Extracellular pH alkalinization by Cl-/HCO3-exchanger is crucial for TASK2 activation by hypotonic...

Extracellular pH alkalinization by Cl�/HCO3� exchanger is crucial for TASK2

activation by hypotonic shock in proximal cell lines from mouse kidney

S. L’Hoste,1 H. Barriere,1 R. Belfodil,1 I. Rubera,1 C. Duranton,1

M. Tauc,1 C. Poujeol,1 J. Barhanin,2 and P. Poujeol11UMR Centre National de la Recherche Scientifique 6548, Universite de Nice-Sophia Antipolis, Nice, and2Institut de Pharmacologie du Centre National de la Recherche Scientifique, Valbonne Sophia-Antipolis, France

Submitted 18 April 2006; accepted in final form 20 September 2006

L’Hoste S, Barriere H, Belfodil R, Rubera I, Duranton C, TaucM, Poujeol C, Barhanin J, Poujeol P. Extracellular pH alkaliniza-tion by Cl�/HCO3

� exchanger is crucial for TASK2 activation byhypotonic shock in proximal cell lines from mouse kidney. Am JPhysiol Renal Physiol 292: F628–F638, 2007. First published Sep-tember 26, 2006; doi:10.1152/ajprenal.00132.2006.—We have previ-ously shown that K�-selective TASK2 channels and swelling-acti-vated Cl� currents are involved in a regulatory volume decrease(RVD; Barriere H, Belfodil R, Rubera I, Tauc M, Lesage F, PoujeolC, Guy N, Barhanin J, Poujeol P. J Gen Physiol 122: 177–190, 2003;Belfodil R, Barriere H, Rubera I, Tauc M, Poujeol C, Bidet M,Poujeol P. Am J Physiol Renal Physiol 284: F812–F828, 2003). Theaim of this study was to determine the mechanism responsible for theactivation of TASK2 channels during RVD in proximal cell lines frommouse kidney. For this purpose, the patch-clamp whole-cell techniquewas used to test the effect of pH and the buffering capacity of externalbath on Cl� and K� currents during hypotonic shock. In the presenceof a high buffer concentration (30 mM HEPES), the cells did notundergo RVD and did not develop outward K� currents (TASK2).Interestingly, the hypotonic shock reduced the cytosolic pH (pHi) andincreased the external pH (pHe) in wild-type but not in cftr �/� cells.The inhibitory effect of DIDS suggests that the acidification of pHi

and the alkalinization of pHe induced by hypotonicity in wild-typecells could be due to an exit of HCO3

�. In conclusion, these resultsindicate that Cl� influx will be the driving force for HCO3

� exitthrough the activation of the Cl�/HCO3

� exchanger. This efflux ofHCO3

� then alkalinizes pHe, which in turn activates TASK2 channels.

regulatory volume decrease; CFTR; KCNE5; potassium and chloridechannels; external and internal pH

IT IS NOW WELL ESTABLISHED that the control of cell volume is anessential phenomenon, which maintains the homeostasis ofnumerous cell functions. Most animal cells ensure maintenanceof their volume by controlling ion movements across theirplasma membrane (19), in response to osmolarity changes inthe extracellular medium. This is achieved through the activa-tion of specific channels and transporters. Epithelial cells, suchas those that constitute the different segments of the nephron,are particularly exposed to variations in extracellular osmolar-ity (11). These cells are subjected to an osmotic shock, eitherby accumulation of active osmolytes inside their cytoplasm(proximal tubule) or by dilution of the tubular fluid (distaltubule). In response to osmotic stress, these cells undergo aregulatory volume decrease (RVD) process characterized bythe exit of Cl� and K� ions, which finally drives water efflux(20). However, despite an impressive quantity of data in the

literature, the molecular identity of the Cl� and K� channelsactivated to achieve this regulation remains unclear (22).Moreover, the mechanisms precluding the activation of thesechannels are still controversial. This is probably due to the factthat, although RVD is an ubiquitous phenomenon, the nature ofthe channels involved in this process could vary depending onthe tissue under study (25). In the mouse proximal tubule, wepreviously demonstrated that during a hypotonic shock CFTRmodulated the swelling-activated Cl� currents by controlling acascade involving apical ATP release, adenosine production,and Ca2� entry (3). Concomitantly, the decrease in tonicityactivated a K� conductance through TASK2 channels (2, 5).Further work provided evidence that TASK2 channels could beinvolved in cell RVD. Moreover, these channels are known tobe sensitive to external pH (pHe) (13, 26, 29, 30, 32). TASK2channels belong to the large family of two-P domain K�

channels, some members of which are activated by cell swell-ing. Nevertheless, the exact mechanism responsible for theactivation of TASK2 channels during RVD is unknown. There-fore, it was of interest to determine why hypotonicity couldincrease TASK2 K� conductance. The high sensitivity of thesecurrents to pHe suggests that the increase in pHe induced byHCO3

� efflux activates TASK2. The related K� permeability ofTASK2 is concomitantly increased by hypotonic shock. In theproximal tubule, an increase in basolateral pHe could be due toHCO3

� transport (42). Such transport is achieved by the Na�-3HCO3

� cotransporter and by the Cl�/HCO3� exchanger. A

specific coupling of TASK2 activity to HCO3� transport

through external alkalinization was demonstrated by Warth etal. (42). Using isotonic conditions, these authors proposed amodel of TASK2 function in the proximal tubule in which Na�

and HCO3� leave the cell by Na�-3HCO3

� cotransport. Underhypotonic conditions, the involvement of HCO3

� in RVD wasalso clearly established in different epithelial-derived cells (28,36). However, although most of the authors agree with theobservation that the presence of HCO3

� is absolutely requiredfor ensuring RVD, the relative participation of the differentHCO3

� transporters is still being discussed. In the presentstudy, we examined the mechanism involved in the exit ofHCO3

� during a hypotonic shock in cell lines originating frommouse proximal tubule.

Here, we show that hypotonicity-induced cell swelling isfollowed by a rapid decrease in internal pH (pHi) and aconcomitant increase in pHe due to an efflux of HCO3

�. ThisCl�-dependent and Na�-independent efflux could be due to

Address for reprint requests and other correspondence: P. Poujeol, UMRCNRS 6548, Universite de Nice-Sophia Antipolis, 06108 Nice Cedex 2,France (e-mail: [email protected]).

The costs of publication of this article were defrayed in part by the paymentof page charges. The article must therefore be hereby marked “advertisement”in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Am J Physiol Renal Physiol 292: F628–F638, 2007.First published September 26, 2006; doi:10.1152/ajprenal.00132.2006.

0363-6127/07 $8.00 Copyright © 2007 the American Physiological Society http://www.ajprenal.orgF628

selective activation of the Cl�/HCO3� exchanger. The increase

in pHe activates TASK2 K� channels, allowing the develop-ment of swelling-activated K� currents.

MATERIALS AND METHODS

Transformation of Primary Cultures with pSV3 neo

The primary cell culture technique has been described in detail inprevious studies (4, 5). Ten-day-old primary cultures of S1 and S2segments of proximal tubules from wild-type, cftr �/� (35), and task2�/� mice were transfected with pSV3 neo using lipofectin (Invitro-gen). After 48 h, selection of the clones was performed by the additionof 500 �g/ml G418. Culture medium M1 containing G418 waschanged every day. Resistant clones were isolated, subcultured, andused after 10 trypsinization steps.

