Activation of the osmo-sensitive chloride conductance involves P21rho ...

Molecular Biology of the Cell Vol. 7, 1419-1427, September 1996

Activation of the Osmo-sensitive Chloride Conductance Involves P21rho and Is Accompanied by a Transient Reorganization of the F-Actin Cytoskeleton Ben C. Tilly,*" Marcel J. Edixhoven,* Leon G.J. Tertoolen,* Narito Morii,§ Yuji Saitoh,§ Shuh Narumiya,§ and Hugo R. de Jonge* *Department of Biochemistry, Cardiovascular Research Institute COEUR, Medical Faculty, Erasmus University, 3000 DR Rotterdam, The Netherlands; tHubrecht Laboratory, Netherlands Institute for Developmental Biology, 3584 CT Utrecht, The Netherlands; and §Department of Pharmacology, Faculty of Medicine, Kyoto University, Sakyo-ku, Kyoto 606, Japan Submitted February 23, 1996; Accepted July 1, 1996 Monitoring Editor: Roger Y. Tsien

Hypo-osmotic stimulation of human Intestine 407 cells rapidly activated compensatory CF- and K+ conductances that limited excessive cell swelling and, finally, restored the original cell volume. Osmotic cell swelling was accompanied by a rapid and transient reorganization of the F-actin cytoskeleton, affecting both stress fibers as well as apical ruffles. In addition, an increase in total cellular F-actin was observed. Pretreatment of the cells with recombinant Clostridium botulinum C3 exoenzyme, but not with mutant enzyme (C3-E173Q) devoid of ADP-ribosyltransferase activity, greatly reduced the activation of the osmo-sensitive anion efflux, suggesting a role for the ras-related GTPase p21rho. In contrast, introducing dominant negative N17-p21rac into the cells did not affect the volume-sensitive efflux. Cell swelling-induced reorganization of F-actin coincided with a transient, C3 exoenzyme-sensitive tyrosine phosphorylation of p125 focal adhesion kinase (pl25FAK) as well as with an increase in phosphatidylinositol-3-kinase (Ptdlns-3kinase) activity. Pretreatment of the cells with wortmannin, a specific inhibitor of PtdIns3-kinase, largely inhibited the volume-sensitive ion efflux. Taken together, our results indicate the involvement of a p2lrho signaling cascade and actin filaments in the activation of volume-sensitive chloride channels. INTRODUCTION Most mammalian cells have developed compensatory mechanisms to respond to the variable osmotic stress caused by changes in the concentrations of osmoactive substances (e.g., glucose, amino acids, lactate) or by variations in the osmolarity of the surrounding medium. Two distinct mechanisms can be recognized: the Regulatory Volume Increase (RVI), leading to a net increase of the osmolarity of the cell, and the Regulatory Volume Decrease (RVD), directing a reduction in cellular tonicity, which are activated by cell shrinkage and cell swelling, respectively. Although the RVI detCorresponding author: Department of Biochemistry, Medical Faculty, Erasmus University, P.O. Box 1738, 3000 DR Rotterdam, The Netherlands.

© 1996 by The American Society for Cell Biology

pends on the activation of ion pumps and carriers, the RVD response involves the concerted opening of K+ and Cl- selective ionic channels, leading to a net efflux of KCI and driving the loss of cellular water (for review, Okada and Hazama, 1989; Grinstein and Foskett, 1990; Al-Habori, 1994). Although the conductive efflux pathways triggered in response to cell swelling have been described in many different cell systems, the molecular mechanism or mechanisms involved in the activation of osmo-sensitive ion channels and transporters have not been clarified in detail and may differ between cell types. Several mechanisms have been suggested to play a role in the RVD response, including a cell swellinginduced rise in intracellular Ca2' activating both Cland K+ channels (McCarty and O'Neil, 1992). Several 1419

B.C. Tilly et al.

cell types, however, including the Intestine 407 cell line used in this study, do not express Ca2+-sensitive Cl- channels, and the osmo-sensitive anion efflux is likely to be elicited independently of a rise in [Ca2+]i (Hazama and Okada, 1988; Grinstein and Smith, 1990; Tilly et al., 1994). As an alternative, a role for the cytoskeleton in regulating swelling-activated ion channels has been proposed. This hypothesis is based primarily on the observation that the volume-regulatory response in a number of cell models was found to be sensitive to pharmacological disruption of the cytoskeleton (Foskett and Spring, 1985; Cornet et al., 1993; Haussler et al., 1994; Hug et al., 1995) and was accompanied by small changes in total cellular F-actin (Hallows et al., 1991; Ziyadeh et al., 1992). Moreover, direct ion channel activation by cytoskeletal elements has been demonstrated in excised membrane patches of toad kidney A6 cells, showing a rapid activation of apical Na+ channels with the addition of exogenous actin (Cantiello et al., 1991). Stimulating responsive cells with mitogens like bombesin, endothelin, thrombin, and lysophosphatidic acid or inducing integrin-receptor dimerization leads to a rapid reorganization of cellular F-actin by using signaling pathways involving the p2lrho family of GTPases (Ridley and Hall, 1992; Nobes and Hall, 1995). In addition, these diverse extracellular stimuli also activate tyrosine phosphorylation of proteins associated with the cytoskeleton, most notably the focal adhesion kinase pl25FAK, a tyrosine kinase associated with focal adhesions (Schaller et al., 1992; Zachary and Rozengurt, 1992). Using the human Intestine 407 cell line as a model, we have previously demonstrated that activation of tyrosine kinase(s) is a prerequisite for activation of the osmo-sensitive CF- conductance (Tilly et al., 1993). Indeed, after hypo-osmotic stimulation, a rapid and transient increase in protein tyrosine phosphorylation was observed, including phosphorylation of the p42 and p44 MAP kinases (Tilly et al., 1993). Hypo-osmotic stimulation of Intestine 407 cells also led to the rapid tyrosine phosphorylation of a 125-130 kDa protein, raising our interest in a possible volume-regulatory role of signaling pathways that regulate the F-actin cytoskeleton in the activation of the RVD response. In the present study, we demonstrate that hypo-osmotic cell swelling is accompanied by a fast but transient reorganization of the F-actin cytoskeleton as well as by an increased phosphorylation of pl25FAK. Both the pl25FAK phosphorylation and the hypotonicity-induced anion efflux were inhibited in Clostridium botulinum C3 exoenzyme-pretreated cells, suggesting the involvement of p2lrho. In addition, it was found that phosphatidylinositol-3-kinase (Ptdlns-3-kinase), a potential substrate of pl25FAK that may play a role in cytoskeletal organization (Chen and Guan, 1994; Wynann and Arcado, 1994), is also activated after osmotic cell swelling. Our results suggest a 1420

