Actin-based cytoskeleton regulates a chloride channel and cell volume in a renal cortical collecting duct cell line.

The regulatory volume decrease (RVD) of a renal cortical collecting duct cell line (RCCT-28A) exposed to a hypotonic solution was studied using electronic cell sizing to measure cell volume and the patch clamp technique to measure Cl- channel activity. Results demonstrate that RVD was mediated in part by KCl loss through separate K+ and Cl- channels. The Cl- channel had a conductance of 305 pS and was activated by cell swelling, membrane stretch, and disruption of F-actin by dihydrocytochalasins. In contrast, stabilizing F-actin with phalloidin prevented swelling and stretch activation of the Cl- channel and inhibited the RVD. Thus, the state of actin polymerization regulates the probability of the 305 pS Cl- channel being open. Short actin filaments activate whereas long actin filaments inactivate the channel. Taken together, our studies suggest that RVD in this renal collecting duct cell line cell is mediated in part by a 305 pS Cl- channel, which is activated, during cell swelling, by a signaling pathway that includes disruption of F-actin.

The regulatory volume decrease (RVD) of a renal cortical collecting duct cell line (RCCT-28A) exposed to a hypotonic solution was studied using electronic cell sizing to measure cell volume and the patch clamp technique to measure C1-channel activity. Results demonstrate that RVD was mediated in part by KC1 loss through separate K' and C1-channels. The C1-channel had a conductance of 305 pS and was activated by cell swelling, membrane stretch, and disruption of F-actin by dihydrocytochalasins. In contrast, stabilizing F-actin with phalloidin prevented swelling and stretch activation of the C1-channel and inhibited the RVD. Thus, the state of actin polymerization regulates the probability of the 305 pS C1-channel being open. Short actin filaments activate whereas long actin filaments inactivate the channel. Taken together, our studies suggest that RVD in this renal collecting duct cell line cell is mediated in part by a 305 pS C1-channel, which is activated, during cell swelling, by a signaling pathway that includes disruption of F-actin.
A variety of second messengers have been suggested as potential mediators of RVD including calcium, calmodulin-dependent protein kinase, protein kinase C, CAMP and protein kinase A, and arachidonic acid and its metabolites (16). Filamentous actin (F-actin) may also play a role in the RVD. For . Preliminary data have been published in ab-* This work was supported by National Institutes of Health Grant stract form (14). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. example, cell swelling induced by exposure to a hypotonic SOlution activates RVD and decreases F-actin content in promyelocytic leukemic cells (17). The decline in F-actin content correlates with the rate of RVD (17). In addition, CAMP activates C1secretion in Madin-Darby canine kidney cells while decreasing cell volume and depolymerizing F-actin (18). Finally, RVD in melanoma cells devoid of actin-binding protein is impaired; however, genetic rescue with actin-binding protein resulted in recovery of cell volume following exposure to a hypotonic solution, a process dependent on activation of a K' channel (19). Presently, however, it is not known if the actinbased cytoskeleton contributes to RVD in renal epithelial cells.
Accordingly, the present study was conducted to determine if RVD in renal epithelial cells exposed to a hypotonic solution is mediated by C1channels and to determine if the actin-based cytoskeleton is an important element in the signal transduction pathway underlying RVD. In a previous study, we characterized C1-channels in RCCT-28A cells, a continuous renal cell line derived from rabbit cortical collecting duct (20, 21). These cells express a large conductance (305 pS) Cl-channel in the apical membrane and an outwardly rectifying (13/96 pS) C1channel in the basolateral membrane (20,22). Using this cell line as a model to study RVD in renal epithelial cells, we now report that hypotonic cell swelling activates the 305 pS C1channel in the apical membrane and that this channel contributes to RVD by a mechanism involving membrane stretch and disruption of F-actin.

MATERIALS AND METHODS
Cell Culture-As described previously, CCD cells were immunodissected from rabbit kidney and infected with an adenovirus 12-simian virus 40 hybrid, resulting in a continuous cell line designated . RCCT-28A cells were grown as described (20) and studied between passages 7 and 25 (20).