BCECF Cell Volume Measurement

The relative cell volume was monitored by image analysis withBCECF-AM used as a fluorescent volume indicator, as previouslyreported (31, 37). At 450 nm (isobestic point), BCECF fluorescence ispH insensitive. The emitted fluorescent signal at 520 nm was thereforerelated to the intracellular dye concentration only and reflected thevariations in cell volume (37). Briefly, proximal cell lines grown in35-mm petri dishes were incubated in the presence of 4 �MBCECF-AM (final concentration) at 37°C for 15 min in a humidifiedatmosphere of 5% CO2-95% air. Cells were then incubated in isotonicHCO3

�-buffered solutions containing (in mM) 80 NaCl, 15 NaHCO3,5 KCl, 1 CaCl2, with low (1 HEPES) or high buffering capacity (30HEPES), pH 7.4, in a 5% CO2 atmosphere (Fig. 1B). These solutionswere adjusted to 290 or 230 mosmol/kgH2O by addition of mannitol.The relative change in cell volume was estimated from the fluorescentsignal by assuming that a 30% decrease in osmolality caused adecrease in the fluorescent signal corresponding to a minimum swell-ing of 30% of the initial volume. The means of relative volumechanges were obtained by analysis of 10–25 cells in each culture. Cellvolume variations were calibrated according to the method describedby Tauc et al. (37).

pHi Measurements

The fluorescent pH indicator BCECF-AM was also used to mea-sure pHi, as described in detail previously (7). Proximal cell linesgrown in 35-mm petri dishes were incubated in the presence of 4 �MBCECF-AM at 37°C for 15 min in a humidified atmosphere of 5%CO2-95% air. Loaded cells were carefully rinsed and placed on thestage of an inverted microscope. The cells were excited successivelyat 450 and 490 nm, and the emitted light was recorded at 520 nm. Atthe end of each experiment, the fluorescence signals related to pHi

changes were calibrated using the K�/H� exchange ionophore nigeri-cin. For this purpose, the cells were perfused with KCl solutionscontaining (in mM) 140 KCl, 1 CaCl2, and 10 HEPES, the pH ofwhich was adjusted to 8.0, 7.0, and 6.0, respectively, with Tris buffer.Finally, 10 �M nigericin was added to all of these solutions.

The cell-buffering capacity was determined by the NH4� technique

(15, 33). pHi was recorded after addition of 20 mM NH4Cl. Thebuffering capacity was calculated according to the method of Roosand Boron (33). From the above two parameters (pHi and bufferingcapacity), the H� efflux as a function of time was calculated asfollow: H� efflux � buffering capacity � �pHi (inmmol � l�1 �min�1). Under all experimental conditions, the fluores-cence was recorded continuously and �pHi was calculated using thefirst 30-s initial rate of H� efflux (6). The initial rate of the change inpHi (�pHi/min) was calculated using linear regression analysis. Toensure an adequate renewal of the medium, the solutions wereperfused at a rate of 2 ml/min.

To determine the activity of the Cl�/HCO3� exchanger at physio-

logical pH values, cells were first incubated in HCO3�-buffered NaCl

solution containing (in mM) 125 NaCl, 15 NaHCO3, 5 KCl, 1 CaCl2,5 glucose, and 10 HEPES, pH 7.4. The solution was then replaced bya Cl �-free solution containing (in mM) 125 sodium gluconate, 15NaHCO3, 5 potassium gluconate, 3 calcium gluconate, 5 glucose, and10 HEPES, pH 7.4. Under these conditions, intracellular Cl� israpidly exchanged against extracellular HCO3

�, and pHi is increased.

pHe Measurement

The nonesterified form of BCECF was used to assess pHe. Proxi-mal cell lines from wild-type, task2 �/�, and cftr �/� mice weregrown on 100-mm petri dishes. After trypsinization, cells were cen-trifuged and resuspended in 500 �l of either isotonic or hypotonicsolutions containing (in mM) 105 NaCl, 5 KCl, 1 CaCl2, 5 glucose,and 1 HEPES, pH 7.0; when necessary, isotonicity was adjusted byadding 72 mM mannitol. The suspension was adjusted to a final

Fig. 1. Effect of pH and buffering capacity of external bath solution onregulatory volume decrease (RVD). A: effect of external pH (pHe) on hypo-tonicity-induced RVD in cultured proximal convoluted tubule (PCT) cell linefrom wild-type mice. Cell volume was measured using BCECF-AM. After acontrol period in an isotonic bath solution (290 mosmol/kgH2O), a hypotonicshock was induced by perfusing the cells with a hypotonic bath solution (230mosmol/kgH2O). Relative volume changes (�SE) as the percentage of initialvolume were plotted against time. Experiments were performed at variousextracellular pH, ranging from 6.0 to 8.0 as indicated. Measurements wereperformed on 8 different monolayers (19–25 random cells each) at each pH. B:effect of buffering capacity of external bath on RVD in proximal cell line fromwild-type mice. After a control period with cells in an isotonic bath solution(290 mosmol/kgH2O), a hypotonic shock was induced by perfusing the cellswith a hypotonic bath solution (230 mosmol/kgH2O) containing either 1 or 30mM HEPES at pH 7.4 or 30 mM HEPES at pH 8.0. Relative volume changes(�SE) as the percentage of initial volume were plotted against time. Measure-ments were performed on 8 different monolayers (21–23 random cells each) ateach HEPES concentration.

F629TASK2 ACTIVATION BY HCO3� DURING HYPOTONIC SHOCK IN PCT

AJP-Renal Physiol • VOL 292 • FEBRUARY 2007 • www.ajprenal.org

number of 2 � 107 cells/ml. Cells were maintained in isotonic orhypotonic medium for 8 min in a humidified atmosphere of 5%CO2-95% air. Cells were then centrifuged at 1,300 g for 2 min, andBCECF (25 �M) was added to the supernatant. The fluorescence ofthe supernatant was rapidly measured using a spectrofluorimeter(Safas, Monaco). For this purpose, the supernatants were excitedsuccessively at 490 and 450 nm and the emitted fluorescence wasrecorded at 520 nm. At the end of each experiment, the fluorescencevalues were converted to pH units using calibration curves performedwith solutions adjusted at pHe ranging from 6.0 to 8.0.

Electrophysiological Studies

Whole-cell currents were recorded from cultured cells grown oncollagen-coated supports (35-mm petri dishes) maintained at 37°C forthe duration of the experiments. The ruptured whole-cell configurationof the patch-clamp technique was used. After formation of a gigaseal,the fast-compensation system of the amplifier was used to compensatefor the intrinsic input capacitance of the head stage and the pipettecapacitance. The membrane was ruptured by additional suction toachieve the conventional whole-cell configuration. Settings availableon the amplifier were used to compensate for cell capacitance. Theseries resistances were not compensated, but experiments in which theseries resistance was �20 M were discarded. The offset potentialsbetween both electrodes were zeroed before sealing, and the liquidjunction potentials were measured experimentally before each exper-iment and corrected accordingly (measured junction potentials werenegligible for Cl� conductance experiments and were 9.96 � 0.91mV for K� conductance experiments). Solutions were perfused in theextracellular bath using a four-channel glass pipette, with the tipplaced as close as possible to the clamped cell. Voltage-clampcommands, data acquisition, and data analysis were controlled via aVP 500 amplifier (Biologic) connected to a computer. The whole-cellcurrents resulting from voltage stimuli were sampled at 2.5 kHz andfiltered at 1 kHz. Cells were held at �50 mV, and 400-ms pulses from�100 to �100 or �120 mV were applied in 20-mV increments.

The composition of the pipette and bath solutions is described inTable 1.

Chemical Compounds

5-Nitro-2-(3-phenylpropylamino)-benzoic acid (NPPB; Calbio-chem) was prepared at 100 mM in DMSO. DIDS or 4–4-dinitros-tilbene-2, 2-disulfonic acid (DNDS) was directly dissolved in themedium at a final concentration of 1 mM. Diphenylamine-2-carbox-ylate (DPC) was prepared at 1 M stock solution in DMSO.

DIDS, DPC, and forskolin were obtained from Sigma-Aldrich(Saint Quentin Fallavier, France), while DNDS, BCECF-AM, andBCECF were obtained from Molecular Probes (Leiden, The Nether-

lands). Clofilium was prepared at 10 mM in 50% DMSO-50% water.Clofilium was a gift from Dr. Barhanin (UMR CNRS 6097).