role for a signaling cascade involving p2lrho in the activation of volume-regulated Cl- channels. MATERIALS AND METHODS Materials Radioisotopes (125I- and y-32P-ATP) and electrochemiluminescence (ECL) Western blotting detection kits were obtained from Amersham Netherlands B.V. ('s Hertogenbosch, Netherlands). Hormones, rhodamine-conjugated phalloidin, phosphatidylinositol, and wortmannin were purchased from Sigma Chemical (St. Louis, MO). Other chemicals were obtained from the following sources: monoclonal anti-phosphotyrosine antibodies from Santa Cruz Biotechnology (Santa Cruz, CA) and KT5926 from Calbiochem (La Jolla, CA). Anti-p125FAK monoclonal 2A7 and polyclonal BC3 antibodies, raised against isolated p125 and a bacterially expressed trpE fusion protein containing the C terminal of the kinase domain, respectively (Kanner et al., 1990; Schaller et al., 1992), were kindly donated by Dr. J.T. Parsons, University of Virginia. Recombinant C. botulinum C3 exoenzyme and mutant C3-E173Q were isolated and prepared as described previously (Kumagai et al., 1993; Saito et al., 1995).

Cell Culture Intestine 407 cells were grown as monolayers on plastic in DMEM supplemented with 25 mM HEPES, 10% fetal calf serum, 1% nonessential amino acids, 40 mg/l penicillin, and 90 mg/ml streptomycin under a 95% 02/5% CO2 atmosphere.

F-Actin Quantification After hypotonic (30%) stimulation, cells were fixed for 15 min in phosphate-buffered saline (PBS) containing 2% formaldehyde and subsequently permeabilized in PBS/Triton X-100 (0.1% vol/vol). After blocking aspecific binding by using PBS plus 0.5% bovine serum albumin, we stained cells with rhodamine-conjugated phalloidin (250 ng/ml). Groups of 8-10 individual cells were identified, and extended focus images were constructed (summated optical sections parallel to the substratum) with a X63 oil immersion objective (Bio-Rad Lasersharp MRC-600 confocal laser microscope, Richmond, CA). Fluorescence was quantitated as the average pixel intensity per cell. During the experiments, background and amplification settings were kept constant. In a typical experiment, intensities of 40-60 cells per time point were quantified.

pl25FAK Immunoprecipitation Cultures were stimulated with histamine or a hypotonic medium as indicated in the legends and washed twice with ice-cold PBS containing 100 ,uM sodium orthovanadate. Thereafter, cells were lysed in RIPA (PBS containing 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM phenylmethylsulfonyl fluoride [PMSF], 1% aprotinin, and 100 mM sodium orthovanadate) for 10 min at 0'C. Lysates were collected and cleared by centrifugation, and the supernatants were incubated with 2A7 monoclonal anti-p125FAK antibodies for 1 h before addition of protein G-Sepharose. One hour later, the immunoprecipitates were washed four times with RIPA and finally resuspended in SDS sample buffer. Precipitates were subjected to SDS-PAGE and electrophoretically transferred to nitrocellulose. Tyrosine phosphorylation of pl25AK was determined with monoclonal anti-phospho-tyrosine antibodies and an ECL Western blotting detection system according to the instructions provided by the manufacturers. Molecular Biology of the Cell

Osmo Regulation of Chloride Channels

Immunoblotting Monolayers of cells were stimulated with hypotonic media, and incubations were terminated by the addition of boiling SDS sample buffer. Proteins were separated by SDS-PAGE, transferred to nitrocellulose, and analyzed as described above.

125I- Efflux Assays Intestinal 407 cells were loaded with 5 ,uCi 1251- for 2 h and washed three times with isotonic buffer containing (in mM) 80 NaCl, 5 KCl, 1.3 CaCl2, 1 MgCl2, 10 glucose, 95 mannitol, and 20 HEPES, pH 7.4, before the assay. Isotope efflux was determined at 37'C by replacing the medium at 1-2 min intervals. Hypotonic buffers were prepared by adjusting the concentration of mannitol. Radioactivity in the media was determined by gamma radiation counting and expressed as fractional efflux per minute, as previously described (Vaandrager et al., 1991).

