This study examined the role of Ca2+ in regulatory volume decrease by Necturus erythrocytes. Hypotonic shock (50% tonicity) stimulated an increase in cytosolic free Ca2+, detected using epi-fluorescence microscopy and the fluorescent Ca2+ indicator fluo-4-AM (10 μM). A similar increase in fluorescence did not occur under isosmotic conditions, unless cells were exposed to the Ca2+ ionophore A23187 (0.5 μM). In addition, a low Ca2+ medium (amphibian Ringer solution with 5 mM EGTA), hexokinase (2.5 U/ml, an ATP scavenger), suramin (100 μM, a P2 receptor antagonist) and gadolinium (10μ M, a stretch-activated channel blocker) each inhibited the swelling-induced increase in Ca2+. Consistent with these studies, a low Ca2+ Ringer solution increased osmotic fragility, whereas volume recovery following hypotonic shock (measured with a Coulter counter) was potentiated with A23187 (0.5 μM). By contrast, a low Ca2+ extracellular medium or buffering intracellular Ca2+ with BAPTA-AM (100 μM) reduced the rate of volume recovery following hypotonic challenge. Finally, a low Ca2+ extracellular Ringer solution inhibited whole-cell currents that are activated during cell swelling (measured with the whole-cell patch clamp technique). Our results are most consistent with hypotonic shock causing an increase in cytosolic free Ca2+, thereby stimulating subsequent volume decrease.
The ability of animal cells to regulate their volume is a fundamental property common to a large number of cell types and has been extensively reviewed (Grinstein and Foskett, 1990; Hoffman, 2000; Land et al., 1998; Lewis and Donaldson, 1990; McCarty and O'Neil, 1992; Strange, 1994). Volume regulation is of importance in cells exposed to anisotonic extracellular conditions, as well as in cells where transport of solutes or pathophysiological conditions change intracellular osmolality (Land et al., 1998; Lang et al., 1998; McManus et al., 1995). Membrane transport pathways associated with volume regulation also have been implicated in processes as diverse as lymphocyte activation, cell cycle regulation and apoptosis (Land et al., 1998; Lang et al., 1998; McManus et al., 1995).
Exposure of vertebrate cells to a hypotonic solution results in an initial increase in cell volume owing to the relatively rapid influx of water. During continuous hypotonic stress, increases in cell volume are followed by a slower, spontaneous recovery towards the pre-shock level, a process known as regulatory volume decrease (RVD). This recovery is accomplished by selectively increasing the permeability of the plasma membrane during cell swelling to allow for efflux of specific intracellular osmolytes, thereby reversing the driving force for water influx (Grinstein and Foskett, 1990; Hoffman, 2000; Lewis and Donaldson, 1990; McCarty and O'Neil, 1992; Strange, 1994). Most vertebrate cells lose K+ and Cl- during RVD (Grinstein and Foskett, 1990; Hazama and Okada, 1990; Hoffman, 2000; Holtzman, 1991; Lewis and Donaldson, 1990; McCarty and O'Neil, 1992; Strange, 1994). This may occur by electroneutral ion transport pathways (Lewis and Donaldson, 1990) or by the separate activation of K+ and Cl- channels (Bergeron et al., 1996; Grinstein and Foskett, 1990; Hoffman et al., 1986; Lewis and Donaldson, 1990; Rubera et al., 1997). Loss of organic anions and osmolytes also may occur during RVD (Kirk and Strange, 1998).
The cellular mechanisms that activate and regulate permeability pathways during RVD are not completely understood and appear to differ between cell types (Land et al., 1998; Lewis and Donaldson, 1990; McCarty and O'Neil, 1992; Strange, 1994). Calcium, however, is a pivotal signaling agent for a wide variety of physiological processes. An elevation in the concentration of intracellular free Ca2+ from a resting level of approximately 50-100 nM to a stimulated level of 1-10 μM occurs as a rapid response to a number of cellular stimuli, including growth factors, neurotransmitters, hormones, peptides, toxins and cell swelling (Cheek et al., 1993).