RESULTS

Effect of pH and Buffering Capacity of External BathSolution on RVD

Relative cell volume variation during a hypotonic shock wasmeasured using fluorescence-videomicroscopy in proximal celllines from wild-type mice. The cell monolayers were bathed inan isotonic solution (290 mosmol/kgH2O) and then perfusedcontinuously with a hypotonic solution (230 mosmol/kgH2O).As illustrated in Fig. 1A, RVD was recorded at different pHe,varying from 6.0 to 8.0. Interestingly, the RVD phenomenonwas minimal at pHe � 6.0 and maximal at pHe � 8.0.

Based on this result, we investigated whether an increase inpHe was necessary to ensure RVD. For this purpose, RVD wasrecorded in cells bathed in weakly (1 mM HEPES) or highly(30 mM HEPES) buffered solutions (Fig. 1B). In the presenceof a highly buffered solution, the cells did not undergo RVD,indicating that pHe variations influenced the RVD response. Toexclude the possibility that high concentrations of HEPESinhibited the RVD phenomenon, relative cell volume variationwas recorded in a highly buffered bath solution (30 mM), withpH adjusted to 8.0. Under this condition, cells underwentclassic RVD, as illustrated in Fig. 1B. Taken together, theseresults confirmed that external alkalinization was required toensure RVD during a hypotonic shock. Experiments were thenperformed to identify the precise mechanism involved in thisalkalinization.

Effect of Cl� and Na� Removal on RVD

In the proximal tubule, pHe is modulated by the exit ofHCO3

� mainly through Na�-3HCO3� cotransport and/or Cl�/

HCO3� exchangers. To test the requirement of Na� and Cl�

during the RVD phenomenon, relative cell volume variationsduring a hypotonic shock were determined in the presence orabsence of Na� or Cl�, respectively. In the experimentsdescribed in Fig. 2A, proximal monolayers from wild-typemice were first incubated for 60 min in a Na�-free bathsolution (pH � 7.4) and the hypotonic shock was then per-formed in the absence of Na� in the bath solution. The removalof Na� did not significantly modify the RVD process. In a

Table 1. Composition of solutions used in whole-cell clamp experiments

Cl� Currents K� Currents

Pipette solutionNMDG-Cl pH � 7.4

Bath solutionNMDG-Cl

Pipette solution Kgluconate pH � 7.4

Bath solutionNMDG-Cl

K gluconate 100 5KCl 20HEPES 10 1/30 10 1/30EGTA 5 5MgATP 5 5NMDG-Cl 140 110 90CaCl2 1 1Glucose 5 5KHCO3

� 25Mannitol 30/60/120 0/30/60Osmolality, mosmol/kgH2O 290 290/350 290 230/290

Concentrations are given in mM. When necessary, bath solutions were made hyperosmotic by addition of mannitol. NMDG, N-methyl-D-glucamine.

F630 TASK2 ACTIVATION BY HCO3� DURING HYPOTONIC SHOCK IN PCT


second set of experiments (Fig. 2B), the hypotonic shock wasperformed in the absence of external Cl� (pH 7.4). Under theseconditions, the cells became swollen but did not exhibit RVD,indicating that Cl� contributed to the RVD. However, theabsence of RVD in Cl�-free solutions could be the conse-quence of two processes. 1) RVD depended on the efflux ofCl� through Cl� channels only. In this case, the massive lossof intracellular Cl� due to the acute removal of external Cl�

could decrease the intracellular Cl� ion pool and stronglyreduced the activity of volume-sensitive Cl� channels. 2) RVDdepended on both an efflux of Cl� through the volume-sensitive Cl� channel and on the efflux of HCO3

� through theCl�/HCO3

� exchanger, which could increase pHe. In this case,the removal of external Cl� could decrease this HCO3

� effluxand could reverse the driving force of Cl�. Consequently,HCO3

� entry precluded pHe alkalinization.To verify these hypotheses, the RVD phenomenon was

assessed in cells suspended in a Cl�-free solution with pHadjusted to 8.0. As illustrated in Fig. 2C, the hypotonic shockwas followed by a RVD. This observation indicated that whenthe external bath pH was maintained at alkaline pH, the acuteCl� removal did not inhibit the RVD process. Thus, in theexperiments described in Fig. 2B, the absence of RVD wasprobably the consequence of an alkalinization defect of pHe

due to a dysfunction of the Cl�/HCO3� exchanger (see hypoth-

esis 2, above).

Involvement of Cl�/HCO3� Exchanger in RVD

The above experiments suggested the involvement of theCl�/HCO3

� exchanger in the RVD phenomenon. To confirmthis hypothesis, relative cell volume variations during a hypo-tonic shock were measured in a proximal tubule cell line in thepresence of DIDS or DNDS (DIDS analog). Two experimentalconditions were used. In a first set of experiments, the hypo-tonic shock was performed with an acute application of DIDSor DNDS. As expected, each inhibitor prevented RVD duringthe hypotonic shock (Fig. 3). In a second set of experiments,cells were preincubated for 60 min with DIDS or DNDS andthoroughly rinsed for 10 min to remove these drugs. Afterward,a hypotonic shock was applied with an inhibitor-free solution.Preincubation with DIDS prevented the RVD (Fig. 3). Incontrast, cells preincubated with DNDS returned to their initialvolume in response to the hypotonic shock. As expected, theabsence of RVD during the acute application of DIDS orDNDS was mainly due to an inhibition of the volume-sensitiveCl� channels since these drugs are potent inhibitors of thesechannels in most cells. By contrast, the absence of RVD incells preincubated with DIDS and rinsed out is probably due toa specific inhibition of the Cl�/HCO3

� exchanger becausesustained DIDS application could irreversibly block this ex-changer.

Cl�/HCO3� Exchanger Activity in Proximal Cell Lines

To further confirm this possibility, we estimated the activityof this exchanger by recording pHi variations during externalCl� removal. For this purpose, a proximal cell line fromwild-type mice loaded with BCECF was maintained in NaClsolutions containing HCO3

� and continuously gassed with 5%CO2. The pHi was recorded by fluorescence microscopy. Fig-ure 4A gives an example of the time course of the pHi

Fig. 2. Effect of Cl� and Na� removal on RVD. A: effect of Na� substitutionon hypotonicity-induced RVD in cultured PCT cell line from wild-type mice.After a control period (isotonic bath solution, 290 mosmol/kgH2O), a hypo-tonic shock was induced by perfusing the cells with a hypotonic bath solution(230 mosmol/kgH2O) in the presence or absence of Na�. Measurements wereperformed on 7 different monolayers with or without Na� (20–23 random cellsfrom each monolayer). B: effect of Cl� substitution on hypotonicity-inducedRVD in cultured PCT cell line from wild-type mice. After a control period(isotonic bath solution, 290 mosmol/kgH2O), a hypotonic shock was inducedby perfusing the cells with a hypotonic bath solution (230 mosmol/kgH2O) inthe presence or absence of Cl�. Measurements were performed on 7 differentmonolayers with or without Cl� (19–22 random cells from each monolayer).C: effect of extracellular alkalinization on hypotonicity-induced RVD incultured PCT cell line from wild-type mice. Experiments were performed as inB using an external bath solution adjusted to pH 8.0 instead of 7.4. Measure-ments were performed on 6 different monolayers with or without Cl� (10–20random cells from each monolayer).



variations. In a first step, external Cl� was replaced by glu-conate, causing pHi to increase over a 6-min period. This effectwas reversed when the gluconate solution was replaced withthe NaCl solution. The initial rates of pHi increase are reportedin Fig. 4B. Compared with the control condition, the acuteapplication of DIDS or DNDS strongly blocked the pHi in-crease induced by Cl� removal. In the second set of experi-ments, the cells were treated with DIDS or DNDS for 60 min.Inhibitors were then removed from the solution, and the pHi

increase was recorded. Figure 4B clearly shows that sustainedincubation with DIDS precluded the pHi increase, whereasincubation with DNDS did not modify the pHi increase in-duced by external Cl� removal. Taken together, these datasuggested that during chronic application, DIDS bound co-valently to the Cl�/HCO3

� exchanger and irreversibly inhibitedits activity. This inhibition resulted in a concomitant inhibitionof the RVD, confirming the involvement of this exchanger inthe control of cell volume during a hypotonic shock.