Vaccinia Virus Infection Recombinant viral growth factor-minus strains of wild-type and dominant negative p21rac-containing vaccinia virus (Peppelenbosch et al., 1995) were used to infect cultures of Intestine 407 cells (10-20 plaque forming units per cell for 60 min in serum-free DMEM). After infection, the medium was replaced by DMEM containing 7.5% fetal calf serum; isotope efflux experiments were performed after 16-18 h.

Phosphatidylinositol-3-kinase Assay PtdIns-3-kinase activity was determined in anti-P-tyr immunoprecipitates as previously described (Soltoff et al., 1994). Briefly, cultures of Intestine 407 cells were stimulated with hypotonic media (30% hypotonicity) and lysed in a buffer containing (in mM) 137 NaCl, 20 Tris-HCl, pH 7.5, 1 MgCl2, 1 CaCl2, 1 PMSF, and 200 ,uM sodium orthovanadate, 10% glycerol, and 0.1% Nonidet-P40. After 10 min, lysates were collected and cleared by centrifugation, and agarose-conjugated anti-P-tyr antibodies were added to the supernatants. After 2 h of incubation at 0'C, the agarose beads were centrifuged, washed three times with 1 ml of PBS plus 1% NonidetP40, twice with 0.5 M LiCl/0.1 M Tris-HCl, and twice with 100 mM NaCl/1 mM EDTA/10 mM Tris-HCl, pH 7.5. Subsequently, the beads were resuspended in a phosphorylation buffer containing (in mM) 10 MgCl2, 1 EGTA, and 20 HEPES, pH 7.5, and 0.4 mg/ml phosphatidylinositol. Phosphorylation was started by addition of y-32P-ATP (15 ,uCi/ml, final concentration 20 ,uM) and terminated after 10 min by addition of 1 M HCl. Lipids were extracted and separated by thin-layer chromatography. Lipid spots were visualized by autoradiography and by exposition to 12 vapor, scraped, and quantitated for 32P-radioactivity by liquid scintillation counting.

stimulated controls, a marked increase in phosphotyrosine immunostaining was observed in p125 precipitates derived from osmotically stimulated cells (30% hypotonicity), confirming that hypotonic conditions enhance the tyrosine phosphorylation of pl25FAK (Figure 1). Phosphorylation of p125F was not restricted to osmotic stimulation, because histamine, a hormone previously found to potentiate the volumeactivated anion efflux (Tilly et al., 1994), was also able to induce pl25FAK phosphorylation (Figure 1). Rapid Changes in the F-Actin Cytoskeleton after Osmotic Cell Swelling Phosphorylation and activation of p125FAK have been implicated in the organization of actin stress fibers and the (re-) assembling of focal adhesions (Burridge et al., 1992; Barry and Critchley, 1994; ChrzanowskaWodnicka and Burridge, 1994). Using rhodamineconjugated phalloidin as a fluorescent probe, we studied changes in the amount and organization of the F-actin cytoskeleton in cell cultures challenged with a 30% hypotonic shock. Immediately after hypo-osmotic stimulation, an -40% increase in rhodamine fluorescence was observed, indicating an increase in total cellular F-actin (Figure 2). The increase in F-actin was transient, returning slightly below basal levels within 5 min after stimulation. Thereafter, no net changes in the total cellular F-actin content were observed. To kDa

200

anti-P-Tyr

-

97 ..

97

anti-p25fAK

69

"uNt

.

bu

-

"

i

RESULTS 46-

Hypotonicity-induced Tyrosine Phosphorylation of pl25FAK

Previously, we have demonstrated that confluent monolayers of Intestine 407 cells respond to extracellular hypotonicity with a transient increase in tyrosine phosphorylation of several proteins, including a 120130 kDa protein comigrating with murine fibroblastderived pl25FAK (Tilly et al., 1993, 1994). To asses putative changes in the phosphorylation state of

pl25FAK more directly, p125FAK was immunoprecipi-

tated from osmotically activated cells and analyzed for its phospho-tyrosine content. As compared with un-

Vol. 7, September 1996

C

Figure

p125FAK

Hypo

Hims

Hypotonicity-induced was immunoprecipitated

1.

phosphorylation

of

p125FK

from lysates of control (C), osmotically stimulated cultures (Hypo; 30% hypotonicity for 2 min), and histamine-stimulated cultures (Hist; 100 ,uM for 5 min) by using 2A7 anti-p125FAK antibodies. Immunoprecipitates were separated by SDS-PAGE followed by Western blotting. Immunoblots were incubated with either anti-P-tyr (lanes indicated as anti-P-Tyr) or BC3 anti-p125FAK antibodies (lane indicated as anti-p125FA ). Arrow indicates the position of p125FAK. 1421

B.C. Tilly et al. 200

z0

150 h

8U) C 0

100 [

8(0 50

F

L shock

0

0

10

5

15

Time (min) Figure 2. Hypotonicity-induced increase in F-actin cytoskeleton. Intestine 407 cells were stimulated with a 30% hypotonic buffer for the indicated times, fixed, permeabilized, and stained with rhodamine-conjugated phalloidin. F-actin content was expressed as a percentage of rhodamine-conjugated phalloidin fluorescence per cell relative to the unstimulated control cells. Data are expressed as mean + SD for 130-170 individual cells. Asterisk indicates a significant difference from the control (p < 0.05, Student t test).

