N=12). The abdominal cavity was cut open lengthwise, and the intestinal tract and swim-bladder removed. Kidney tissue was gently excised, weighed (FW mean mass: 0.328±0.091 g, N=13; SW mean mass: 0.399±0.118 g, N=12) and immediately transferred to ice-cold saline containing 150 mmol l-1 NaCl, 1 mmol l-1 Hepes, 1 mmol l-1 1,4-dithiothreitol (DTT) and 0.1 mmol l-1 EDTA, adjusted to pH 8.0 with Tris. All further steps were performed at 0–4 °C. The renal tissue was disrupted by 30 strokes with a Dounce homogenizer equipped with a loosely fitting pestle in isotonic sucrose buffer containing 250 mmol l-1 sucrose, 10 mmol l-1 Hepes, 1 mmol l-1 DTT and 100 trypsin inhibitor units ml-1 aprotinine, adjusted to pH 7.4 with Tris, with approximately 0.7 g of kidney tissue in 15 ml of buffer. This homogenization disrupts the complex renal tissue but leaves most blood cells intact. The homogenate was centrifuged for 10 min at 1400 g (Heraeus Sepatech Omnifuge 2.0RS, BS4402/A rotor, 2850 revs min-1) to remove nuclei, cellular debris and blood cells. The resulting supernatant was brought to 37 % sucrose (w/w) by mixing (5 strokes with the Dounce homogenizer) with 1.25 volumes of sucrose (60 % w/w) dissolved in 10 mmol l-1 Hepes/Tris (pH 7.4). 9 ml of this suspension was overlaid with 3 ml of isotonic sucrose buffer and centrifuged isopycnically for 90 min at 200 000g (Beckman L8–80, SW40 Ti rotor, 40,000 revs min-1). The membranes on the interface of the sucrose block and the isotonic buffer were collected in a volume of 0.5 ml and mixed with 12 ml of isotonic buffer, containing the basic ingredients of the assay medium:
150 mmol l-1 NaCl (for Na+/Ca2+ exchange) or 150 mmol l-1 KCl (for ATP-dependent Ca2+ uptake) and 20 mmol l-1 Hepes/Tris (pH 7.4). The membranes were pelleted by centrifugation for 30 min at 150 000g (50Ti rotor, 45 000 revs min-1), rinsed with isotonic buffer and resuspended by 20 passages through a 23 gauge needle in 0.5 ml of assay medium. Membrane preparations contained approximately 2.1 mg ml-1 protein and were used on the day of isolation without being frozen. Protein concentration was determined with a commercial reagent kit (Biorad), using bovine serum albumin as a reference.
Enzyme assays
The marker enzymes used were Na+/K+-ATPase for basolateral plasma membranes (Mircheff and Wright, 1976), aminopeptidase for brush border membranes (George and Kenny, 1973; Pfleiderer, 1970), NADH–cytochrome c reductase for endoplasmic reticulum (ER) (Omura and Takesue, 1970) and succinic acid dehydrogenase for mitochondrial fragments (Flik et al. 1983). Enzyme activities were assayed after treatment with a detergent, 0.2 mg ml-1 saponin (10 min, 25 °C), at a protein concentration of 1 mg ml-1, to unmask enzyme activity that was latent as a result of membrane resealing (Flik et al. 1990).
Membrane orientation was determined as described previously (Flik et al. 1990). The percentage of inside-out orientated vesicles (IOV) was determined on the basis of acetylcholine esterase activity, using digitonin (0.1 % w/v, 10 min at 25 °C) to unmask latent enzyme activity. Determination of the percentage of rightside-out orientated vesicles (ROV) was based on the specific trypsin sensitivity of the cytosol-oriented part of the Na+/K+-ATPase. Trypsin was used at 4500 BAEE units mg-1 membrane protein, where BAEE is N-benzoyl-arginine ethyl ester, for 30 min at 25 °C. After quenching the trypsin activity with 25 mg ml-1 soybean trypsin inhibitor, trypsin-insensitive Na+/K+-ATPase activity (representing the ROV membrane fraction) was revealed by treatment with detergent. In controls, trypsin inhibitor was added before the addition of trypsin to allow assessment of total Na+/K+-ATPase activity.