Further experiments were also performed to identify whetherthe Cl�/HCO3

� exchanger was active in two different cell linesalready described as being unable to regulate their volumeduring hypotonic shock (task2 �/� and cftr �/�) (2, 3). Asillustrated in Fig. 4C, the rate of rise in pHi following Cl�

removal was identical in cells from wild-type, task2 �/�, andcftr �/� mice. These data clearly indicated that the Cl�/HCO3

�

exchanger was functional in the different cell lines indepen-dently of the presence or absence of CFTR or TASK2 chan-nels.

Measurement of pHi and pHe Variations During RVD

To further prove that the Cl�/HCO3� exchanger was in-

volved in the pHe increment during RVD, we monitored pHi

and pHe changes during a hypotonic shock.

pHi. As reported in Table 2, the buffering capacity was notsignificantly different from one cell line to another. The pHi

behavior of cells bathed in HCO3� medium is illustrated in Fig.

5A. In proximal tubule cells from wild-type mice, the hypo-tonic shock reduced pHi from 7.51 � 0.01 to 7.01 � 0.05 (n �7) within 4–5 min. Afterward, pHi increased to stabilize at7.48 � 0.03, a value not significantly different from the valuerecorded just before the onset of the shock. In cells from task2�/� mice which did not undergo RVD (2), the hypotonic shockinduced a similar decrease in pHi (7.42 � 0.02 to 6.81 � 0.03,

Fig. 4. Measurement of internal (cytosolic) pH (pHi) variations in proximalcell lines from wild-type, task2 �/�, and cftr �/� mice during extracellular Cl�

removal. A: cells were loaded for 15 min with BCECF, and the fluorescencewas recorded during repetitive replacements of the bath by Na gluconatesolution (Cl� free) or NaCl solution in the absence or presence of DIDS (1mM). The sequence of bath solution substitution is indicated (top). Allsolutions were buffered with NaHCO3 (15 mM), and experiments wereperformed at constant PCO2 (5%).Values are means � SE of 6 monolayers. B:initial rate of pHi increment (�pH/min) was measured after replacement ofexternal NaCl solution by a Cl�-free solution (Na-gluconate) in the absence(control) or presence of DIDS (1 mM) or DNDS (1 mM). Additional experi-ments were performed with cells preincubated for 1 h with DIDS and DNDSand then thoroughly washed to remove the inhibitors. Values are means � SEof 6 different experiments for each experimental conditions. C: �pH/min wasmeasured after replacement of external NaCl solution by a Cl�-free solution(Na-gluconate) in proximal cell lines from wild-type, task2 �/�, and cftr �/�

mice. Values are means � SE of 6 monolayers/each mouse strain.

Fig. 3. Involvement of Cl �/HCO3� exchanger activity on RVD in cultured

PCT cell lines from wild-type mice. Cell volume was measured by anelectronic sizing technique with a CASY 1 cell counter (Scharfe System). Cellswere suspended in casyton solution (NaCl isotonic solution) for controlcondition (F) and then in 30% diluted casyton solution in the absence orpresence of DIDS or 4–4-dinitrostilbene-2, 2-disulfonic acid (DNDS). Valuesare expressed as percentage of cell volume variations measured during hypo-tonic shock in the absence of inhibitors (E), in the presence of DIDS (1 mM,�), or DNDS (1 mM, ■ ). Cell volume measurements were also performed withcells preincubated for 1 h with DIDS (1 mM, ‚) or DNDS (1 mM, �) and thenthoroughly washed to remove the inhibitors. Values are means � SE of 5–7different experiments.



n � 7), which was also followed by pHi recovery (pHi �7.44 � 0.05). In both cell lines, the application of 1 mM DIDScompletely prevented the pHi changes during the hypotonicshock.

The pHi variation during the hypotonic shock was alsomeasured in a proximal tubule cell line from cftr �/� mice. Inthese cells, the hypotonic shock did not induce significant pHi

change (Fig. 5A) and did not activate volume-sensitive Cl�

channels (3). Therefore, the pHi decrement induced by hypo-tonicity in wild-type cells was probably related to the presenceof functional volume-sensitive Cl� channels.

pHe. Experiments were performed to test whether a hypo-tonic shock could also increase pHe during the time course ofthe RVD. pHe was determined in cells obtained by trypsiniza-tion of proximal monolayers from wild-type, cftr �/�, andtask2 �/� mice and resuspended in a weakly buffered solution(1 mM HEPES) maintained at pH � 7.0. The pHe changeswere measured 8 min after the onset of the hypotonic shock. Asillustrated in the histogram in Fig. 5B, the hypotonic shockperformed on cells from wild-type and task2 �/� mice induceda significant increment of pHe (control: 7.01 � 0.02; wild-typecells: 7.35 � 0.03, and task2 �/� cells: 7.32 � 0.05, for n �6 for all experimental conditions). This increment of pHe wasnot detected in the presence of 1 mM DIDS, 100 �M NPPB, or1 mM DPC. Moreover, the hypotonic shock was unable topromote pHe variations in experiments performed with cftr �/�

cells (Fig. 5B).The inhibitory effects of DIDS suggested that the acidifica-

tion of pHi and the alkalinization of pHe induced by hypoto-nicity in wild-type and task2 �/� cell lines could be due to anexit of HCO3

�. Moreover, the inhibitory effect of NPPB andDPC in wild-type cells indicated also the involvement of Cl�

channels in the pHe alkalinization mediated by the Cl�/HCO3�

exchanger. This hypothesis is supported by the inability of thecftr �/� cell line to alkalinize the pHe or to acidify the pHi

during the hypotonic shock. These experiments suggested thatalkalinization of pHe by the Cl�/HCO3

� exchanger was aprerequisite to generate the RVD during a hypotonic shock.

In the following study, whole-cell experiments were per-formed to assess the effect of external pH variations on thevolume-sensitive Cl� and K� conductances during RVD.

Effect of Buffering Capacity of External Bath Solution onCl� and K� Currents Activated During RVD

During these experiments the mean cell membrane capaci-tances were 20.1 � 1.4, 18.6 � 1.3, and 19.2 � 1.1 (n � 20)

for wild-type, task2 �/�, and cftr �/� proximal cells, respec-tively, and they did not differ significantly between the celllines (paired t-test).

Cl� currents. To test the effect of buffering capacity onswelling-activated Cl� currents, whole-cell Cl� conduc-tance was recorded during a hypotonic shock. Whole-cellcurrents were recorded with Ca2�-free (5 mM EGTA)NMDGCl pipette solution (osmolality of 290 mosmol/kgH2O). In Fig. 6A, control currents were measured with anextracellular weakly buffered solution (1 mM HEPES, os-molality of 350 mosmol/kgH2O). The monolayers were thenperfused with a 290 mosmol/kgH2O solution. In more than95% of the cells, whole-cell currents increased within 1 minand reached maximum amplitude after 4 –5 min. These largeoutwardly rectifying currents showed a time-dependent in-activation at depolarizing potentials �60 mV. When thecells were exposed to 100 �M NPPB, the currents returnedto the control value within 2–3 min. Increasing the buffering

Fig. 5. Measurement of pHi and pHe variations during RVD in proximal celllines from wild-type, task2 �/�, and cftr �/� mice. A: for pHi measurement, thecells were loaded for 15 min with BCECF-AM and washed in isotonic NaClsolution. pHi variations were then recorded during hypotonic shock induced byperfusion of a hypotonic NaCl bath solution (230 mosmol/kgH2O). A pHcalibration protocol was performed at the end of each experiment by theperfusion of nigericin containing solutions adjusted to various pH values asindicated. Values are means � SE of n different cell cultures. B: for pHe

measurement, the cells were incubated for 8 min in isotonic or hypotonic NaClsolutions. After centrifugation, BCECF (25 �M) was added to the supernatant.The fluorescence of the supernatant was rapidly measured using a spectrofluo-rimeter. Fluorescence units were converted in pH units using a pH calibrationcurve. Values are means � SE of 6 different cell cultures. NPPB, 5-Nitro-2-(3-phenylpropylamino)-benzoic acid; DPC, diphenylamine-2-carboxylate;