Calcium has been shown to play a key role during cell volume regulation in a number of cell types (Foskett, 1994; Hoffman, 2000; McCarty and O'Neil, 1992; Tinel et al., 2000). In some instances, RVD strictly depends on Ca2+ influx across the plasma membrane (Montrose-Rafizadeh and Guggino, 1991; Wong et al., 1990), whereas in other cell types the Ca2+ response is mediated by Ca2+ release from intracellular stores (Negulescu et al., 1992; Terreros and Kanli, 1992). In addition, although it has been suggested that Ca2+ can directly activate ion channels during RVD (Hazama and Okada, 1990; McCarty and O'Neil, 1992), there also is evidence that several Ca2+-dependent intracellular messengers and enzymes (e.g., calmodulin, phospholipase A2, eicosanoids and protein kinases) are involved with cell volume regulation (Hoffman, 2000; McCarty and O'Neil, 1992; Rubera et al., 1997; Strange, 1994; Tinel et al., 2000). Furthermore, the cytoskeleton is thought to play a role in volume regulation and actin polymerization is dependent on intracellular Ca2+ levels (Hoffman, 2000; Mills et al., 1994).
Recent studies in our laboratory indicate that RVD in Necturus erythrocytes depends on a quinine-inhibitable K+ conductance that is regulated during cell swelling by a calmodulin-dependent mechanism (Bergeron et al., 1996) and by a 5-lipoxygenase metabolite of arachidonic acid (Light et al., 1997), as well as extracellular ATP activation of P2 receptors (Light et al., 1999; Light et al., 2001). Because calmodulin, phospholipase A2 and 5-lipoxygenase are Ca2+ sensitive (Holtzman, 1991) and because P2X receptors are ligand-gated, Ca2+-permeable cation channels (Fredholm et al., 1994), the purpose of this study was to investigate a more definitive role for Ca2+ in the regulation of cell volume following hypotonic challenge. To this end, we used four different approaches: (1) hemolysis studies to examine osmotic fragility; (2) fluorescence microscopy to detect changes in intracellular Ca2+ levels; (3) a Coulter counter to measure the volume of osmotically stressed cells; and (4) the whole-cell patch clamp technique to study membrane currents.
Materials and Methods
Mudpuppies (Necturus maculosus) were obtained from local vendors (Lemberger Co., Oshkosh, WI and Nasco Science, Ft. Atkinson, WI) and kept in well aerated, aged tap water at 5-10°C for no more than 6 days prior to use. They were anesthetized with 3-aminobenzoic acid ethyl ester (MS-222, 1%) and sacrificed by decapitation. Blood was obtained from a mid-ventral incision and collected into heparin (10,000 units/ml) coated tubes. Immediately following esanguination, the blood was spun in a centrifuge (Hermel-Z230, National Labnet Co., Woodbridge, NJ) at 100 g for 1 minute. The supernatant was aspirated and replaced with an equal volume of amphibian Ringer solution. This process of centrifuging and washing the cells was repeated twice.
Osmotic fragility was examined by determining the degree of cell lysis for a suspension of RBCs in hypotonic Ringer. The level of hemolysis was determined via a turbidity shift (cloudy to clear) that occurs when the integrity of the plasma membrane is compromised. This was detected with a spectrophotometer (Spectronic 20D, Milton Roy Co.) 10, 15 or 20 minutes after blood (30-50 μl) was added to saline solutions (3 ml) of different osmolalities and compositions. Spectrophotometric experiments were conducted at 625 nm because this wavelength provided the greatest difference in optical density (OD) between intact and lysed cells (Bergeron et al., 1996).
A percent hemolytic index (HI) was determined using the formula: HI(%)=(OD of test compound—OD of negative control)/(OD of positive control—OD of negative control) × 100, where OD of test compound refers to the OD of a cell suspension in diluted Ringer to which a test compound was added, OD of negative control refers to the OD of a cell suspension in diluted Ringer and OD of positive control refers to the OD of a cell suspension in distilled water. All reported hemolytic indices were calculated using a concentration of Ringer that gave an OD reading between 0.025 and 0.040. We chose this concentration range because it was sufficiently dilute to lyse approximately half the cells in suspension. Consequently, we could assess whether a test compound increased osmotic fragility by a subsequent reduction in OD compared to the negative control solution. Conversely, a rise in OD indicated that a test compound decreased osmotic fragility.
Intracellular free Ca2+ levels were monitored using a Nikon diaphot microscope, equipped with Hoffman DIC optics (400×) and epifluorescence (mercury lamp and FITC filter cube), and the fluorescent Ca2+ indicator fluo-4-AM (10 μM, Molecular Probes, Eugene, OR). This indicator has a high-affinity binding for Ca2+ (Kd=345 nM) and a very large fluorescence intensity increase in response to Ca2+ binding (>100 fold). The acetoxymethyl (AM) ester derivative permeates cell membranes and, once inside a cell, the lipophilic blocking groups are cleaved by non-specific esterases. This results in a charged form that is relatively impermeable.