Patch Clamp Analysis of Single Cl-Channels-Single channel currents in the apical membrane were measured at 25 "C with a currentto-voltage converter (Warner Instrument Corp., model PC-501, Hamden, CTj, low pass filtered at 300 Hz, and digitized at 1 kHz with an Everex AT computer as described in detail previously using PClamp Version 5.51 software (Axon Instruments, Foster City, CAI (20, 23). Briefly, the single channel current amplitude was calculated by constructing amplitude histograms of single channel currents (i). Channels were considered open when the current was larger than i/2. Data were recorded for a minimum of one, 10-s trial every minute during control and experimental periods. In control and experimental periods, currents were recorded until a steady state was observed (steady state is defined here as a minimum of three consecutive 1-min periods in which the single channel open probability (Po) did not vary by more than 10%).
The Po was defined as the total time the channel was open divided by the total time of data collection. Single channel currents were also displayed continuously on a strip chart recorder throughout the control and experimental periods. The patch pipettes were filled with (in 111~): 140, NaCl; 5, KCl; 1, CaC12; 1, MgC12; 10, Hepes pH 7.4. The same solution was present in the bath during gigaohm seal formation. When the bath solution contained (in mM): 5, NaCI; 140, KC1; 0.0001, CaClz membrane patches were excised to form the inside-out configuration, (buffered with 2 m M EGTA); 1, MgC12; 10, Hepes pH 7.4.
Whole Cell Patch Clamp-Whole cell at C1-currents were measured 7081 a t 25 "C by the patch clamp technique as previously described (24). To generate I-V plots, the membrane was clamped to a holding voltage of 0 mV and then stepped in 20 mV increments between 2100 mV for 137.5 ms using PClamp 5.51 software. Three I-V plots were constructed every minute during control and experimental periods, and the plots were averaged to yield a mean I-V plot for each minute. In control periods, currents for I-V plots were recorded for a minimum of five consecutive minutes. During experimental maneuvers, currents for I-V plots were recorded until a steady state was observed (steady state is defined as three to five consecutive periods in which the slope of the I-V plot (i.e. conductance) did not vary by more than 10%). Whole cell conductance was calculated from the slope of the I-V plots. Currents were stored on the hard drive of an Everex AT computer and analyzed unfiltered using PClamp 5.51 routines. Current magnitude was measured during the last 20 ms of each 137.5-ms voltage pulse. In some experiments, drugs were added to the pipette solution. Upon obtaining the whole cell configuration, the drugs diffused into the cell with a t1,2 of 30 s to 1 min (25).
We measured cell capacitance as previously described ( Hepes; 1, CaC1,; 60, sucrose; pH 7.4. In preliminary experiments, we found that addition of sucrose to the bath solution prevented swelling activation of C1-currents and oscillations in C1-currents in the absence of experimental maneuvers. Because these solutions do not contain K+ or Na', the observed whole cell currents are referable primarily to C1-(see below). Measurement of Cell VoZume-Cell volume was measured by electronic cell sizing as described in detail previously (26)(27)(28). Briefly, cells were grown on tissue culture-treated polystyrene flasks and were harvested by a mild trypsinization protocol (0.05% trypsin and 0.53 m~ EDTA in a 0 Ca2+ and 0 Mg2+ phosphate-buffered saline solution). 8 x lo6 cells were suspended in 4 ml of an isotonic cell solution (in m: 70, NaCl; 2.4, KCl; 0.6, MgC12; 1.2, CaC12; 0.6, KH2P04; 7.5, Hepes; 10, glucose; 160, sucrose; pH 7.4; osmolality, 325 mOsmkg H20) for 60 min. Thereahr, 500-9 aliquots of the cells in the isotonic solution were suspended in 20 ml of the isotonic solution, a hypotonic solution (isotonic solution above without sucrose; pH 7.4; osmolality, 165 mOsmkg H20), or the hypotonic solution containing either C1-or K+ channel inhibitors. Because the isotonic and hypotonic solutions had the same ionic strength, any regulatory volume decrease resulting from KC1 loss cannot be attributed to changes in the chemical driving forces for these ions across the cell membrane. Cell volume was measured for each condition (duplicate 500-pl aliquots of the 4-ml cell solution; 1 x lo6 celldaliquot) a t 15 s and 2, 5, 10, 15, 25, and 60 min using a Coulter counter equipped with a Channelyzer (model ZBI-Chemlyzer 11; Coulter Corporation; Hialeah, FL) and a counting probe with a 100-pm aperture. Experiments were performed at 25 "C. Calibration of the instrument was verified by measuring the diameter of polystyrene microspheres (Epics Division of Coulter Corp.) of known dimensions as described in detail previously (27, 28).