Ca2+ transport
Transport of Ca2+ was assayed by means of a rapid filtration technique (Flik et al. 1990; Van Heeswijk et al. 1984). The composition of the assay medium was 150 mmol l-1 KCl, 0.5 mmol l-1 EGTA, 0.5 mmol l-1 HEEDTA, 0.5 mmol l-1 nitrilotriacetic acid (NTA), 0.8 mmol l-1 free Mg2+, 3 mmol l-1 ATP (Tris-salt, Sigma), 20 mmol l-1 Hepes/Tris (pH 7.4) and 5 μg ml-1 oligomycin B. CaCl2 was added to obtain the calculated free Ca2+ concentrations of 1.0X10×8 to 2.0X10×6 mol l-1. 45CaCl2 (specific activity 24 TBq mol-1, Amersham) was added to the incubation medium to make up a radioactive concentration of approximately 74 kBq ml-1. In order to achieve a uniform specific activity of all the Ca2+ species in the incubation medium, the radiotracer was added at least 30 min prior to experimentation. ATP-dependent uptake was determined as the difference between Ca2+ uptake in the presence and in the absence of ATP. Pilot experiments demonstrated that Ca2+ uptake was linear for approximately 1 min. Therefore, 30 s incubations were used to estimate the initial rate of Ca2+ uptake. When Na+/Ca2+ exchange was assayed, ATP was omitted from the assay medium and in some cases KCl was replaced by NaCl. Incubations were carried out at 37 °C for optimum enzyme activity. Free Ca2+ and free Mg2+ concentrations were calculated according to Schoenmakers et al. (1992). In experiments where oxalate (2 mmol l-1) was included in the assay medium, Ca2+ and Mg2+ binding to oxalate was taken into account. Thapsigargin was added from a 1 mmol l-1 stock solution in ethanol. A23187 (calcimycin) was dissolved in dimethyl sulphoxide (DMSO) and added to the assay medium at 5 μg ml-1. The concentration of solvents in the assay medium did not exceed 0.1 % (v/v). The reaction was quenched by adding 1 ml of ice-cold stop buffer: 150 mmol l-1 KCl (or NaCl when Na+/Ca2+ exchange was assayed), 20 mmol l-1 Hepes/Tris, pH 7.4, 0.8 mmol l-1 MgCl2, 0.1 mmol l-1 LaCl3 to 0.15 ml of incubate. A volume of 1 ml, equivalent to 10–20 μg of membrane protein, was then filtered (Schleicher and Schuell, ME25; pore size: 0.45 μm). Filters were rinsed twice with 2 ml of stop buffer and dissolved in 4 ml of scintillation fluid. 45Ca specific activity was determined by counting the radioactivity in 0.05 ml of vesicle suspension. Radioactivity was determined in a Pharmacia Wallac 1410 liquid scintillation counter.
Calculations and statistics
Values are expressed as mean ± standard deviation (S.D.). Data were analyzed with a nonlinear regression data analysis program (Leatherbarrow, 1987). Data were analyzed statistically by the Mann-Whitney U-test or the Student’s t-test when appropriate. Statistical significance was accepted at P<0.05.
Results
Basolateral membrane isolation and orientation
In Table 1, protein recovery and the recovery and enrichment of several marker enzymes are listed for freshwater and seawater membrane preparations. Recovery and enrichment of the basolateral membrane marker Na+/K+-ATPase did not differ for freshwater and seawater preparations: specific enzymatic activity (SA; expressed as the rate of phosphate release) was 121±28 μmol h-1 mg-1 protein for freshwater preparations and 91±17 μmol h-1 mg-1 protein for seawater preparations (P=0.061). The aminopeptidase activity indicates that recovery of brush border membranes is relatively low. NADH– cytochrome c reductase recovery, however, was considerable and of special importance since this fraction exhibits Ca2+ transport activity (see next section for further details).