Aliquots of fluo-4-AM were mixed with DMSO and diluted to give a final concentration of 10 μM. The non-ionic detergent Pluronic F-127 was used to assist in dispersion of the non-polar AM ester in aqueous media. This was accomplished by mixing an aliquot of AM ester stock solution in DMSO with an equal volume of 20% (w/v) Pluronic in DMSO before dilution into the loading medium. Cells were incubated with the AM ester for 60-90 minutes at room temperature [incubation at this temperature helps reduce potential compartmentalization of dye within cells (Molecular Probes). However, we did not test for compartmentalization and therefore cannot rule out this possibility]. Cells were then washed in indicator-free medium to remove any dye that was not specifically associated with the cell surface and then incubated for another 30-60 minutes to allow for complete de-esterification of intracellular AM esters.
Our experimental approach was designed to provide a qualitative assessment of fluorescence intensity while still maintaining identical imaging parameters for all conditions. This was accomplished by taking photographs with a Nikon camera (N2000) mounted directly to the microscope and using the same exposure time (3 seconds) and film speed (ASA 400) for all pictures. All photographs of swollen cells were taken 5 minutes after hypotonic shock, a point at which cells were maximally swollen.
Cell volume distribution curves were obtained by electronic sizing using a Coulter counter model Z2 with channelyzer (Coulter Electronics, Hialeah, FL). Mean cell volume was taken as the mean volume of the distribution curves. The diameter of the aperture tube orifice was 200 μm and the metered volume was 0.5 ml. Absolute cell volumes were obtained using polystyrene latex beads (20.13 μM diameter or 4.271×103 fl volume) as standards (Coulter). Experiments with the latex beads showed that measured volumes were unaffected by changes in osmolality and ionic composition within the ranges used for this study. Cell suspensions were diluted to give a final cell density of approximately 5,000-7,000 cells per ml.
Relative cell volume is defined as the average volume of cells compared to that in an isotonic medium. As described by others (Jorgensen et al., 1997; Wang et al., 1996), a percent volume recovery at X minutes after hypotonic exposure was calculated as [(Vmax—VX min)/(Vmax—V0)]×100, where Vmax is the peak relative cell volume, V0 is the initial relative volume (or one) and VX min is the relative cell volume measured X minutes after hypotonic exposure. A percent volume decrease was calculated as [(percent recoveryexperimental)/(percent recoverycontrol)]×100, where maximal recovery in hypotonic Ringer is 100%.
Patch pipettes were fabricated from Kovar sealing glass (Corning model 7052, 1.50 mm OD, 1.10 mm ID; Garner Glass, Claremont, CA) using a two-pull method (Narishige PP-7). Pipette tips were fire polished (Narishige MF-9) to give a direct current resistance of approximately 5-8 MΩ in symmetrical 100 mM KCl solutions (the large size of Necturus RBCs allows for the use of relatively wide tip pipettes). All pipette solutions were filtered immediately before use with a 0.22 μm membrane filter (Millex-GS, Bedford, MA), and the pipettes were held in a polycarbonate holder (E. W. Wright, Guilford, CT). Membrane currents were measured with a 1010 Ω feedback resistor in a headstage (CV-201A, Axon Instruments, Foster City, CA) with a variable gain amplifier set at 1 mV/pA (Axopatch 200A, Axon Instruments, Foster City, CA). The current signals were filtered at 1 kHz through a four-pole low-pass Bessel filter and digitized at 5 kHz with an IBM-486 computer.
Acquisition and analysis of data were conducted with P-Clamp® (version 6, Axon Instruments, Foster City, CA). Data were acquired during 100 msecond voltage pulses, and the command potential was set to -15 mV (close to the resting potential for RBCs) for 100 msec between each pulse. All voltage measurements refer to the cell interior.
RBCs, attached to glass coverslips (5 mm diameter, Bellco Biotech., Vineland, NJ) with poly-D-lysine (150,000-300,000; 1 mg/ml), were placed in a specially designed open-style chamber (250 μl volume, Warner Instruments Corp., Hamden, CT). The bath solution could be changed by a six-way rotary valve (Rheodyne Inc., Cotati, CA). The whole-cell configuration was achieved following formation of a gigaohm seal (cell-attached configuration) by applying suction to disrupt the patch of membrane beneath the pipette or by applying a large voltage (>200 mV) to the patch. A sudden increase in the capacitance current transient accompanied disruption of the membrane. (N.B., although the perforated patch method is preferable when trying to maintain normal intracellular Ca2+ levels, we were not successful at obtaining stable GΩ seals with either nystatin or amphotericin B in patch pipettes. Nonetheless, we were able detect a response that was internally consistent with our other three experimental protocols when we manipulated extracellular calcium levels using the standard whole-cell technique.)