In some experiments we examined the effect of phalloidin, a fungal toxin that stabilizes actin filaments, on cell volume. Because phalloidin is impermeable to renal epithelial cell membranes we used the bacterial toxin streptolysin 0 (Life Technologies, Inc.) to permeabilize the cell membrane to phalloidin. Cells were washed 3 times in a 0 Ca2+ and 0 Mg2+ phosphate-buffered saline solution and then incubated in a permeabilization buffer solution containing, in m: 137, NaCl; 100, Pipes; 2.7, KCl; 2.7, EGTA; 5.6, glucose; 1 Na,-ATP; pH 7.4,0.1% bovine serum albumin, 1 unitfml streptolysin 0, and 1 n~ phalloidin for 5 min at 25 "C. Control cells were incubated in the permeabilization buffer solution minus phalloidin. Thereafter, the cells were washed 3 times in the cell culture media containing fetal bovine serum and transferred to the isotonic cell volume solution for 30-45 min before measurements of cell volume were initiated. Preliminary experiments were conducted to determine the shortest incubation time and lowest streptolysin 0 concentration that permeabilized the cell membrane to phalloidin without causing cell death. All chemicals, except where noted, were purchased from Sigma.
Statistics-Data are presented as mean 2 S.E. The statistical significance of an experimental procedure was determined using paired Student's t test, unpaired Student's t test, or analysis of variance followed by a Student-Newmann-Keuls or Bonfferoni test as appropriate. A P value less than 0.05 was considered significant.

RESULTS
Characterization of RVD-The first series of experiments were conducted to determine if RCCT-28A cells volume regulate in response to a hypotonic solution and, if so, to determine if C1-channels participate in RVD. Fig. 1 illustrates the effects of a hypotonic solution (i.e. reduction of the bath osmolality from 325 to 165 mOsmkg H20 by removing sucrose from the isotonic solution) on cell volume. The volume response can be divided into three phases: 1) an initial and rapid, osmotic swelling reaching a peak 52% above control volume at 5 min; 2) a rapid shrinking in the following 10 min; 3) a more gradual decrease in cell volume reaching a steady state -15% above control volume a t 60 min.
To determine if the RVD involved activation of C1-channels, the C1-channel blocker, DIDS M), was added to the hypotonic bath solution. DIDS increased the initial, rapid osmotic swelling, attenuated the shrinking phase, and inhibited the gradual, slower decrease in cell volume observed in cells not exposed to DIDS (Fig. 1). To determine if the RVD also involved K+ channels, the K ' channel blocker, barium M), was added to the bath solution. The effects of barium were similar to those observed with DIDS ( Fig. 1). Taken together, these observations suggest that RVD involves KC1 loss through separate C1-and K+ conductive pathways. However, because DIDS inhibits C1-channels and the CI-/HC03-exchanger, it cannot be excluded that DIDS inhibited RVD by blocking a Cl-/HCO3exchanger (see below).
Effects of Cell Swelling on C1t Channels-To provide more definitive evidence that cell swelling activates a C1channel that contributes to RVD, we conducted cell-attached patch clamp recordings on the apical membrane. Fig. 2 illustrates the results of a representative experiment. Active C1-channels were rarely observed (<1% of patches) in cell-attached patches of cells bathed in the isotonic solution. When the osmolality of the bath solution was reduced from 300 t o 272 m o s m k g HzO, a 10% dilution routinely experienced by renal cortical collecting duct cells in the deep cortex, C1-channels activated after a variable delay of 2-5 min. The channels usually remained active for 5-10 min. The channels activated by the hypotonic solution were identified as 305 pS C1-channels described previously by our laboratory (20). The channels had a C1-to Na+ permeability ratio of 10   by the C1channel inhibitors DIDS (5 x M), diphenylamine apical membrane that were activated by cell swelling; however, carboxylic acid M), and 5-nitro-2-(3-phenylpropylamino) K' channels will be described in another report. benzoic acid (3 x M), but not by the K+ channel blocking the highest values between 6 and 10 min after exposure to the hypotonic solution, and declined slowly thereafter. Comparison of Figs. 1 and 3 reveals a good correlation between activation of the C1-channel and the onset of RVD. Because of this correlation and because DIDS inhibits both RVD and the C1-channel, it is reasonable to conclude that the 305 pS C1-channel contributes, a t least in part, to the RVD.