Plasma membrane orientation of freshwater preparations was 29±4 % IOV (N=5), 44±7 % ROV (N=7) and 27±11 % leaky membranes (calculated value). For seawater preparations, the corresponding percentages are 42±6 % (N=5), 34±7 % (N=5) and 24±13 %, respectively. The seawater preparations contained significantly more IOV orientated vesicles (P=0.004) and fewer ROV orientated vesicles (P=0.035) than the freshwater preparations.
Thapsigargin inhibition of ATP-dependent Ca2+ uptake
Fig. 1 shows the dose-dependent inhibition of Ca2+ uptake by thapsigargin. Thapsigargin partly inhibited ATP-dependent Ca2+ uptake. Maximal inhibition occurred at a thapsigargin concentration of 0.5 μmol l-1. Increasing the thapsigargin concentration to 5 μmol l-1 did not have any further effect on Ca2+ uptake rates. The residual, thapsigargin-insensitive, activities at a Ca2+ concentration of 500 nmol l-1 amounted to 26.1±0.4 % for freshwater preparations and 36.2±1.3 % for seawater preparations of the total Ca2+ uptake. These thapsigargin-insensitive activities divided by the respective IOV percentages for freshwater (29 %) and seawater (42 %) preparations yield the total thapsigargin-insensitive Ca2+ pump activities of the plasma membranes. These corrected thapsigargin-insensitive activities amount to, for freshwater preparations, 55 % (14 nmol min-1 mg-1 protein) and, for seawater preparations, 57 % (5.6 nmol min-1 mg-1 protein) of the total Ca2+ pump activity (consisting of the thapsigargin-sensitive activity plus the corrected thapsigargin-insensitive activity). Thapsigargin at 1 μmol l-1 completely abolished oxalate stimulation of ATP-dependent Ca2+ uptake in freshwater and seawater preparations (Fig. 2), indicating complete inhibition of endoplasmic reticulum (ER)-related Ca2+ uptake. Thapsigargin-insensitive Ca2+ accumulation was ATP-dependent and was reversed by the Ca2+ ionophore A23187 (Fig. 3).
Kinetics of thapsigargin-insensitive Ca2+ uptake
Kinetic analysis of the [Ca2+]-dependence of ATP-dependent Ca2+ uptake was performed on individual preparations. Fig. 4 shows the pooled data of five freshwater and five seawater preparations. For freshwater preparations a Vm of 4.50±0.89 nmol min-1 mg-1 protein and a Km of 57±17 nmol l-1 were calculated. For seawater preparations Vm decreased to 2.96±0.26 nmol min-1 mg-1 protein (P=0.008) and Km was 63±20 nmol l-1. When appropriate corrections for the percentages of IOV present are made, the calculated values for Vm are 16 nmol min-1 mg-1 protein for freshwater preparations and 7.0 nmol min-1 mg-1 protein for seawater preparations. This amounts to a 56 % decrease of renal Ca2+ pump activity upon transfer to sea water.
Na+/Ca2+ exchange
Table 2 summarizes data on Ca2+ uptake in membrane vesicles from freshwater-adapted tilapia, in the absence of ATP. Stimulation of Ca2+ uptake by applying a Na+ gradient could not be demonstrated. No dependence on Ca2+ concentration was indicated and data did not allow further kinetic analysis.