Amphibian Ringer solution consisted of (in mM) 110 NaCl, 2.5 KCl, 1.8 CaCl2, 0.5 MgCl2, 5 glucose and 10 HEPES (titrated to pH 7.4 with NaOH, 235 mosm/kg H2O). A `low Na+ Ringer' was prepared by substituting choline chloride for NaCl (used for all experiments with gramicidin), and a 0.5× Ringer was obtained by reducing NaCl accordingly. A `low Ca2+ Ringer' was obtained by adding 5 mM ethylene glycol-bis(2-aminoethylether)-N,N,N′,N′-tetraacetic acid (EGTA) to amphibian Ringer, which reduced the free Ca2+ concentration to a calculated level of 35 nM (MaxChelator: http://www.standford.edu/∼cpatton/). Cells were loaded with 1,2-bis(o-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid tetra(acetoxymethyl) ester (BAPTA-AM) in a manner similar to that described above for fluo-4-AM. A stock solution of gramicidin was dissolved in methanol, and a stock solution of A23187 was prepared in DMSO, both at 1000× the final concentration and then diluted to give an appropriate working concentration. All stock aqueous solutions were diluted 100× to give an appropriate final concentration.
Patch pipettes were filled with an intracellular Ringer solution containing (in mM) 100 KCl, 3.5 NaCl, 1.0 MgCl2, 1.0 CaCl2, 2.0 EGTA, 5 glucose, 1.0 Mg-ATP, 0.5 GTP and 5.0 HEPES (pH 7.4 with KOH, 235 mosm/kg/H2O, calculated free Ca2+ level of 60 nM). During seal formation, the extracellular solution contained (in mM) 110 NaCl, 2.5 KCl, 1.8 CaCl2, 0.5 MgCl2, 5 glucose and 10.0 HEPES (pH 7.4). A hypotonic (0.5×) high K+ bath contained (in mM) 2.5 NaCl, 50 KCl, 1.8 CaCl2, 0.5 MgCl2, 5 glucose and 10.0 HEPES (pH 7.4, 120 mosm/kg/H2O).
For hemolysis experiments, pharmacological agents or their vehicle were present prior to the addition of cells. For cell volume studies, pharmacological agents were added with hypotonic exposure (0 minutes) or at peak cell volume (5 minutes after hypotonic challenge). Osmolality of solutions was measured with a vapor pressure osmometer (#5500, Wescor, Logan, UT). Chemicals were purchased from Sigma Chemical Co. (St. Louis, MO), Alexis Biochemicals (San Diego, CA), and ICN Pharmaceuticals Inc. (Costa Mesa, CA). All experiments were conducted at room temperature (21-23°C).
Data are reported as means±s.e.m. The statistical significance of an experimental procedure was determined by a paired Student's t-test or least significant difference test with paired design of analysis of variance (ANOVA)/multivariate ANOVA (MANOVA), as appropriate (Data Desk® software, Ithaca, NY). A P<0.05 was considered significant. Each animal served as its own control, and cell volumes at specific times were tested against each other. For patch clamp studies, each cell served as its own control.
Osmotic fragility studies
Although osmotic fragility depends on several factors, we first examined this property as one assessment of a cell's ability to regulate volume in a hypotonic medium. The OD was first measured at a concentration of amphibian Ringer (30.0±1.3 mosmol/Kg H2O) that caused approximately 50% of the cells in suspension to lyse (i.e., a 7.8-fold dilution of physiological concentrations, resulting in an extracellular Ca2+ concentration of 230 μM). Under these conditions, the OD was 0.078±0.004 (n=10 experiments, Fig. 1). (N.B., amphibian erythrocytes have a much lower osmotic fragility than their mammalian counterparts, which typically lyse with only a two-fold dilution.)
To determine whether osmotic fragility depended on extracellular Ca2+, we repeated this assay with a `low Ca2+ Ringer' (amphibian Ringer solution containing 5 mM EGTA). In this case, the OD measured at the same level of dilution as the control (resulting in an extracellular Ca2+ concentration of <10 nM) was 0.011±0.002 (n=10, P<0.001, Fig. 1), which gave a hemolytic index of 86%. By contrast, addition of the cationophore gramicidin (5 μM) to a diluted `low Ca2+ choline Ringer solution' blocked the inhibitory effect of bathing cells in low Ca2+ (n=10, Fig. 1; gramicidin is a cationophore that was used to maintain a high K+ permeability).