Effect of Membrane Stretch on the 305 p S C1C Channel-Some ion channels activated by cell swelling are also sensitive to membrane stretch (29-31). Accordingly, we examined the effects of membrane stretch on the 305 pS C1-channel. As illustrated by the representative current records in Fig. 4, application of negative pressure to the patch pipette increased the Po of the channel in inside-out patches. Increasing negative pressure from 0 to 5, 10 and 15 cm of HzO progressively increased the Po. Removal of negative pressure partially reversed the increase in the Po. Fig. 5 summarizes the effects of negative pressure on the 305 pS C1-channel in 19 experiments. Two types of responses were noted. Fig. 5A summarizes the effects of negative pressure on channels that were inactive following patch excision to form the inside-out configuration (i.e. Po = 01, and Fig. 5B summarizes the effects of negative pressure on channels with a Po > 0 following patch excision to the inside-out configuration. In membrane patches that contained quiescent channels ( n = l l ) , -2 cm of H 2 0 pressure failed to increase the Po. However, -5 cm of HzO pressure increased the Po. Ten cm of HzO negative pressure maximally increased the Po. Application of -15 cm of HzO pressure failed to increase Po further and often resulted in loss of the gigaohm seal. Two channels with an initial Po = 0 were not activated by negative pressure (data included in analysis). Fig. 5B summarizes the effects of negative pressure on channels with a Po > 0 following patch excision to the inside-out configuration ( n = 8). Two cm of HzO negative pressure increase the Po by -100%. Five to ten cm of H20 negative pressure failed to increase Po further. Sustained, ten cm of HzO negative pressure often resulted in the loss of the gigaohm seal. Thus, channels that were activated during conversion from the cell-attached to the inside-out configuration were more sensitive to small increases in negative pressure than channels that were not activated during formation of the inside-out configuration. In both subsets of channels, negative pressure increased the Po within 1 min, and release of pressure reduced the Po within seconds. Although the Po declined as negative pressure was released, the Po returned to control values only in the subset of channels that were activated by patch excision (Fig. 5B 1. Effects of F-actin Disruption on the 305 p S C1-Channel-Structural integrity of the plasma membrane is maintained in part by the actin-based cytoskeleton (32). Many transport proteins including epithelial Na' channels interact with the cytoskeleton (33) and disruption of F-actin activates Na' channels in renal A6 epithelial cells (34). Because F-actin is disrupted by cell swelling (6, 17,18,[35][36][37][38], we tested the hypothesis that the 305 pS C1-channel is activated by disruption or fragmentation of F-actin. To this end, we used cytochalasins to disrupt andor fragment F-actin (39-43) in inside-out patches of the apical membrane. A representative experiment illustrating the effects of dihydrocytochalasin B (DHCB) on a C1-channel is shown in Fig. 6. DHCB increased the Po of quiescent and excision-activated C1-channels after a delay of 3-5 min. The Cytochalasins B and D also activated the C1-channel (Fig. 7). In contrast, chaetoglobosin C M), a cytochalasin that does not depolymerize actin, failed to increase the Po (n = 5).

Cytoskeleton and Chloride Channels 7085
Because disruption of F-actin activates the C1-channel, it follows that repolymerization of F-actin should inactivate the channel. Accordingly, we examined the effects of A T P -M e , a catalyst of actin polymerization, on the C1-channel after the Po was increased by DHCB. ATP M) reduced the Po of channels activated by DHCB (Fig. 8). Washing DHCB from the bath in the absence of ATP did not reduce Po. In previous experiments we showed that ATP-Mg2' alone had no effect on the Po of active 305 pS C1-channels (20). Thus, it is unlikely that the action of ATP involved the activation of protein kinases or ATPases. Taken together, these experiments are most consistent with the conclusion that disruption of F-actin activates the 305 pS C1-channel, whereas polymerization of actin inactivates the channel.