Table 2. Cell-buffering capacity

Cell-Buffering Capacity,mmol H� � l�1 � pH�1

Number ofExperiments

TCP task2�/� 29.7�0.9 6TCP wild-type 28.4�0.4 7TCP cftr�/� 29.1�0.7 6

Values are means � SE. The cell-buffering power was determined by theNH4

� technique (15, 33). Intracellular pH (pHi) was recorded after addition of20 �M NH4Cl. The buffering capacity was calculated according to the methodof Roos and Boron (33). From the above 2 parameters (pHi and bufferingpower), the H� efflux as a function of time may be calculated as follows: H�

efflux � buffering power � �pHi (in mmol � l�1 � min�1). In all experimentalconditions, the fluorescence was recorded continuously, and �pHi measuredduring the first 30 s was used to calculate the initial rate of H� efflux (6).



capacity of the bath solution (30 mM HEPES) did notmodify the development of the swelling-activated Cl� cur-rents (Fig. 6B). Finally, these swelling-activated Cl� cur-rents did not depend on the buffering capacity of theextracellular bath. Proximal cells from task2�/� mice ex-hibited swelling-activated Cl� currents sharing similar char-acteristics (data not given).

K� currents. It has previously been suggested (2, 5) thatTASK2 could be the K� channel involved in cell volumeregulation. Since TASK2 is activated at alkaline pH, weinvestigated the effect of the buffering capacity of the bathsolution on the activation of K� conductances during a hypo-tonic shock in proximal cell lines from wild-type and task2 �/�

mice. Whole-cell recordings were performed on confluentmonolayers bathed in a HCO3

�-free, weakly (1 mM HEPES,pH � 7.4) or highly (30 mM HEPES, pH � 7.4 or 8.0)buffered solutions. The currents were recorded with Ca2�-freepipette solutions containing 25 mM HCO3

� (osmolality of 290mosmol/kgH2O). Experiments were performed in the presenceof 100 �M NPPB to avoid the development of volume-activated Cl� currents. Figure 7A shows K� currents recordedin proximal cells from wild-type mice bathed in weakly buff-ered solution (1 mM HEPES, pH 7.4, osmolality of 290mosmol/kgH2O). The voltage-step protocol elicited time-inde-pendent outwardly rectifying currents, with a reversal potentialof �61.2 � 4.0 mV and a maximal slope conductance of 3.6 �0.6 nS (n � 5, Fig. 7A). The monolayers were then perfusedwith a 230 mosmol/kgH2O solution. This hypotonic shockinduced the development of K� currents within 4–5 min.These large outwardly rectifying currents reversed at �62.3 �4.1 mV with a maximum slope conductance of 26.6 � 1.6 nS(n � 5, Fig. 7A). These currents were strongly inhibited by theapplication of 10 �M clofilium (Fig. 7A).

When similar hypotonic shock experiments were performedin a highly buffered bath solution (30 mM HEPES, pH 7.4, 230mosmol/kgH2O), no significant increase in the outward K�

currents was observed (Fig. 7B). To test whether a high HEPESconcentration modified the swelling-activated K� currents, themonolayers were perfused with a highly buffered hypotonicbath solution (30 mM HEPES, pH 8.0). Under this condition(Fig. 7B), large outwardly rectifying K� currents were re-corded with a reversal potential of �57.3 � 3.4 mV and aslope conductance of 30.2 � 3.1 pS (n � 5). These dataindicated that high HEPES concentrations did not inhibitTASK2 K� currents.

As expected for a K� current flowing through TASK2channels, no swelling-activated K� currents were observed intask2 �/� cells (Fig. 7C).

Finally, the data strongly suggest that cell swelling inducedan increase in pHe, which activated TASK2 K� currents,leading finally to the RVD process. The increase in pHe isprobably due to the extrusion of HCO3

� through the Cl�/HCO3�

exchanger.

Involvement of Cl�/HCO3� Exchanger in Activation of K�

Currents During RVD

To test whether the Cl�/HCO3� exchanger was involved

in the activation of K� conductance, whole-cell currentswere recorded during a hypotonic shock in a weakly buff-ered (1 mM HEPES) bath solution. The experimental solu-tions were chosen to promote a Cl�/HCO3

� exchange byincreasing the outward gradient of HCO3

� and the inwardgradient of Cl�. Under these control conditions, the hypo-tonic shock activated TASK2 K� currents (see Fig. 7A). Incontrast, preincubation of proximal cells with 1 mM DIDSfor 60 min followed by intensive washout impaired thedevelopment of outwardly rectifying K� conductance dur-ing the hypotonic shock (Fig. 8A). Thus the irreversibleinhibition of the Cl�/HCO3

� exchanger due to covalentDIDS binding strongly decreased TASK2 K� conductance.To discard the possibility that this inhibition of TASK2 K�

conductance could be indirectly due to a blockade of Cl�

Fig. 6. Effect of buffering capacity on development of hypotonicity-inducedCl� currents in proximal cell lines from wild-type mice. A: whole-cell currentsin PCT cell line from wild-type mice were recorded in weakly buffered bathsolutions (1 mM HEPES). Cl� currents were measured in control hypertonicsolution (350 mosmol/kgH2O), 4–5 min after replacement of the bath by anisotonic solution (290 mosmol/kgH2O), and finally in the presence of NPPB(100 �M). Membrane voltage was held at �50 mV and stepped to testpotential of �100 to �120 mV in 20-mV increments. Shown are whole-cellrecordings and corresponding current-voltage (I-V) relationships. Values aremeans � SE of 5 cells from 5 different monolayers. B: whole-cell currents inPCT cell line from wild-type mice were recorded in highly buffered bathsolutions (30 mM HEPES). Cl� currents were measured in control hypertonicsolutions (350 mosmol/kgH2O), 4–5 min after replacement of the bath by anisotonic solution (290 mosmol/kgH2O), and finally in the presence of NPPB(100 �M). Shown are whole-cell recordings and corresponding I-V relation-ships. Values are means � SE of 5 cells from 4 different monolayers.



conductance by DIDS, further experiments were performedwith another Cl� channel blocker. The presence of 100 �MNPPB during the hypotonic shock did not modify thedevelopment of K� conductance (Fig. 8B).

Role of pHe in Activation of K� Currents by Hypotonicity

To further demonstrate the effect of pHe on the activation ofK� currents by a hypotonic shock, whole-cell currents wererecorded in weakly buffered bath solutions (1 mM HEPES)containing 1 mM DIDS. The results are reported in Fig. 9. Inan isotonic bath solution (290 mosmol/kgH2O), increasing pHe

from 7.4 to 8.0 enhanced outward K� conductance (maximalslope conductance � 3.9 � 0.7 nS at pHe � 7.4 and 18.1 � 1.2nS at pHe � 8.0, n � 5, Fig. 9). Interestingly, the perfusion ofa hypotonic solution (230 mosmol/kgH2O) at pHe � 8.0induced a further increase in the outward K� current with amaximal slope conductance of 29.9 � 2.8 nS (n � 5). Thiscurrent was markedly blocked by clofilium (10 �M).