investigate the reorganization of the cytoskeleton in more detail, we studied the intracellular distribution of the F-actin cytoskeleton by constructing optical sections from the apical and basal regions of the cells with a confocal laser microscope. As shown in Figure 3, osmotic cell swelling resulted in an immediate and rather dramatic increase in basal stress-fiber formation, accompanied by a loss in apical F-actin likely to be associated with plasma membrane ruffles. Numerous small and poorly organized fibers were observed within 1 min of osmotic stimulation, whereas a clear increase in actin stress fibers was found after 2 min (Figure 3). Subsequently, stress-fiber formation declined, and the amount of F-actin fibers returned to prestimulatory levels after 5-15 min. In addition to an increase in basal stress-fiber formation, a marked decrease in apical F-actin was observed (Figure 3). Like stress-fiber formation, the decrease in apical F-actin was transient rather than sustained, and, as compared with unstimulated controls, no clear differences were observed after 5 min of stimulation. Taken together, osmotic stimulation of Intestine 407 cells led to an immediate reorganization of the F-actin cytoskeleton, affecting both the basal and apical poles of the cells.

Inhibition of the Volume-sensitive Anion Efflux and

pl25FAK Phosphorylation by Pretreatment with

Clostridium botulinum C3 Exoenzyme It is now well established that members of the Rho family of small GTP-binding proteins, including 1422

p21rho and p2lrac, play an important role in the organization of the F-actin cytoskeleton, stress-fiber formation, and cell ruffling (Ridley and Hall, 1992, 1994; Barry and Critchley, 1994). To study the involvement of Rho-type GTPases in the cellular response to hypotonicity, we pretreated Intestine 407 cells with C. botulinum C3 exoenzyme, a toxin that enters the cell through nonspecific endocytosis (Kamata et al., 1994) and specifically promotes ADP-ribosylation and, thereby, inactivation of p2lrho (Ridley and Hall, 1992; Aktories and Just, 1995). As shown in Figure 4A, pretreatment of the cells with C3 exoenzyme largely inhibited the hypotonicity-induced activation of 1251- efflux, most plausibly reflecting an inhibition of volume-sensitive Cl- channels. Importantly, no inhibition of the volume-sensitive I2S- efflux was observed after pretreatment of the cells with C3-E173Q mutant exoenzyme, which is devoid of ADP-ribosyltransferase activity (Figure 4A). Introducing a dominant negative species of the Rho-related p2lrac protein by using recombinant vaccinia virus expressing N17p2lrac did not affect the cell swelling-activated anion efflux (peak 1251- efflux after 30% hypotonic stimulation; mock-infected: 33 ± 2%/min; WT-Rac-infected: 29 ± 1%/min; N17-Rac-infected: 30 ± 1%/min). Inhibition of the volume-sensitive anion efflux by C3 exoenzyme was accompanied by a decrease in hypotonicity-induced pl25FAK phosphorylation while leaving the cell swelling-activated phosphorylation of most other proteins unaffected (Figure 4B). Activation of Phosphatidylinositol-3-Kinase by Osmotic Cell Swelling Phosphatidylinositol-3-kinase (PtdIns-3-kinase) recently has been recognized as a potential substrate for p125FAK and as a mediator of F-actin reorganization (Chen and Guan, 1994; Shibasaki et al., 1994; Wymann and Arcado, 1994). In addition, PtdIns-3kinase activation previously has been found to be sensitive to C3 exoenzyme pretreatment (Kumagai et al., 1993). To explore a possible connection with the p21rho/p125FAK signaling pathway in more detail, PtdIns-3-kinase activity was determined in cell cultures exposed to an osmotic shock (30% hypotonicity). Figure 5 shows that activation of PtdIns-3-kinase, as determined by the in vitro phosphorylation of phosphatidylinositol, was already observed after 1 min of osmotic stimulation and reached a peak level after 2 min (approximately fourfold increase in activity). Thereafter, the kinase activity slowly declined but did not return to basal levels within the time period of the measurements (10 min). As has been shown for cell swelling-induced isotope efflux and pl25FAK phosphorylation, the hypotonicity-provoked Ptdlns-3-kinase activity was also blocked by C3 exoenzyme pretreatment (Figure 5C). Importantly, as is evident from Molecular Biology of the Cell


Figure 3. Hypotonicity-induced changes in cellular distribution of F-actin. Cultures were stimulated with a hypotonic buffer for the indicated time and stained with rhodamine-conjugated phalloidin, as described in MATERIALS AND METHODS. Thereafter, optical sections were prepared from the basal (stress fibers) as well as apical (ruffles) parts of the cells. Stress fibers, in control cells some stress fibers can be observed. After 1 min of osmotic stimulation (30% hypotonicity), a dramatic increase in polymerized actin can be seen, followed by an intense stress-fiber formation (2 min). Thereafter, stress-fiber breakdown becomes apparent (5 min), a process that is completed at t = 15 min. Ruffles, resting cells do have extensive ruffles. Ruffling is strongly diminished at the time of maximal cell swelling (1 min) but is gradually restored and back at control levels within 5 min. Data presented are representative for 15 sections obtained from three independent experiments. Vol. 7, September 1996

1423

B.C. Tiliv ct til.

was manifest at 1 ,M. Notably, in intact cells, PtdIns3-kinase activation was found to be equally sensitive to wortmannin pretreatment (5-10% and 60-70% reduction in PtdIns-3-P formation after pretreatment with 100 nM and 1 ,uM, respectively). Because the micromolar concentrations of wortmannin (1 ,tM) might also affect the activity of myosin light chain kinase, similar experiments were performed in the presence of KT5926, a specific inhibitor of myosin light chain kinase (Nakanishi et al., 1990). However, at concentrations up to 10 ,uM, KT5926 seemed unable to

A 40 - 30

c 1-

xIs

20

U) 10

6

9

12

A

Time (min)

B kDa C 200

-

97

-

1

2

5

10

Time (min)

B

500

a C

c

0 69

400

0

o 300 46

-

m 0

a.L Bot. C3

-

-

+

CO

Shock

-

+

+

-o

a.