We also conducted experiments to determine if negative pressure applied to the membrane patch activates C1-channels by a mechanism involving the disruption of F-actin. We examined the effect of negative pressure on channels in membrane patches exposed either to DHCB or to phalloidin, which stabilizes F-actin (39,40,44,45). As summarized in Table I, -10 cm of H 2 0 suction failed to increase the Po of channels preactivated with DHCB. In contrast, DHCB added to membrane patches containing channels preactivated by negative pressure (-10 cm of H 2 0 pressure) increased the Po (Table I). These observations suggest that DHCB is more effective than negative pressure in activating the C1-channel. In contrast, when membrane patches were treated first with phalloidin (2 x 10"' M), -10 cm of H 2 0 suction failed to activate the 305 pS C1channel (Table I). These results suggest that the mechanism of activation of the 305 pS C1-channel by membrane stretch involves disruption of the cortical actin filament network.
Whole Cell C1-Conductance-Whole cell C1-currents were monitored to provide additional support for our single channel records indicating that cell swelling and disruption of F-actin activates C1-channels. Under the experimental conditions described under "Materials and Methods," whole cell currents in RCCT-28A cells are referable exclusively to C1- (22). A 10% reduction in bath osmolality (350-315 mOsmkg H20; dilution with distilled H20) increased the whole cell C1-conductance from a steady-state value of 8.6 2 2.3 nS to a peak value of 11.9 2 2.8 nS at 1.5 min ( Fig. 9; p < 0.05; n = 4). A 25% reduction in bath osmolality (350-262 mOsm/kg H20, bath dilution with distilled H20) increased the C1-conductance from a steadystate value of 9. the increase in whole cell C1-conductance from 73 2 10% to 22 2 15% in response to a reduction in bath solution osmolality (325-165 mOsm/kg H20; sucrose removed from the isotonic bath solution was p < 0.05; n = 3). F-actin Regulates Swellingactivated Whole Cell Cl-Currents-Additional studies, using the whole cell patch clamp technique, were conducted to determine if the actin-based cytoskeleton participates in the swelling-induced stimulation of C1-channels (Table 11). DHCB elicited a sustained increase in C1-conductance of cells bathed in the isotonic solution. In DHCB-treated cells, however, the hypotonic solution failed to elicit an additional increase in whole cell C1conductance. Thus, when actin is disrupted by DHCB, cell swelling does not activate C1-channels. This observation is consistent with our single channel experiments demonstrating that cell swelling activates C1-channels by a mechanism involving the disruption of F-actin. To provide additional support for this conclusion, experiments were performed with phalloidin (Table 11) din also blocked the regulatory volume decrease in cells exposed to a hypotonic solution (ie. reduction of the bath osmolality from 325 to 165 mOsm/kg H20 by removing sucrose from the isotonic solution). As depicted in Fig. 10, cells treated with phalloidin and placed in the hypotonic solution exhibited an initial and rapid osmotic swelling reaching a peak 56% above control volume at 5 min. However, in contrast to control cells placed in the hypotonic solution, cells treated with phalloidin and placed in the hypotonic solution did not exhibit a regulatory volume decrease. Taken together, these results indicate that disruption of F-actin is required for swelling-induced activation of C1-channels and RVD. Is Exocytosis Involved in RVDB-In many epithelia, stimulation of ion transport involves the exocytic insertion of transport

MI.
The voltage across the membrane patch was -20 mV. *, p < 0.05 versus control.

ATP
proteins from a vesicular pool into the apical plasma membrane by a microtubule-dependent mechanism (46,471. To determine if swelling-induced activation of Cl-currents occurs by a mechanism involving exocytosis and the delivery of 305 pS C1channels, andor a regulatory protein, from an intracellular vesicular pool to the plasma membrane we monitored exocytosis by measuring capacitance using the whole cell patch clamp technique. Membrane capacitance is proportional to the cell surface area and can be monitored to detect increases in plasma membrane area as intracellular vesicles fuse with the plasma membrane. As illustrated in Fig. 11, a reduction in the osmolality of the bath solution (25% reduction in osmolality by distilled HzO) increased C1-conductance; however, whole cell capacitance did not change. Similar observations were made in cells exposed to a 10% reduction in osmolality ( n = 4, data not shown).  Colchicine ( M), a drug that inhibits the polymerization of tubulin monomers and blocks microtubule-dependent exocytosis, had no effect on basal C1-conductance (15.0 5 1.3 nS in control versus 12.9 2 1.6 nS in the presence of colchicine) or on the increase in C1-conductance induced by a reduction in bath osmolality from 350 to 290 mOsm/kg H20 (osmolality reduced by removing sucrose from the isotonic solution). In colchicinetreated cells the hypotonic bath solution increased the C1-conductance from 12.9 + 1.6 to 19.7 ? 2.4 nS (n = 4, p < 0.05). As reported above, a similar decrease in osmolality in the absence of colchicine increased C1-conductance from 15.0 ? 1.3 nS to a peak value of 25.9.2 1.5 nS ( p < 0.001; n = 14). These experi- ments are consistent with the view that swelling-induced activation of C1-currents occurs by a mechanism that is independent of microtubules, exocytosis, and the delivery of C1channels, and/or a regulatory protein, from an intracellular vesicular pool to the plasma membrane (see "Discussion").