Fig. 8. Involvement of Cl�/HCO3� exchanger activity in activation of K�

currents during RVD in proximal cell lines from wild-type mice. A: whole-cellcurrents in DIDS (1 mM)-preincubated PCT cells from wild-type mice wererecorded in control condition (isotonic solution, 290 mosmol/kgH2O) and after4–5 min of extracellular perfusion of a hypotonic solution (230 mosmol/kgH2O). Before the patch-clamp experiments were performed, the preincu-bated cells were thoroughly washed to eliminate unbound DIDS. Membranevoltage was held at �50 mV and stepped to test potential of �100 to �120mV in 20-mV increments. Shown are whole-cell recordings and correspondingI-V relationships. Values are means � SE of 6 cells from 4 different mono-layers. B: whole-cell currents in a PCT cell line from wild-type mice wererecorded in the continuous presence of NPPB (100 �M) in control condition(isotonic solution, 290 mosmol/kgH2O) or after 4–5 min of extracellularperfusion of a hypotonic solution (230 mosmol/kgH2O). At the end of theexperiment, a hypotonic solution containing DIDS (1 mM) was perfused.Shown are whole-cell recordings and corresponding I-V relationships. Valuesare means � SE of 6 cells from 4 different monolayers.

Fig. 7. Effect of buffering capacity of external bath solution on K� currents activatedduring RVD in proximal cell lines from wild-type and task2 �/� mice. A: whole-cellcurrents in PCT cells line from wild-type mice were recorded in weakly buffered bathsolutions (1 mM HEPES, pHe � 7.4). K� currents were measured in control isotonicsolutions (290 mosmol/kgH2O), 4–5 min after replacement of the bath by a hypotonicsolution (230 mosmol/kgH2O), and finally in the presence of clofilium (10 �M). Thesolutions contained NPPB (100 �M). Membrane voltage was held at �50 mV andstepped to test potential of �100 to �120 mV in 20-mV increments. Shown arewhole-cell recordings and corresponding I-V relationships. Values are means � SE of5 cells from 5 different monolayers. B: whole-cell currents in PCT cells line fromwild-type mice were recorded in highly buffered bath solutions (30 mM HEPES). K�

currents were measured in control isotonic solution (290 mosmol/kgH2O, pHe � 7.4),4–5 min after replacement of the bath by a hypotonic solution (230 mosmol/kgH2O,pHe � 7.4), 4–5 min after replacement of the bath by a hypotonic solution (230mosmol/kgH2O, pHe � 8.0), and finally in the presence of clofilium (10 �M). Thesolutions contained NPPB (100 �M). Shown are whole-cell recordings and corre-sponding I-V relationships. Values are means � SE of 5 cells from 4 monolayers. C:whole-cell currents in PCT cell line from task2 �/� mice were recorded in weaklybuffered bath solutions (1 mM HEPES, pHe � 7.4). K� currents were measured incontrol isotonic solution (290 mosmol/kgH2O) and 4–5 min after replacement of thebath by a hypotonic solution (230 mosmol/kgH2O).The solutions contained NPPB(100 �M). Shown are whole-cell recordings and corresponding I-V relationships.Values are means � SE of 5 cells from 6 different monolayers.



Therefore, these results strongly suggested that alkaliniza-tion of the extracellular medium is a prerequisite to trigger theactivation of TASK2 during a hypotonic shock.

DISCUSSION

In the different segments of the nephron, cells are subjectedto osmotic shocks, either by accumulation of active osmolytesin their cytoplasm (proximal tubule) or by dilution of thetubular fluid (distal tubule). In response to such osmotic stress,these cells undergo a RVD process by activating swelling-sensitive Cl� and K� conductances (5). Of the K� channelsinvolved in the osmotic response, TASK2 K� channels play anessential role in the proximal tubule (2). However, the mech-anisms responsible for TASK2 activation during a hypotonicshock are not yet established. Interestingly, the TASK2 K�

current shows a strong dependence on pHe, being activated atalkaline pH within physiological ranges of variation (21). Thisproperty leads us, first, to investigate whether the activation ofTASK2 by a hypotonic shock could result from an increase inpHe. For this purpose, we have developed immortalized celllines from primary cultures of proximal tubules from wild-typeand task2 �/� mice. All the cell lines formed monolayers, andthe wild-type cells were capable of RVD mediated by Cl� andK� conductances. These conductances shared similar biophys-ical and pharmacological features with those described inprimary cultures of mouse proximal tubules (2, 5). Severalfindings in the present study indicate that a pHe incrementcould mediate the RVD. For example, we demonstrated thatthe RVD phenomenon was strongly dependent on pHe, with aninhibition at acidic pH and activation at alkaline pH. Interest-ingly, an effect of pH on RVD has already been reported inEhrlich cells submitted to a hypotonic shock (18). If we linkthis finding to the observation that a high buffering capacityalso blocks RVD, then we could assume that the hypotonicshock triggered an increase in pHe. Direct measurement of pHe

during RVD confirmed this assumption since the hypotonicshock induced an increase in pHe by �0.3 pH units.

In the present study, it was demonstrated that swelling-activated Cl� channels were insensitive to pHe. Therefore, the

pHe sensitivity of the RVD phenomenon is probably conferredby swelling-activated TASK2 K� conductance. The blockadeof TASK2 currents by highly buffered concentrations duringhypotonicity corroborated the essential role of this channel.

In isotonic conditions, the activation of TASK2 channelswas demonstrated to be mediated by the rise in pHe induced byHCO3

� efflux (42). It is probably also the case in hypotonicconditions since the TASK2 current was not observed in thepresence of DIDS. However, the increase pHe alone was notsufficient to account for the increase in K� conductance duringthe hypotonic shock. Reciprocally, the hypotonic shock wasinefficient to enhance TASK2-mediated K� currents when theincrease in pHe was prevented. Thus alkalinization of extra-cellular medium is a prerequisite to trigger the activation ofTASK2 during a hypotonic shock. Moreover, the observationthat RVD occurred concomitantly with an intracellular DIDS-sensitive acidification suggested that the hypotonic shock in-duced an exit of HCO3

� (14, 28). The role of HCO3� in RVD (8)

and the involvement of the Cl�/HCO3� exchanger in HCO3

�

transport during RVD (9, 16, 28, 39) have already beensuggested in previous experiments. This is shown to be thecase in the present study because RVD was insensitive to Na�

suppression and was inhibited by external Cl� removal. IndeedDIDS completely prevented the RVD, but its action could bedue to a simultaneous effect on both the Cl�/HCO3

� exchangerand swelling-activated Cl� conductance. The observation thatRVD was inhibited in cells preincubated with DIDS and notwith DNDS corroborates the central role of Cl�/HCO3

�. Infact, it has been demonstrated that DIDS but not its analogDNDS covalently binds the AE1 exchanger when incubated for60 min or more (8, 34, 38). Therefore, in the present study theCl�/HCO3

� exchanger remained blocked once DIDS was re-moved and the cells were unable to undergo RVD, althoughCl� conductance was still functional.

In proximal tubule cell lines from wild-type, task2 �/�, andcftr �/� mice, the strong increase in pHi induced by externalCl� removal in a HCO3

� medium could well be mediated bythe activation of the Cl�/HCO3

� exchanger. However, it isinteresting to note that in cftr �/� cells the hypotonic shock

Fig. 9. Role of pHe in the activation of K� currents byhypotonicity. Whole-cell currents in PCT cell line from wild-type mice were recorded in the continuous presence of DIDS (1mM) in an isotonic solution (control pH 7.4, 290 mosmol/kgH2O), 4–5 min after extracellular perfusion of an isotonicsolution at pH 8.0, 4–5 min after perfusion of a hypotonicsolution at pH 8.0 (230 mosmol/kgH2O), and finally in thepresence of clofilium (10 �M). Shown are whole-cell record-ings and corresponding I-V relationships. Values are means �SE of 5 cells from 5 different monolayers.



was insufficient to decrease pHi while the Cl�/HCO3� ex-

changer remained functional. Moreover, these cells were un-able to undergo RVD upon the hypotonic shock due to a lossof the signaling cascade that controls swelling-activated Cl�

channels (3). Therefore, it is clear that an increase in Cl�

conductance induced by a hypotonic shock is required toactivate the Cl�/HCO3

� exchanger. This observation was con-firmed in wild-type and task2 �/� cells in which the applicationof the Cl� channel blocker NPPB prevented the pHi decreaseupon hypotonicity.