By contrast, we also conducted experiments that raised intracellular Ca2+ levels. Addition of the Ca2+ ionophore A23187 (0.5μ M) to the original diluted amphibian Ringer (230 μM extracellular Ca2+) increased the OD from 0.062±0.006 to 0.083±0.006, indicating a 34% decrease in lysed cells compared to the control (n=10, P<0.01, Fig. 1).
We next examined the effect of blocking P2 receptor activation. This was accomplished by hydrolyzing extracellular ATP with hexokinase (2.5 U/ml), an enzyme that traps ATP by transferring its γ-phosphoryl group to a variety of C6 sugars (Schwiebert et al., 1995). In this case, the OD decreased from 0.071±0.007 to 0.042±0.006 (n=10, P<0.01, Fig. 2), giving a hemolytic index of 41%. As illustrated in Fig. 2, the presence of gramicidin (5 μM) prevented the inhibitory effect of hexokinase (n=10). In addition, the general P2 receptor antagonist suramin [100μ M, (Fredholm et al., 1994)] decreased OD from 0.075±0.002 to 0.047±0.003 (n=10, P<0.001, Fig. 2), giving a hemolytic index of 37%. As with hexokinase, the presence of gramicidin (5 μM) prevented the inhibitory effect of suramin (n=10). In addition, A23187 (0.5 μM) also prevented suramin from increasing osmotic fragility (n=10). (N.B., variation in the average OD between control groups resulted from adding different volumes of blood for some experiments; volumes ranged from 30-50 μl). However, the same volume of blood per volume of extracellular medium was used within an experimental group.
Fluorescence microscopy studies
Having established that osmotic fragility was dependent on extracellular Ca2+ and activation of P2 receptors, we next determined whether intracellular Ca2+ levels changed during hypotonic challenge. As shown in Fig. 3A, cells loaded with fluo-4-AM (10 μM) did not display fluoresce under isosmotic conditions (n=6). By contrast, addition of A23187 (0.5 μM) stimulated fluorescence under isosmotic conditions, indicating cells were properly loaded with dye and we could detect qualitative changes in the level of intracellular free Ca2+ (n=6, Fig. 3B).
We next examined the effect of hypotonic shock on the level of cytosolic Ca2+. Exposure of cells to a hypotonic (0.5×) Ringer increased the level of fluorescence, indicating a rise in intracellular Ca2+ in swollen cells (n=6, Fig. 3A versus 3C). By contrast, exposing cells to a `low Ca2+ hypotonic Ringer' did not result in an increase in fluorescence associated with hypotonic challenge (n=4, Fig. 3C versus 3D). Additionally, the presence of hexokinase (2.5 U/ml) or suramin (100 μM) in hypotonic Ringer prevented an increase in fluorescence in swollen cells (n=4, Fig. 4). This inhibitory effect was blocked with the addition of A23187 (0.5 μM, n=4, Fig. 4). Further, the stretch-activated channel antagonist gadolinium (10 μM) (Yang and Sachs, 1989) inhibited the fluorescence associated with hypotonic swelling (n=4; data not shown). Finally, the calcium channel blockers verapamil (10 μM, n=4) and nifedipine (10 μM, n=4) had no effect on fluorescence following hypotonic challenge (data not shown).
Cell volume studies
When RBCs were placed in a hypotonic amphibian Ringer, they quickly swelled and then slowly and spontaneously decreased in volume (Fig. 5). When repeated with the `low Ca2+ hypotonic Ringer', the rate of volume recovery was reduced (n=8, P<0.05 after 40 minutes compared to control values), lowering the percent volume decrease to 57% of control values at 120 minutes. Similarly, when intracellular Ca2+ was buffered with BAPTA-AM (100 μM), volume recovery also was inhibited (n=8, P<0.05 after 40 minutes compared to control), reducing the percent volume decrease to 51% of control values at 120 minutes (Fig. 5). There was no significant difference at any time between the `low Ca2+ hypotonic Ringer' and the BAPTA-AM groups. Further, the inhibitory affect of BAPTA was prevented by adding gramicidin (5 μM, n=8, ionophore was added 5 minutes after hypotonic shock when cells were maximally swollen, Fig. 5, P<0.05 within 5 minutes following addition of ionophore compared to the other three groups).