Discussion
To study the involvement of active transporters in Ca2+ reabsorption in the fish kidney, we isolated a membrane fraction from a renal tissue homogenate of tilapia. The separation techniques applied were based on studies of ion transporters in the mammalian kidney (reviewed by Mürer and Gmaj, 1986) and were successfully used in our laboratory for isolation of the basolateral membrane fraction of tilapia enterocytes (Flik et al. 1990). The enrichment factor and the specific enzymatic activity of the basolateral membrane enzyme marker Na+/K+-ATPase derived for our preparation are comparable to values reported earlier for tilapia intestinal membranes (Flik et al. 1990). Recovery, enrichment and specific activity of the Na+/K+-ATPase were not significantly different in freshwater or seawater renal membrane preparations of tilapia, although a decrease of Na+ pump activity upon transfer to high ambient salinity has been reported for several other euryhaline species (Doneen, 1993; Trombetti et al. 1990; Venturini et al. 1992). The high recovery of the ER enzyme marker NADH–cytochrome c reductase indicated that our final membrane preparation still contained a considerable proportion of ER fragments. This was confirmed by the 45Ca uptake studies, which showed that oxalate, which is preferentially transported into ER-derived vesicles (Ponnappa et al. 1981), was able to stimulate 45Ca accumulation by trapping 45Ca in vesicles. Attempts to enhance the purification of plasma membranes relative to the ER fraction by other initial tissue fragmentation techniques or by further centrifugation on sucrose block gradients ranging from 31 % to 43 %, or on a self-generating Percoll gradient as reported by Van Heeswijk et al. (1984), were unsuccessful (results not shown).
We used the highly specific ER Ca2+-ATPase inhibitor thapsigargin to inhibit 45Ca uptake into ER-derived vesicles (Thastrup et al. 1990). Our study shows that thapsigargin abolished oxalate-stimulated Ca2+ uptake, indicating complete inhibition of all ER-related Ca2+ pump activity. Maximal inhibition was reached at a thapsigargin concentration of 0.5 μmol l-1. The residual Ca2+ uptake was not affected by an increase in the thapsigargin concentration to 5 μmol l-1, indicating that at these low concentrations thapsigargin inhibits ER Ca2+-ATPase activity, but does not have an ionophoretic effect on the membrane vesicles, as reported by Favero and Abramson (1994). Thapsigargin-insensitive Ca2+ uptake was stimulated by ATP, and could be reversed by the addition of the Ca2+ ionophore A23187, indicating uphill Ca2+ transport. The Km values derived for freshwater and seawater preparations indicate that this Ca2+ pump is activated at intracellular Ca2+ levels. This strongly suggests that this Ca2+ pump activity originates from the plasma membrane fraction of the membrane preparation and reflects a mechanism for Ca2+ extrusion.
We were unable to demonstrate Na+/Ca2+ exchange activity in renal epithelium of tilapia: no significant difference was found between Ca2+ uptake in the absence and in the presence of a Na+ gradient. Results obtained in our laboratory, with membrane preparations of tilapia gill and intestine, using a similar experimental setup, show that application of a Na+ gradient stimulates Ca2+ uptake in these preparations by at least a factor of two. When assaying preparations of kidney and intestine from the same fish simultaneously, only in the intestine preparation could we demonstrate Na+-driven Ca2+ transport, which excludes a methodological origin for our inability to demonstrate Na+/Ca2+ exchange in renal preparations (results not shown). Furthermore, the addition of the K+ ionophore valinomycin to prevent the build-up of a potential difference did not affect Ca2+ transport rates (results not shown). We tentatively conclude that Na+/Ca2+ exchange activity is low or absent in tilapia kidney and, in the case of low activity, may be obscured by a proportionally large, non-specific Ca2+ uptake or Ca2+ binding.