200

100

0

Figure 4. Inhibition of hypotonicity-induced 125I efflux and p125FAK phosphorylation by C. botulinum C3 exoenzyme. Intestine 407 cells were pretreated with 50 ,zg/ml C3 or C3-E173Q exoenzyme for 48 h before the experiments. Thereafter, hypotonicityinduced (30% hypotonic shock) activation of (A) 1251- efflux and (B) protein tyrosine phosphorylation (10 ,ug of protein per lane) was studied, as described in MATERIALS AND METHODS. Data in A are expressed as mean ± SD for n 2 3; asterisk indicates a significant difference from the control (p < 0.05). Arrow indicates time point of changing from an isotonic to a hypotonic medium.

the much larger activation attainable on exposure of Intestine 407 cells to EGF (-100-fold increase in activity), only a minor subfraction of the enzyme is responsive to hypotonic conditions. To investigate whether PtdIns-3-kinase is involved in the osmo regulation of Cl- channels, we pretreated the cultures with wortmannin, a specific inhibitor of Ptdlns-3-kinase (Yano et al., 1993). Figure 6 shows that partial inhibition of the cell swelling-activated 1251- efflux was found after pretreatment of the cultures with 100 nM wortmannin, whereas almost complete inhibition of the responses 1424

2

10

Time (min)

C

Bot. C3 Shock

-

+

+ +

Figure 5. Activation of PtdIns-3-kinase by osmotic cell swelling. PtdIns-3-kinase activity was determined in vitro after stimulating the cells for the indicated times with a 30% hypotonic buffer, as described in MATERIALS AND METHODS. (A) Arrowhead denotes the position of standard PtdInsP; 0 = origin. (B) 32P-radioactivity of the Ptdlns-3-P spot expressed as percentage of stimulation relative to the control. At the 2-min time point, a 405 ± 93% (n = 4) increase in 32P-Ptdlns-3-P was observed. (C) Inhibition of the osmotic swelling-induced Ptdlns-3-kinase activity after pretreatment with C3 exoenzyme (50 ,ug/ml, 48 h). No inhibition was observed when the cells were treated with C3-E173Q. Molecular Biology of the Cell


0 C

100

T

--,r-

-F

0 -4

75

x

50 G)

25 -

0

0

-1

100nM 1 PM

Wortmannin

0

10pM

KT5926

Figure 6. Inhibition of the osmo-sensitive 125I efflux by wortmannin. Intestine 407 cells were pretreated with the indicated concentrations of wortmannin or KT5926 for 30 min before hypo-osmotic stimulation (30% hypotonicity). Data are expressed as a percentage of the untreated control (mean ± SD for n = 3). The absolute 125efflux from untreated cultures was 35 ± 2%/min and 32 ± 3%/min for the wortmannin and the KT5926 experiment, respectively.

mimic the inhibitory effect of wortmannin on the volume-sensitive 1251- efflux (Figure 6). DISCUSSION Osmotic swelling of Intestine 407 cells activates conductive K+ and Cl- efflux pathways that limit excessive cell swelling and, finally, restore cellular volume (Okada and Hazama, 1989; Tilly et al., 1993, 1994). Previous experiments have demonstrated that osmo regulation of the CF- channels could be achieved independently of changes in [Ca2+]i (Hazama and Okada, 1988; Tilly et al., 1994) but, instead, needed the activation of one or more tyrosine kinase(s) (Tilly et al., 1993). Here we report that hypotonic challenge leads to an immediate reorganization of the actin cytoskeleton and to the activation of cytoskeleton-associated signal transduction pathways. Involvement of the cytoskeleton in cell volume regulation has been proposed previously. Foskett and Spring (1985) reported that the RVD response in gallbladder epithelia was inhibited in the presence of cytochalasin B and suggested that an intact F-actin filament system is a prerequisite for RVD. Since then, similar results have been obtained in studies with cultured cells, including Ehrlich ascites tumor cells, HSG cells, and bronchial epithelial cells (Cornet et al., 1993; Fatherazi et al., 1994; Hug et al., 1995). In other cell models, however, disruption of the F-actin cytoskeleton did not affect the RVD response (Hallows et al., 1991) or was even found to potentiate the activity of volume-sensitive Cl- channels (Haussler et al., 1994). Hypo-osmotic swelling of Intestine 407 cells resulted both in an increase in total cellular F-actin Vol. 7, September 1996