DISCUSSION
The major observation of this study is that RCCT-28A cells volume regulate in response to hypotonic cell swelling, at least in part, by activating 305 pS C1-channels. Our data suggest that cell swelling activates the 305 pS C1-channel by a signaling pathway involving the disruption of F-actin.
RCCT-28A cells also express outwardly rectifying C1-channels (48). Accordingly, we cannot rule out the possibility that the outwardly rectifying C1-channel may, in addition to the 305 pS C1-channel, also contribute to RVD. However, it is unlikely that the outwardly rectifying C1-channel plays a major role in cell volume regulation, because we found that cell swelling increases a whole cell GIconductance exhibiting a linear I-V relationship, not an outwardly rectifying I-V relationship.
Cl-Channels a n d Cell Volume Regulation-C1-conductive pathways contribute to RVD in a variety of cell types (7, 9-12, 14, 15,49,50). However, only a few studies have characterized C1channels involved in RVD (51-54). In neuroblastoma cells, hypotonic cell swelling activates a 200-400 pS C1-channel that is permeable to HC03-(PCI-:PHCO3 = 2.4:l) and rapidly inactivates when the voltage is changed to values more negative than -30 mV (51). The properties of this channel are similar to the 305 pS C1-channel in . Cell swelling activates a 3 pS C1-channel in lymphocytes, a 23 pS C1-channel in Ehrlich ascites tumor cells, and an outwardly rectifying Does Membrane Stretch and Cell Swelling Activate the 305 pS Cl-Channel by Depolymerizing F-actin?-Our studies suggest that membrane stretch activates the 305 pS C1-channel by a mechanism involving disruption of F-actin. Several lines of evidence support a role for F-actin in regulating ion channels in other cell types. Fragmentation of F-actin activates Na+-selective channels in cell-attached and inside-out patches of the apical membrane of A6 cells (34). Phalloidin blocks CAMP-activated C1-secretion and CAMP-induced rearrangement of Factin organization in T84 cells (44). DHCB and cell swelling activate a C1-conductance in cardiocytes (55). Because DHCB and swelling were not additive it was suggested that F-actin was involved in the swelling activation mechanism (55). Hypotonic cell swelling activates RVD in promyelocytic leukemic cells while simultaneously depolymerizing F-actin (17). Cytochalasins and CAMP depolymerize F-actin in Madin-Darby canine kidney cells and decrease cell volume by activating K+ and C1-conductances ( 18). Finally, RVD in melanoma cells devoid of actin-binding protein is impaired; however, genetic rescue with actin-binding protein resulted in recovery of cell volume following exposure to a hypotonic solution, a process dependent on activation of a K' channel (19). Do Cell Swelling and Negative Pressure Activate the 305 p S Cl-Channel by the Same Mechanism?-The most parsimonious interpretation of our single channel and whole cell patch clamp data is that cell swelling activates C1-channels by a mechanism involving membrane stretch and disruption of the actin-based cytoskeleton. Making a few reasonable assumptions, it can be calculated that -20 cm of HzO suction applied to an inside-out membrane patch produces a transmembrane pressure that is equivalent to the pressure generated by hypotonic cell swelling. According to Laplace's law, -20 cm of Hz0 suction applied to a 1.5-pm diameter membrane patch will produce a tension of 2 dynedcm2 (56-58). Using equation A4 from Sackin (57) and the lowest estimate of the plasma membrane area elasticity constant (130 dyneskm') (581, a n increase in cell volume of only 1% should be sufficient to produce a membrane tension of 0.8 dynedcmz, a value comparable with a tension of 2 dynes/cm2 created by suction (57). This calculation implies that cell swelling creates a pressure gradient across the cell membrane (direct measurements reveal a n increase in intracellular pressure in Xenopus oocytes exposed to a hypotonic solution (59)). Although it is not yet possible to measure intracellular pressure in smaller cells normally used as models to study cell volume regulation, these calculations indicate that cell swelling is likely to increase membrane tension to levels equivalent to or even greater than those produced by application of negative pressure to membranes by patch clamp electrodes.