In the original study of Dellasega and Grantham (12), invitro experiments clearly indicated that a hypotonic bath influ-enced cell volume in nonperfused proximal tubules. In re-sponse to this shock, the proximal tubule underwent RVD. It isclear that the model and the experimental conditions were quitedifferent in this study. A 50% hypotonic shock was applied inthe basolateral compartment, and the extracellular volume wasone million times larger than the volume of the tubule (12).Under these conditions, it was difficult to measure a pHe

change. However, on the basis of our experiments it is possiblethat the pHe variation was restricted to the immediate vicinityof the cells. Such a local phenomenon could be sufficient toincrease the local pHe and activate TASK2. This latter hypoth-esis could partially explain the activation of TASK2 K�

conductance recorded in whole-cell clamp experiments duringa hypotonic shock. However, in vitro studies performed inintact proximal tubules (43) showed that intracellular HCO3

�

depletion did not alter RVD. Such an observation is at variancewith our present data since we attributed a crucial role toHCO3

� in the alkalinization of the extracellular compartment.This discrepancy could probably be explained by the hugedifference between the HCO3

� concentrations released by thecells compared with the external bath buffering capacity.Interestingly, the HCO3

� requirement in the RVD process hasalready been described in cells of proximal tubules of Necturus(24) and mouse (41), cells of the thin descending limb ofHenle’s loop in the rabbit (23), and in other cell types such assingle osteosarcoma UMR-106–01 cells (36) and the epithe-lial-derived human breast cancer cell line ZR-75–1 (28).

The relationship between CFTR and HCO3� secretion has

been well documented in secretory epithelia. Different mech-anisms have been identified such as a Cl�-dependent secretionassociated with apical membrane Cl�/HCO3

� exchange, or acAMP-induced secretion through the Cl�-permeable pore ofCFTR (40). In primary cultures of mouse proximal tubules, ithas been demonstrated that the application of forskolin did notstimulate any Cl� currents (3). This feature has also beenfound in the mouse proximal cell line used in the present study(data not shown). The absence of cAMP-stimulated Cl� chan-nels suggests that CFTR does not conduct HCO3

� directly.The data from the present study are summarized in Fig. 10.

In proximal tubule cells, a hypotonic shock induces the acti-vation of swelling-activated Cl� channels in the presence ofCFTR. The molecular nature of these channels is not yetknown, but CFTR controls them by modulating autocrineadenosine production (5). The resulting exit of Cl� woulddecrease cytosolic Cl� activity, providing the driving force forthe activation of the Cl�/HCO3

�exchanger. The efflux ofHCO3

� then alkalinizes pHe, which activates the TASK2 chan-nels, allowing for a further increment of the K� conductanceinduced by the hypotonicity. This sensitivity to hypotonic

shock has already been described in several studies whichdemonstrated that TASK2 could be a swelling-activated K�

channel (2, 17, 29).Finally, the RVD process is achieved by an exit of KCl

controlled indirectly by CFTR (3). It is clear that furtherexperiments will be necessary to confirm this hypothesis.Notably, the membrane localization of the different transport-ers and channels involved in the RVD is not fully established.Even if TASK2 channels are located mainly in the proximaltubule (2), their basolateral localization has been determined infunctional studies only (2, 42). Numerous lines of evidencesuggest that the Cl�/HCO3

� exchanger could contribute tobasal efflux-mediating HCO3

� reabsorption across the basolat-eral membrane (1). Concerning the swelling-activated Cl�

channel, the fact that it could be modulated by CFTR-depen-dent autocrine control suggested a common localization withCFTR. At present, CFTR has been detected in the apicalmembrane by immunofluorescence study but patch-clamp re-cordings localize its activity to the basolateral membrane of theproximal tubule (10).

In conclusion, the present study attributes a central role toHCO3

� in the control of RVD in proximal tubule cells. Theoriginality of the model proposed here is based on the assumptionthat TASK2-mediated K� permeability is stimulated by hypo-tonic shock and the concomitant increase in pHe is induced byHCO3

� efflux. This mechanism is physiologically relevant in theproximal tubule because the active apical absorption of ions andorganic solutes increases osmolytes inside the cytoplasm. Theproximal tubule exhibits high water permeability of both cellmembranes (apical and basolateral) due to the presence of aqua-porin-1 (27). Therefore, the solute entries are accompanied bywater fluxes in the same direction and RVD must occur tocontinuously maintain cell volume. The RVD in proximal tubulescells is the result of a KCl efflux via Cl� and K� channels.However, in the proximal tubule, the reabsorption of HCO3

� inexcess of water alkalinizes the blood in the peritubular capillaries,allowing the activation of TASK2 channels. By using an in vitrosystem, we differentiated 1) the exit of HCO3

� induced by cellswelling from 2) the exit of HCO3

� resulting from its transcellulartransport. The first is driven by a Cl�/HCO3

� exchanger, whereasthe second is driven by a Na�-3HCO3

� cotransporter (42). In theproximal tubule in vivo, both systems work together to ensure

Fig. 10. Working model of TASK2 and CFTR during RVD in proximal celllines. Putative mechanisms for cell volume regulation during a hypotonicshock are shown.



HCO3� reabsorption and cell volume regulation. TASK2 K�

channels are continuously activated by cell swelling and basolat-eral HCO3

� accumulation, allowing the cell to maintain its volumeand its membrane potential, which is depolarized by Cl� exit(through swelling-activated Cl� channels) and electrogenic exit ofNa� and HCO3

� (through the Na�-3HCO3� cotransporter).

ACKNOWLEDGMENTS

We thank Dr. K. Mitchell and Prof. Dr. W. Skarnes for generouslyproviding task2 �/� mice.

REFERENCES

1. Alpern RJ. Cell mechanisms of proximal tubule acidification. Physiol Rev70: 79–114, 1990.

2. Barriere H, Belfodil R, Rubera I, Tauc M, Lesage F, Poujeol C, GuyN, Barhanin J, Poujeol P. Role of TASK2 potassium channels regardingvolume regulation in primary cultures of mouse proximal tubules. J GenPhysiol 122: 177–190, 2003.

3. Barriere H, Belfodil R, Rubera I, Tauc M, Poujeol C, Bidet M, PoujeolP. CFTR null mutation altered cAMP-sensitive and swelling-activated Cl�

currents in primary cultures of mouse nephron. Am J Physiol RenalPhysiol 284: F796–F811, 2003.

4. Barriere H, Rubera I, Belfodil R, Tauc M, Tonnerieux N, Poujeol C,Barhanin J, Poujeol P. Swelling-activated chloride and potassium con-ductance in primary cultures of mouse proximal tubules. Implication ofKCNE1 protein. J Membr Biol 193: 153–170, 2003.

5. Belfodil R, Barriere H, Rubera I, Tauc M, Poujeol C, Bidet M, PoujeolP. CFTR-dependent and -independent swelling-activated K� currents inprimary cultures of mouse nephron. Am J Physiol Renal Physiol 284:F812–F828, 2003.

6. Bidet M, Merot J, Tauc M, Poujeol P. Na�-H� exchanger in proximalcells isolated from kidney. II. Short-term regulation by glucocorticoids.Am J Physiol Renal Fluid Electrolyte Physiol 253: F945–F951, 1987.

7. Bidet M, Tauc M, Koechlin N, Poujeol P. Video microscopy of intra-cellular pH in primary cultures of rabbit proximal and early distal tubules.Pflugers Arch 416: 270–280, 1990.

8. Cabantchik ZI, Greger R. Chemical probes for anion transporters of mam-malian cell membranes. Am J Physiol Cell Physiol 262: C803–C827, 1992.

9. Civan MM, Peterson-Yantorno K, Coca-Prados M, Yantorno RE.Regulatory volume decrease by cultured non-pigmented ciliary epithelialcells. Exp Eye Res 54: 181–191, 1992.