content as well as in a rapid but transient reorganization of the actin cytoskeleton. It is now well established that alterations of the actin cytoskeleton may involve activation of GTP-binding proteins of the p2lrho family, which includes Rho, Rac, and Cdc42 (Nobes and Hall, 1995). Furthermore, evidence is accumulating that inactivation of these GTPases prevents hormone-induced changes in cell morphology (Ridley and Hall, 1992; Jalink et al., 1994). Because pretreatment of Intestine 407 cells with C. botulinum C3 exoenzyme almost completely blocks the cell swelling-induced 1251- efflux, p2lrho activation may play an essential role in the signaling pathway leading to the opening of volume-sensitive Cl- channels, most plausibly through the induction of cytoskeletal alterations. At least two possible mechanisms may underlie the regulation of ionic channels by cytoskeletal rearrangements. First, generation of actin filaments of a specific size may be crucial for activation. This notion is supported by our observations that pretreatment of the cells with cytochalasin B, but not colchicine or nocadozole, potentiated the hypotonicity-induced 125I- efflux by approximately twofold (Tilly and de Jonge, unpublished results). Modulation of the activity of ionic channels by actin filaments is not unprecedented: in toad kidney A6 cells an Na+ channel has been described that could be regulated directly by exogenous F-actin (Cantiello et al., 1991). In melanoma cells lacking actin-binding protein (ABP), however, activation of a volume-sensitive K+ channel was markedly decreased but could be rescued by transfecting the cells with the wild-type ABP gene (Cantiello et al., 1993), suggesting that a protein or proteins associated with the F-actin skeleton, rather than the actin filaments themselves, act as regulatory elements. As an alternative for a direct role of F-actin in ion channel regulation, modulation of signal transduction pathways associated with cytoskeletal rearrangements could also be involved. Indeed, hypotonicity-induced changes in the cytoskeleton are accompanied by an increase in tyrosine phosphorylation of several proteins, including the focal adhesion kinase pl25FAK. Phosphorylation and activation of p125FAK have been observed in response to a variety of extracellular stimuli (compare Zachary and Rozengurt, 1992; Schaller and Parsons, 1993). pl25FAK has been implicated in a number of cellular functions, including growth and differentiation, cell motility and adhesion, and regulation of cell shape (Zachary and Rozengurt, 1992), but its precise mechanism of action is still primarily unknown. Pretreatment of Intestine 407 cells with C. botulinum C3 exoenzyme almost completely and selectively abolished hypotonicityinduced tyrosine phosphorylation of p125FAK. A similar inhibition of p125FAK phosphorylation by C3 exoenzyme has been reported previously for hormone or lysophosphatidic acid-stimulated fibroblasts 1425

B.C. Tilly et al.

(Kumagai et al., 1993; Rankin et al., 1994). Because protein tyrosine phosphorylation is a prerequisite for osmo regulation of Cl- channels (Tilly et al., 1993), our combined data suggest that the p2lrho/pl25FAK signaling cascade is critically involved in cell volume regulation. Ptdlns-3-kinase, another signaling enzyme implicated in cytoskeletal reorganizations, chemotaxis, and membrane ruffling (Kotani et al., 1994; Shibasaki et al., 1994; Wennstrom et al., 1994), is likewise activated upon stimulation of cells with growth factors and mitogens. This lipid kinase recently has been suggested to serve as a substrate for pl25FAK (Chen and Guan, 1994), a notion supported by the observation that C3 exoenzyme-catalyzed ADP-ribosylation of p2lrho abolished Ptdlns-3-kinase activation (Kumagia et al., 1993; Figure 5), as well as by our finding in Intestine 407 cells that histamine, in addition to promoting pl25FAK phosphorylation (Figure 1), also increased PtdIns-3-kinase activity by approximately threefold (Tilly, unpublished results). In comparison, hypotonicity was found to elicit a transient stimulation of Ptdlns-3-kinase activity up to approximately fourfold. Moreover, pretreatment of the cells with low (100 nM) concentrations of Ptdlns-3-kinase inhibitor wortmannin partially inhibited hypotonicity-induced 1251- efflux, whereas higher concentrations (1 ,uM) almost completely abolished the osmo-sensitive anion efflux (Figure 6). At micromolar concentrations wortmannin is no longer specific for Ptdlns-3-kinase but, at least in vitro, also inhibits myosin light chain kinase (Yano et al., 1993). Two observations, however, suggest that PtdIns-3-kinase, and not myosin light chain kinase or other potential targets, is involved in the osmo regulation of Cl- channels. 1) In intact cells, the dose dependency of wortmannin inhibition was similar for hypotonicity-induced PtdIns-3-P formation and for the osmo-sensitive 125I- efflux (Figure 6). 2) KT5926, a specific inhibitor of myosin light chain kinase, did not affect the volume-sensitive 1251- efflux. Furthermore, 100 nM wortmannin completely inhibited Ptdlns-3-kinase activity when added to the lysates, suggesting a rather poor penetration into Intestine 407 cells. The results of the present study demonstrate that a signal transduction pathway known to be exploited by hormones for remodeling of the F-actin cytoskeleton resulting in changes in cell morphology, i.e., the p2lrho/p125FAK cascade and Ptdlns-3-kinase, is switched on also in response to hypotonic conditions, whereas inhibition of this pathway leads to an impaired activation of volume-sensitive Cl- efflux. Hormonal activation of pl25FAK and/or Ptdlns-3-kinase under isotonic conditions, as observed in the presence of histamine (pl25FAK) and/or EGF (Ptdlns-3-kinase), is not sufficient to activate volume-sensitive Cl- channels (compare Tilly et al., 1993, 1994), indicating that 1426

an additional factor, perhaps directly related to membrane stretch, is also essential. Linking the signal transduction pathways involved in Cl- channel activation to an independent, putative "volume sensor" has two important physiological consequences. First, activation of the channels occurs only during cell swelling, thereby preserving the specificity of the response; second, by coupling the volume response to pathways activated upon hormonal stimulation, small changes in cell volume that may occasionally take place during hormone-induced changes in cellular metabolism (Haussinger and Lang, 1991) are corrected more adequately.