We previously demonstrated that addition of A23187 to the extracellular solution potentiates volume recovery following hypotonic challenge (Light et al., 1999). In this study we assessed whether the effectiveness of pharmacologically increasing intracellular Ca2+ has a time-dependent nature. This was accomplished by adding A23187 (0.5 μM) 5, 40 and 70 minutes after hypotonic shock. In all three cases, this agent quickly reduced cell volume to near basal levels (n=8, Fig. 6, P<0.05 compared to control values immediately following addition of ionophore, regardless of when it was applied). In addition, although gadolinium (10 μM, n=6) inhibits volume recovery (Light et al., 1999), in this study we found both nifedipine (10 μM, n=6) and verapamil (10 μM, n=6) had no affect.
We also have shown in a previous study that the ATP scavenger hexokinase blocks cell volume recovery following hypotonic shock, and both A23187 and gramicidin prevent this inhibitory response (Light et al., 1999). Here we examined whether inhibition of RVD with hexokinase (2.5 U/ml) has a time-dependent response by adding it 5, 40 and 70 minutes after cell swelling. In all three instances, volume recovery was inhibited to virtually the same extent (n=8, Fig. 7, P<0.05 compared to the control immediately following addition of hexokinase at the 5 minute mark, and P<0.05 after 10 and 5 minutes when hexokinase was added at the 70 and 40 minute marks, respectively). There was no significant difference between the three hexokinase groups once this enzyme exhibited an inhibitory effect that was significantly different from control values.
Patch clamp studies
We previously reported that the K+ channel antagonist quinine, the calmodulin inhibitors pimozide and W-7, the stretch-activated channel blocker gadolinium and the P2 receptor antagonist suramin each inhibit whole-cell currents that are activated under hypotonic conditions (Bergeron et al., 1996; Light et al., 2001). Here we determined the effect of exposing swollen cells to a low Ca2+ hypotonic Ringer solution. For these experiments we used a high KCl Ringer in the pipette and a 0.5× KCl Ringer in the bath. The only major ions of significance with these solutions were K+ and Cl-, and the equilibrium potentials for perfect cation- and anion-selective conductances were -16.2 mV and +14.7 mV, respectively. Addition of 5 mM EGTA to the extracellular medium reduced whole-cell conductance by 72%, from 20.1±2.4 nS to 5.6±1.4 nS (n=5, P<0.01, Fig. 8). However this maneuver did not significantly shift the reversal potential.
The major finding of this study is that hypotonic swelling of Necturus erythrocytes stimulated a rise in intracellular Ca2+. This conclusion was demonstrated using epi-fluorescence microscopy and the fluorescent Ca2+ indicator fluo-4-AM. There was virtually no fluorescence for cells exposed to isosmotic conditions, indicating low levels of cytosolic free Ca2+. By contrast, hypotonic shock noticeably increased the level of fluorescence, demonstrating a rise in intracellular Ca2+ associated with cell swelling.
Our finding that hypotonic shock increased cytosolic Ca2+ is consistent with reports for several other cell types. For instance, reducing the osmolarity of the bathing medium causes intracellular Ca2+ to increase in cultured toad bladder cells (Wong et al., 1990), rabbit medullary thick ascending limb cells (Montrose-Rafizadeh and Guggino, 1991), gastric parietal cells (Negulescu et al., 1992) and Intestine 407 cells (Hazama and Okada, 1990). Recent reviews provide multiple examples of different cell types that elevate intracellular Ca2+ in response to acute osmotic swelling (Foskett, 1994; McCarty and O'Neil, 1992).
Our findings also show that extracellular Ca2+ was necessary for the swelling-induced rise in cytosolic free Ca2+. This was demonstrated by a lack of fluorescence associated with hypotonic shock when cells were exposed to a `low Ca2+ medium'. It should be noted, however, that once internalized, fluo-4 is diluted as cells swell in response to hypotonic shock. This in turn decreases fluorescence intensity in a corresponding manner. Because we did not measure relative fluorescence intensity concomitant with cell volume changes, it is possible there was a small rise in intracellular Ca2+ in swollen cells exposed to the `low Ca2+ Ringer' that we could not detect. However, we do not believe this was a significant problem in our study. For example, cells in a normal Ca2+ hypotonic Ringer displayed bright fluorescence even though they initially swelled to the same extent as cells in a low Ca2+ hypotonic Ringer. Conversely, although cells in a low Ca2+ hypotonic Ringer had an inhibited regulated volume decrease, they still initially swelled to approximately the same extent as cells in normal Ca2+ hypotonic Ringer, yet displayed virtually no fluorescence. Taken together, these observations support our conclusion that extracellular Ca2+ is necessary for an increase in fluorescence associated with cell swelling. We cannot, however, rule out the presence of a Ca2+-induced calcium release from intracellular stores. The importance of extracellular Ca2+ also has been shown for other cell types, including kidney cells (Montrose-Rafizadeh and Guggino, 1991) and toad bladder cells (Wong et al., 1990), which do not display a rise in cytosolic Ca2+ following hypotonic shock in the presence of low extracellular Ca2+.