10. Crawford I, Maloney PC, Zeitlin PL, Guggino WB, Hyde SC, TurleyH, Gatter KC, Harris A, Higgins CF. Immunocytochemical localizationof the cystic fibrosis gene product CFTR. Proc Natl Acad Sci USA 88:9262–9266, 1991.

11. De Smet P, Simaels J, Van Driessche W. Regulatory volume decrease ina renal distal tubular cell line (A6). I. Role of K� and Cl. Pflugers Arch430: 936–944, 1995.

12. Dellasega M, Grantham JJ. Regulation of renal tubule cell volume inhypotonic media. Am J Physiol 224: 1288–1294, 1973.

13. Duprat F, Lesage F, Fink M, Reyes R, Heurteaux C, Lazdunski M.TASK, a human background K� channel to sense external pH variationsnear physiological pH. EMBO J 16: 5464–5471, 1997.

14. Gleeson D, Corasanti JG, Boyer JL. Effects of osmotic stresses onisolated rat hepatocytes. II. Modulation of intracellular pH. Am J PhysiolGastrointest Liver Physiol 258: G299–G307, 1990.

15. Grinstein S, Cohen S, Rothstein A. Cytoplasmic pH regulation in thymiclymphocytes by an amiloride-sensitive Na�/H� antiport. J Gen Physiol83: 341–369, 1984.

16. Hara C, Satoh H, Usui T, Kunimi M, Noiri E, Tsukamoto K, Tanigu-chi S, Uwatoko S, Goto A, Racusen LC, Inatomi J, Endou H, Fujita T,Seki G. Intracellular pH regulatory mechanism in a human renal proximalcell line (HKC-8): evidence for Na�/H� exchanger, Cl�/HCO3

� ex-changer and Na�-HCO3

� cotransporter. Pflugers Arch 440: 713–720,2000.

17. Hoffmann EK. Intracellular signalling involved in volume regulatorydecrease. Cell Physiol Biochem 10: 273–288, 2000.

18. Hougaard C, Jorgensen F, Hoffmann EK. Modulation of the volume-sensitive K� current in Ehrlich ascites tumour cells by pH. Pflugers Arch442: 622–633, 2001.

19. Kanli H, Norderhus E. Cell volume regulation in proximal renal tubulesfrom trout (Salmo trutta). J Exp Biol 201: 1405–1419, 1998.

20. Knoblauch C, Montrose MH, Murer H. Regulatory volume decrease bycultured renal cells. Am J Physiol Cell Physiol 256: C252–C259, 1989.

21. Lesage F, Guillemare E, Fink M, Duprat F, Lazdunski M, Romey G,Barhanin J. A pH-sensitive yeast outward rectifier K� channel with twopore domains and novel gating properties. J Biol Chem 271: 4183–4187,1996.

22. Lionetto MG, Giordano ME, De Nuccio F, Nicolardi G, HoffmannEK, Schettino T. Hypotonicity induced K� and anion conductive path-ways activation in eel intestinal epithelium. J Exp Biol 208: 749–760,2005.

23. Lopes AG, Amzel LM, Markakis D, Guggino WB. Cell volume regu-lation by the thin descending limb of Henle’s loop. Proc Natl Acad SciUSA 85: 2873–2877, 1988.

24. Lopes AG, Guggino WB. Volume regulation in the early proximal tubuleof the Necturus kidney. J Membr Biol 97: 117–125, 1987.

25. Millar ID, Hartley JA, Haigh C, Grace AA, White SJ, Kibble JD,Robson L. Volume regulation is defective in renal proximal tubule cellsisolated from KCNE1 knockout mice. Exp Physiol 89: 173–180, 2004.

26. Morton MJ, Abohamed A, Sivaprasadarao A, Hunter M. pH sensingin the two-pore domain K� channel, TASK2. Proc Natl Acad Sci USA102: 16102–16106, 2005.

27. Nejsum LN. The renal plumbing system: aquaporin water channels. CellMol Life Sci 62: 1692–1706, 2005.

28. Nicholl AJ, Killey J, Leonard MN, Garner C. The role of bicarbonatein regulatory volume decrease (RVD) in the epithelial-derived humanbreast cancer cell line ZR-75–1. Pflugers Arch 443: 875–881, 2002.

29. Niemeyer MI, Cid LP, Barros LF, Sepulveda FV. Modulation of thetwo-pore domain acid-sensitive K� channel TASK-2 (KCNK5) bychanges in cell volume. J Biol Chem 276: 43166–43174, 2001.

30. Niemeyer MI, Cid LP, Sepulveda FV. K� conductance activated duringregulatory volume decrease. The channels in Ehrlich cells and theirpossible molecular counterpart. Comp Biochem Physiol A Mol IntegrPhysiol 130: 565–575, 2001.

31. Raat NJ, De Smet P, van Driessche W, Bindels RJ, Van Os CH.Measuring volume perturbation of proximal tubular cells in primaryculture with three different techniques. Am J Physiol Cell Physiol 271:C235–C241, 1996.

32. Reyes R, Duprat F, Lesage F, Fink M, Salinas M, Farman N, Laz-dunski M. Cloning and expression of a novel pH-sensitive two poredomain K� channel from human kidney. J Biol Chem 273: 30863–30869,1998.

33. Roos A, Boron WF. Intracellular pH. Physiol Rev 61: 296–434, 1981.34. Schopfer LM, Salhany JM. Characterization of the stilbenedisulfonate

binding site on band 3. Biochemistry 34: 8320–8329, 1995.35. Snouwaert JN, Brigman KK, Latour AM, Malouf NN, Boucher RC,

Smithies O, Koller BH. An animal model for cystic fibrosis made by genetargeting. Science 257: 1083–1088, 1992.

36. Star RA, Zhang BX, Loessberg PA, Muallem S. Regulatory volumedecrease in the presence of HCO3

� by single osteosarcoma cells UMR-106–01. J Biol Chem 267: 17665–17669, 1992.

37. Tauc M, Le Maout S, Poujeol P. Fluorescent video-microscopy study ofregulatory volume decrease in primary culture of rabbit proximal convo-luted tubule. Biochim Biophys Acta 1052: 278–284, 1990.

38. Taylor AM, Grobner G, Williamson PT, Watts A. Binding properties ofthe stilbene disulfonate sites on human erythrocyte AE1: kinetic, thermo-dynamic, and solid state deuterium NMR analyses. Biochemistry 38:11172–11179, 1999.

39. Teleky B, Hamilton G, Cosentini E, Bischof G, Riegler M, Koperna T,Feil W, Schiessel R, Wenzl E. Intracellular pH regulation of humancolonic crypt cells. Pflugers Arch 426: 267–275, 1994.

40. Vidyasagar S, Rajendran VM, Binder HJ. Three distinct mechanisms ofHCO3

� secretion in rat distal colon. Am J Physiol Cell Physiol 287:C612–C621, 2004.

41. Volkl H, Lang F. Ionic requirement for regulatory cell volume decreasein renal straight proximal tubules. Pflugers Arch 412: 1–6, 1988.

42. Warth R, Barriere H, Meneton P, Bloch M, Thomas J, Tauc M,Heitzmann D, Romeo E, Verrey F, Mengual R, Guy N, Bendahhou S,Lesage F, Poujeol P, Barhanin J. Proximal renal tubular acidosis inTASK2 K� channel-deficient mice reveals a mechanism for stabilizingbicarbonate transport. Proc Natl Acad Sci USA 101: 8215–8220, 2004.

43. Welling PA, Linshaw MA. Importance of anion in hypotonic volumeregulation of rabbit proximal straight tubule. Am J Physiol Renal FluidElectrolyte Physiol 255: F853–F860, 1988.



Extracellular pH alkalinization by Cl-/HCO3-exchanger is crucial for TASK2 activation by hypotonic...

Documents

Transcript of Extracellular pH alkalinization by Cl-/HCO3-exchanger is crucial for TASK2 activation by hypotonic...