ACKNOWLEDGMENTS The authors thank Dr. J.T. Parsons (University of Virginia) for his generous gift of anti-p125FAK antibodies, Dr. N. van der Berghe (University of Leiden) for donating wortmannin, and Drs. J.L. Bos and J.-P. Medema (University of Utrecht) for help with the vaccinia virus experiments. B.C.T. is a fellow of the Royal Dutch Academy of Sciences, The Netherlands.

REFERENCES Aktories, K., and Just, I. (1995). In vitro ADP-ribosylation of Rho by bacterial ADP-ribosyltransferases. Methods Enzymol. 256, 184-195. Al-Habori, M. (1994). Cell volume and ion transport regulation. Int. J. Biochem. 26, 319-334. Barry, S.T., and Critchley, D.R. (1994). The RhoA-dependent assembly of focal adhesions in Swiss 3T3 cells is associated with increased tyrosine phosphorylation and the recruitment of both ppl25FAK and protein kinase C-8 to focal adhesions. J. Cell Sci. 107,2033-2045. Burridge, K., Turner, C.E., and Romer, L.H. (1992). Tyrosine phosphorylation of paxillin and ppI25FAK accompanies cell adhesion to extracellular matrix: a role in cytoskeletal organization. J. Cell Biol. 119, 893-903. Cantiello, H.F., Prat, A.G., Bonventre, J.V., Cunningham, C.C., Hartwig, J.H., and Ausiello, D.A. (1993). Actin-binding protein contributes to cell volume-regulatory ion channel activation in melanoma cells. J. Biol. Chem. 268, 4596-4599. Cantiello, H.F., Stow, J.L., Prat, A.G., and Ausiello, D.A. (1991). Actin filaments regulate epithelial Na+ channel activity. Am. J. Physiol. 261, C882-C888. Chen, H.-C., and Guan, J.-L. (1994). Association of focal adhesion kinase with its potential substrate phosphatidylinositol-3-kinase. Proc. Natl. Acad. Sci. USA 91, 10148-10152. Chrzanowska-Wodnicka, M., and Burridge, K. (1994). Tyrosine phosphorylation is involved in reorganization of the actin cytoskeleton in response to serum or LPA stimulation. J. Cell Sci. 107, 3643-3654. Comet, M., Lambert, I.H., and Hoffmann, E.K. (1993). Relation between cytoskeleton, hypo-osmotic treatment, and volume regulation in Ehrlich ascites tumor cells. J. Membr. Biol. 131, 55-66. Fatherazi, S., Izutsu, K.T., Wellner, R.B., and Belton, C.M. (1994). Hypotonically activated chloride current in HSG cells. J. Membr. Biol. 142, 181-193. Foskett, J.K., and Spring, K.R. (1985). Involvement of calcium and cytoskeleton in gallbladder epithelial cell volume regulation. Am. J. Physiol. 248, C27-C36. Grinstein, S., and Foskett, J.K. (1990). Ionic mechanisms of cell volume regulation in leukocytes. Annu. Rev. Physiol. 52, 399-414. Molecular Biology of the Cell


Grinstein, S., and Smith, J.D. (1990). Calcium-independent cell volume regulation in human lymphocytes. J. Gen. Physiol. 95, 97-120. Hallows, K.R., Packman, C.H., and Knauf, P.A. (1991). Acute cell volume changes in anisotonic media affect F-actin content in HL-60 cells. Am. J. Physiol. 261, C1154-C1161. Haussinger, D., and Lang, F. (1991). Cell volume in the regulation of hepatic function: a mechanism for metabolic control. Biochim. Biophys. Acta 1071, 331-350. Haussler, U., Rivet-Bastide, M., Fahlke, C., Muller, D., Zachar, E., and Rudel, R. (1994). Role of the cytoskeleton in the regulation of Cl- channels in human embryonic skeletal muscle cells. Pflugers Arch. 428, 323-330. Hazama, A., and Okada, Y. (1988). Ca21 sensitivity of volumeregulatory K+ and Cl- channels in cultured human epithelial cells. J. Physiol. 402, 687-702. Hug, T., Koslowsky, T., Ecke, D., Greger, R., and Kunzelmann, K. (1995). Actin-dependent activation of ion conductances in bronchial epithelial cells. Pflugers Arch. 429, 682-690. Jalink, K., van Corven, E.J., Hengeveld, T., Morii, N., Narumiya, S., and Moolenaar, W.H. (1994). Inhibition of lysophosphatidate- and thrombin-induced neurite retraction and neuronal cell rounding by ADP-ribosylation of the small GTP-binding protein Rho. J. Cell Biol. 126, 801-810. Kamata, Y., Nishiki, T., Matsumura, K., Hiroi, T., and Kozaki, S. (1994). Morphological effects, rate of incorporation, and enzymatic action of ADP-ribosyltransferase, known as C3 exoenzyme, on human GOTO cells. Microbiol. Immunol. 38, 421-428. Kanner, S.B., Reynolds, A.B., Vines, R.R., and Parsons, J.T. (1990). Monoclonal antibodies to individual tyrosine-phosphorylated protein substrates of oncogene-encoded tyrosine kinases. Proc. Natl. Acad. Sci. USA 87, 3328-3332. Kotani, K., et al. (1994). Involvement of phosphoinositide 3-kinase in insulin- or IGF-1-induced membrane ruffling. EMBO J. 13, 23132321. Kumagai, N., Morii, N., Fujisawa, K., Nemoto, Y., and Narumiya, S. (1993). ADP-ribosylation of Rho p21 inhibits lysophosphatidic acidinduced protein tyrosine phosphorylation and phosphatidylinositol-3-kinase activation in cultured Swiss 3T3 cells. J. Biol. Chem. 268, 24535-24538. McCarty, N.A., and O'Neil, R.G. (1992). Calcium signalling in cell volume regulation. Physiol. Rev. 72, 1037-1061. Nakanishi, S., Yamada, K., Iwahashi, K., Kuroda, K., and Kase, H. (1990). KT5926, a potent and selective inhibitor of myosin light chain kinase. Mol. Pharmacol. 37, 482-488. Nobes, C.D., and Hall, A. (1995). Rho, Rac, and Cdc 42 GTPases regulate the assembly of multimolecular focal complexes associated with actin stress fibers, lamellipodia, and filopodia. Cell 81, 53-62. Okada, Y., and Hazama, A. (1989). Volume-regulatory ion channels in epithelial cells. News Physiol. Sci. 4, 238-242. Peppelenbosch, M.P., Qiu, R.-G., de Vries-Smits, A.M.M., Tertoolen, L.G.J., de Laat, S.W., McCormick, F., Hall, A., Symans, M.H., and Bos, J.L. (1995). Rac mediates growth factor-induced arachidonic acid release. Cell 81, 849-856. Rankin, S., Morii, N., Narumiya, S., and Rozengurt, E. (1994). Botulinum C3 exoenzyme blocks the tyrosine phosphorylation of