Earlier studies report that euryhaline species maintain Ca2+ homeostasis upon transfer from fresh water to sea water by reducing renal Ca2+ reabsorption (Elger et al. 1987; Foster, 1976; Schmidt-Nielsen and Renfro, 1975). To evaluate the physiological significance of the identified Ca2+ pump in renal Ca2+ handling, we compared its activity in freshwater-and seawater-adapted tilapia. In our experiments, in which freshwater-and seawater-kidney preparations were isolated and assayed simultaneously, the renal Ca2+ transport capacity of freshwater-adapted fish significantly exceeded that of seawater-adapted fish, indicating that Ca2+-pump activity correlates with Ca2+ reabsorption. At the estimated cytosolic Ca2+ concentration of 100 nmol l-1 and at 25 °C, and correcting for the percentage of IOVs present and assuming an activation energy of 33±4 kJ mol-1 (Van Heeswijk et al. 1984), we calculate a Ca2+ pump activity of 5.9 nmol min-1 mg-1 protein for freshwater preparations. From a protein yield of 1.1±0.3 mg and the percentage found for basolateral membrane recovery, it follows that the Ca2+ transport capacity amounts to 6.0 μmol h-1 kg-1 fish.
This transport capacity of the Ca2+ pump is well in line with data on renal Ca2+ handling in freshwater-adapted fish. From the estimated glomerular filtration rate of 4 ml h-1 kg-1 fish for freshwater-adapted species (Hickman and Trump, 1969) and an ultrafiltrable plasma Ca2+ concentration of 1.82±0.29 mmol l-1 (N=3), a renal Ca2+ filtration rate of 7.3 μmol h-1 kg-1 fish can be calculated. Estimating the Ca2+ excretion at 2.5 μmol h-1 kg-1 fish (Butler, 1993; Elger et al. 1987; Oikari and Rankin, 1985; Schmidt-Nielsen and Renfro, 1975; Hickman, 1968), we calculate that a net reabsorption of 4.8 μmol h-1 kg-1 fish occurs, which amounts to 80 % of the transporting capacity of the Ca2+ pump. The high capacity of the Ca2+ pump relative to the estimated rate for net Ca2+ reabsorption suggests that in fish, in contrast to mammalian species (Friedman and Gesek, 1993), the transcellular route for Ca2+ reabsorption prevails. Paracellular reabsorption of solutes may be relatively small because of the low water permeability of the tubular epithelium (as reabsorption of water must be minimized) in the distal part of the nephron (Nishimura and Imai, 1982).
For seawater-adapted fish, a Ca2+ transporting capacity of 2.7 μmol h-1 kg-1 fish is calculated. The observation that seawater-adapted fish maintain Ca2+ pump activity even though they have no requirement for net Ca2+ reabsorption, suggests that part of the pumping activity is required for intracellular Ca2+ homeostasis. Consequently, the relative contribution of the Ca2+ pump to reabsorption may be smaller than estimated on the basis of total pump activity and, therefore, reabsorption may require additional Ca2+ transport mechanisms.
Ca2+ transport via the intestine and gill epithelium of freshwater-and seawater-adapted tilapia has been characterized by Schoenmakers et al. (1993) and Verbost et al. (1994), respectively. Whereas in the epithelium of the intestine Na+/Ca2+ exchange is the predominant mechanism for transcellular Ca2+ transport, in the gill epithelium the activity of the exchanger was estimated to be only 50 % of that of the ATP-driven pump (at prevailing cytosolic Ca2+ concentrations). This study indicates, however, that transcellular Ca2+ transport in renal epithelium of tilapia is fully dependent on an ATP-driven pump. The picture that emerges is that Ca2+ handling by the euryhaline tilapia is controlled by the regulated and differential expression of both Ca2+ pump and Na+/Ca2+ exchange activities in the three major Ca2+-transporting epithelia, i.e. gill, intestine and kidney. It enables the euryhaline tilapia to maintain Ca2+ homeostasis over a wide range of ambient Ca2+ concentrations, which is a prerequisite for successful salinity adaptation.
ACKNOWLEDGEMENTS
The authors gratefully acknowledge Professor J. J. M. de Goeij for his critical comments during the preparation of the manuscript.