Additionally, we found the rise in intracellular Ca2+ that accompanied hypotonic challenge also was correlated with subsequent volume decrease. For example, application of Ca2+ ionophore caused a decrease in osmotic fragility and also potentiated the rate of volume recovery following hypotonic challenge. In addition, we previously demonstrated that ionophore causes isosmotic cells to shrink (Light et al., 1999). It is possible, however, that the rise in Ca2+ resulting from A23187 may have stimulated solute efflux unrelated to a normal RVD process. However, our results are more consistent with a rise in cytosolic Ca2+ being necessary, at least in part, for regulated volume decrease. We make this claim because a low Ca2+ Ringer solution increased osmotic fragility and also reduced the rate of volume decrease in swollen cells, demonstrating the importance of extracellular calcium. Furthermore, buffering intracellular Ca2+ with BAPTA had a similar response to chelating extracellular Ca2+. Thus, if a rise in Ca2+ during cell swelling was an epiphenomenon accompanying RVD, as described by others (Pasantes-Morales and Morales-Mulia, 2000), then buffering cytosolic Ca2+ should not have inhibited volume decrease. Finally, a low Ca2+ hypotonic Ringer also reduced whole-cell currents that are normally activated during cell swelling.
We conducted experiments to assess the specific nature of a Ca2+ entry pathway during cell swelling. Interestingly, we found conventional Ca2+ channel blockers, such as verapamil and nifedipine, had no effect on fluo-4 fluorescence nor on cell volume recovery and whole-cell conductance. This is in contrast to several other reports that show these antagonists inhibit a rise in intracellular Ca2+ following hypotonic challenge (Montrose-Rafizadeh and Guggino, 1991; Wong et al., 1990). On the other hand, in this study gadolinium, a stretch-activated channel antagonist (Yang and Sachs, 1989), blocked the increase in fluorescence associated with cell swelling. We previously reported that this lanthanide increases osmotic fragility and inhibits volume recovery and whole-cell currents activated by hypotonic challenge (Bergeron et al., 1996; Light et al., 1999). Further, all the inhibitory actions of gadolinium were prevented with the Ca2+ ionophore, indicating that the Ca2+ entry and Ca2+-dependent steps are `downstream' to the gadolinium-sensitive site. On the basis of this information, a stretch-activated channel is a reasonable candidate for a permeability pathway for Ca2+ influx during cell swelling.
However, gadolinium has also been shown to inhibit P2X receptors (Nakazawa et al., 1997), which are ATP-gated, Ca2+-permeable, non-selective cation channels (Fredholm et al., 1994). Because we have shown that Necturus erythrocytes express P2X receptors and activation of these receptors play a role in cell volume decrease (Light et al., 1999; Light et al., 2001), it is possible the affect of gadolinium in this study was due to antagonism of P2X receptors. In fact, this supposition was supported by our studies using hexokinase and suramin. That is, both of these agents inhibited the fluorescence increase associated with cell swelling, indicating a lack of Ca2+ influx in the absence of P2 receptor activation.
Further, our previous findings have shown that hexokinase and suramin increase osmotic fragility, inhibit regulated volume decrease and block whole-cell currents associated with hypotonic swelling (Light et al., 1999; Light et al., 2001). Finally, in all cases the inhibitory nature of hexokinase and suramin was prevented with Ca2+ ionophore, indicating that Ca2+-dependent processes are `downstream' of the site of action of hexokinase and suramin, similar to observations made with cultured neurons (Garcia-Lecea et al., 1999). Taken together, our results are most consistent with extracellular ATP activation of P2X receptors leading to a rise in cytosolic Ca2+. It also seems likely that the receptor acts as a permeability pathway for Ca2+ influx during cell swelling.