Vol. 7, September 1996

pI25FAK and paxillin induced by bombesin and endothelin. FEBS Lett. 354, 315-319. Ridley, A.J., and Hall, A. (1992). The small GTP-binding protein Rho regulates the assembly of focal adhesions and actin stress fibers in response to growth factors. Cell 70, 389-399. Ridley, A.J., and Hall, A. (1994). Signal transduction pathways regulating Rho-mediated stress-fiber formation: requirement for a tyrosine kinase. EMBO J. 13, 2600-2610. Saito, Y., Nemoto, Y., Ishizaki, T., Watanabe, N., Morii, N., and Narumiya, S. (1995). Identification of Glu173 as the critical amino acid residue for the ADP-ribosyltransferase activity of Clostridium botulinum C3 exoenzyme. FEBS Lett. 371, 105-109. Schaller, M.D., Borgman, C., Cobb, B.S., Vines, R.R., Reynolds, A.B., and Parsons, J.T. (1992). ppI25FAK, a structurally distinctive protein tyrosine kinase associated with focal adhesions. Proc. Natl. Acad. Sci. USA 89, 5192-5196. Schaller, M.D., and Parsons, J.T. (1993). Focal adhesion kinase: an integrin-linked protein tyrosine kinase. Trends Cell Biol. 3, 258-262. Shibasaki, T., Fukami, K., Fukui, Y., and Takenawa, T. (1994). Phosphatidylinositol 3-kinase binds to a-actinin through p85 subunit. Biochem. J. 302, 551-557. Soltoff, S.P., Carraway K.L., Prigent, S.A., Gullik, W.G., and Cantley, L.C. (1994). ErbB3 is involved in activation of phosphatidylinositol 3-kinase by epidermal growth factor. Mol. Cell. Biol. 14, 3550-3558. Tilly, B.C., Edixhoven, M.J., van den Berghe, N., Bot, A.G.M., and de Jonge, H.R. (1994). Ca2+-mobilizing hormones potentiate hypotonicity-induced activation of ionic conductances in intestine 407 cells. Am. J. Physiol. 267, C1271-C1278. Tilly, B.C., van den Berghe, N., Tertoolen, L.G.J., Edixhoven, M.J., and de Jonge, H.R. (1993). Protein tyrosine phosphorylation is involved in osmo regulation of ionic conductances. J. Biol. Chem. 268, 19919-19922. Vaandrager, A.B., Bapnath, R., Groot, J.A., Bot, A.G.M., and de Jonge, H.R. (1991). Ca + and cAMP activate different chloride efflux pathways in HT29-cl.19A colonic epithelial cell line. Am. J. Physiol. 261, G958-G965. Wennstrom, S., Siegbahn, A., Yokote, K., Arvidsson, A.-K., Heldin, C.-H., Morii, S., and Cleasson-Welsh, L. (1994). Membrane ruffling and chemotaxis transduced by the PDGF ,-receptor required the binding site for phosphatidylinositol 3' kinase. Oncogene 9, 651660. Wymann, M., and Arcado, A. (1994). Platelet-derived growth factorinduced phosphatidylinositol 3-kinase activation mediates actin rearrangements in fibroblasts. Biochem. J. 298, 517-520. Yano, H., Nakanishi, S., Kimura, K., Hanai, N., Saitoh, Y., Fukui, Y., Nonomura, Y., and Matsuda, Y. (1993). Inhibition of histamine secretion by wortmannin through the blockade of phosphatidylinositol 3-kinase in RBL-2H3 cells. J. Biol. Chem. 268, 25846-25856. Zachary, I., and Rozengurt, E. (1992). Focal adhesion kinase (p125FK): a point of convergence in the action of neuropeptides, integrins, and oncogenes. Cell 71, 891-894. Ziyadeh, F.N., Mills, J.W., and Kleinzeller, A. (1992). Hypotonicity and cell volume regulation in shark rectal gland: role of organic osmolytes and F-actin. Am. J. Physiol. 262, F468-F479.

1427