Signaling Mechanism of Swelling-induced Activation of the 305 p S CZ-Channel-Although volume-sensitive ion channels have been described in a variety of cell types, little information is available concerning the mechanisms by which cell swelling activates ion channels. Some evidence suggests that the actinbased cytoskeleton is involved. In particular, disruption of the filamentous actin network below the apical plasma membrane may increase channel currents by: 1) removing a physical barrier preventing the exocytic insertion of vesicles containing channels and/or a channel regulator from an intracellular pool into the plasma membrane; 2) liberating second messengers ; 3) changing the conformation of the channel protein due to a n alteration in the channel-actin-based cytoskeleton interaction. For example, in some cells, RVD involves the insertion of vesicles into the plasma membrane, from a cytoplasmic pool, by a mechanism that is sensitive to cytochalasins and microtubule-disrupting agents (60-62). However, in the present study the hypotonic bath solution had no measurable effect on whole cell capacitance (i.e. membrane area). Furthermore, colchicine, a drug that inhibits microtubule-dependent exocytosis, did not prevent swelling activation of the whole cell C1-conductance. Accordingly, our data are most consistent with the conclusion that exocytosis, and in particular microtubule-dependent exocytosis, is not an important component of RVD in RCCT-28A cells. On the other hand, it is possible that cell swelling stimulated exocytosis and increased membrane area but that the change was too small to detect. Furthermore, it must also be considered that swelling stimulated parallel increases in exocytosis and endocytosis such that cell capacitance remained unchanged. Additional experiments are required to explore these alternative hypotheses.
In preliminary experiments we observed that diacylglycerol production by RCCT-28A cells is stimulated by the hypotonic bath solution and that calphostin C, an inhibitor of protein kinase C, and pertussis toxin block the swelling-induced activation of the 305 pS C1-channel (4). Furthermore, calphostin C and pertussis toxin prevented disruption of actin filaments in cells exposed to the hypotonic solution.2 These observations are consistent with previous studies on RCCT-28Acells in which we demonstrated that the pertussis toxin-sensitive G protein, Gai.3, and protein kinase C activate the 305 pS C1-channel (20). Accordingly, our preliminary data suggest that cell swelling activates C1-channels by a sequential signaling pathway that includes cell swelling, Gai-3, protein kinase C, and disruption of F-actin.
It also must be considered that other signaling pathways may contribute to swelling-induced activation of the 305 pS C1channel. Cell swelling increases intracellular calcium in some cells, and it i s known that a rise in intracellular calcium activates C1-channels either directly or indirectly via calmodulin or calmodulin-dependent protein kinases (16). Furthermore, exposure to a hypotonic solution increases the production of arachidonic acid and its metabolites and increases cyclic AMP.
Thus, other signaling mechanisms, not yet examined, may also be involved in RVD in RCCT-28A cells.
Finally, changes in the conformation of the C1-channel protein due to alterations in channel-actin cytoskeleton interactions may also regulate channel activity. Ion channels and transporters are linked to actin via spectrin and ankyrin in many cell types (33,(63)(64)(65)(66)(67)(68)(69)(70)(71)(72). This interaction is known to anchor these integral membrane proteins in specific domains within a cell; however, it is not known whether cytoskeletal proteins directly affect transporter function. As integral membrane proteins, ion channels have numerous membrane-spanning domains that are stabilized by electrostatic attractions and subtle conformations. Alteration of these conformations by changes in the actin-based cytoskeleton may change the physical dimensions of the ion channel protein, alter its conformation, and open or close the channel pore. Adenylyl cyclase, a membrane-spanning protein with an ion channel motif and function, is activated by mechanical deformation of the cells, resulting in an increase in intracellular concentrations of cyclic AMP (16). This activation is enhanced by cytochalasins and by microtubule-disrupting agents such as colchicine and vinblastine (73,74). However, the mechanism whereby actin regulates transmembrane proteins is unknown and is a challenge for future research.
gift of the RCCT28A cells, Sabine Dietl for performing cell volume