It is interesting to note that the effect of a low Ca2+ Ringer on osmotic fragility was greater than the effect of blocking P2 receptors with suramin. We can only speculate as to why this might be so. For example, although suramin is considered a general P2 receptor antagonist (Fredholm et al., 1994), it may not completely block all P2 receptors. However, the observation that hexokinase has virtually the same effect as suramin, and a lesser effect than a low Ca2+ Ringer, suggests another explanation. There may have been a suramin- and hexokinase-insensitive Ca2+ permeability pathway that was open during cell swelling, independent of P2 receptors. Alternatively, a low Ca2+ Ringer may have additional effects on osmotic fragility, independent from regulation of permeability pathways, such as altering membrane integrity.
Some reports indicate that the increase in cytosolic free Ca2+ following cell swelling is transient (Negulescu et al., 1992; Wong et al., 1990). In addition, others have reported that there is a time-dependence concerning the activation of transport pathways used for RVD. For example, Ehrlich ascites tumor cells show activation of a Cl- conductance during cell swelling that becomes inactivated within the next 10 minutes (Hoffman et al., 1986). With this in mind, we examined whether there was a time dependence for Ca2+-stimulated volume decrease in Necturus erythrocytes. This was accomplished by adding Ca2+ ionophore at different points in time following hypotonic challenge. In all cases, this procedure potentiated volume recovery. In addition, we also examined the converse; that is whether lack of P2 receptor activation, and therefore Ca2+ influx, has a time-dependent nature. Application of hexokinase at different points in time following hypotonic shock also did not display a time-dependent affect, at least not during the time course of our experiments. Therefore, Necturus erythrocytes do not display a narrow window of time regarding activation of RVD mechanisms, at least as they relate to Ca2+.
Our finding that intracellular Ca2+ levels increase during cell swelling, which thereby stimulates volume decrease, is consistent with our previous observations that volume regulation is inhibited by agents that antagonize calmodulin (Bergeron et al., 1996). Additionally, we reported that volume regulation depends on phospholipase A2 and 5-lipoxygenase activity (Light et al., 1997), two Ca2+-dependent enzymes involved with arachidonic acid metabolism and the generation of leukotrienes (Holtzman, 1991). Thus, on the basis of our previous work, it is not surprising that we found buffering Ca2+ with BAPTA or EGTA inhibited volume decrease.
Our study also provides evidence that a rise in intracellular Ca2+ leads to an increase in K+ efflux. This was shown pharmacologically using the cationophore gramicidin with a choline Ringer. With this solution, K+ and Cl- were the only two permeable ions of significance, and addition of gramicidin ensured a continual high K+ permeability. Gramicidin consistently prevented the inhibitory effect of hexokinase and suramin, as well as that of BAPTA and EGTA. The reason for examining the effect of gramicidin 5 minutes after hypotonic shock is because that point in time corresponded with maximum cell swelling, indicating it took several minutes for endogenous K+ channels to activate. We also previously reported that addition of gramicidin causes cells to shrink under isosmotic conditions (Light et al., 1999). Taken together, our observations are consistent with this cell type having a low K+ permeability under isotonic conditions and an elevated K+ permeability during hypotonic stress in response to a rise in cytosolic Ca2+.
Moreover, our electrophysiological studies demonstrated that cell swelling leads to an increase in K+ permeability, via a conductive pathway, in response to a rise in intracellular Ca2+. We have previously shown that hypotonic shock activates a K+ conductance that is necessary for volume recovery (Bergeron et al., 1996), and this conductance is inhibited by agents that antagonize calmodulin and phospholipase A2 (Light et al., 1996; Light et al., 1997). In this study, we show that bathing cells in a low Ca2+ solution significantly decreased the swelling-activated whole-cell conductance. However, because the reversal potential did not change, we cannot rule out the possibility that lowering Ca2+ also inhibited a Cl- conductance concomitantly with a K+ channel. Further, it also is possible that an increase in intracellular Ca2+ may have led to membrane depolarization. This in turn could have resulted in activation of ion channels required for volume recovery.
In conclusion, hypotonic challenge activates a P2 receptor, which leads to a rise in cytosolic Ca2+, thereby stimulating volume decrease by activating a K+ conductance. The coupling of a P2 receptor to Ca2+ and cell volume decrease represent a novel mechanism for osmotic regulation of cell function.
We thank Robert L. Wallace (Ripon College) for helpful discussions and suggestions on the manuscript. Research support was provided by the National Science Foundation (MCB-9603568 and MCB-0076006). Portions of this study were presented in abstract form at Experimental Biology `02, New Orleans, LA., April, 2002.
- Accepted September 27, 2002.
- © The Company of Biologists Limited 2003