chloride dependence of ph modulation by beta-adrenergic

M Désilets, M Pucéat and G VassortChloride dependence of pH modulation by beta-adrenergic agonist in rat cardiomyocytes.

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 1994 American Heart Association, Inc. All rights reserved.is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation Research

doi: 10.1161/01.RES.75.5.8621994;75:862-869Circ Res.

http://circres.ahajournals.org/content/75/5/862World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://circres.ahajournals.org//subscriptions/

is online at: Circulation Research Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. Further informationEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculation Researchin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:

by guest on March 2, 2014http://circres.ahajournals.org/Downloaded from by guest on March 2, 2014http://circres.ahajournals.org/Downloaded from

http://circres.ahajournals.org/content/75/5/862

http://circres.ahajournals.org/content/75/5/862

http://www.ahajournals.org/site/rights/

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://www.lww.com/reprints



http://circres.ahajournals.org/




862

Chloride Dependence of pH Modulation by-Adrenergic Agonist in Rat Cardiomyocytes

Michel Desilets, Michel Puceat, Guy Vassort

Abstract The effects of /3-adrenergic agonists on pHi werestudied on single ventricular myocytes isolated from adult ratheart and loaded with the acetoxymethyl ester (AM) form ofthe pH indicator SNARF-1. In modified Krebs' solutioncontaining 20 mmol/L HEPES and 4.4 mmol/L HCO3-, iso-proterenol (1 /lLmol/L) caused a significant decrease of steady-state pHi from 7.20±0.02 to 7.13±0.02 (mean±SEM) within 2minutes. This acidification, which was also observed in myo-cytes that were preloaded with the Ca2 chelator BAPTA andsuperfused with nominally Ca2'-free solution, was blocked bypropranolol as well as by the specific 8,-antagonist CGP 20712A but not by the $3-antagonist ICI 118,551. Forskolin (10,mol/L) induced a similar reversible decrease of pH, (averagedecrease, 0.11±0.02 pH unit). Furthermore, adenosine (100,umol/L) substantially attenuated the isoproterenol-induceddecrease of pHi. The effect of isoproterenol was not preventedby inhibitors of the Na+-HW antiport, amiloride (1 mmol/L)and 2-N,N-hexamethylene amiloride (20 tmol/L). On theother hand, blockers of Cl- transport mechanisms, DIDS (200,umol/L) and probenecid (100 ,mol/L), inhibited this acidifi-cation. Isoproterenol also failed to induce a decrease of

steady-state pH1 in myocytes incubated in Cl--free medium.Rather, the initial rate of rise of pHi observed on removal ofexternal Cl- ions was significantly increased in the presence ofisoproterenol or dibutyryl cAMP. Because the alkalinizationinduced by removal of Cl- ions is mainly due to reversal of theCl--HCO3- exchanger, the augmentation of this initial rate ofpH, rise directly points to a /-adrenergic stimulation of theexchanger. Furthermore, the pHi recovery following NH4Clexposure was accelerated by isoproterenol in the presence ofprobenecid, indicating that the Na+-HCO3, cotransportand/or the Na+-H+ antiport also could be activated. In con-clusion, the present results demonstrate that ,B-adrenergicagonist-induced acidification of rat ventricular myocytes oc-curs mainly through alteration of Cl- transport systems, mostlikely via a cAMP-dependent stimulation of the ClW-HCO3-exchanger. Since the alkalinizing mechanisms are also stimu-lated, an increased apparent cellular buffering capacity isexpected. (Circ Res. 1994;75:862-869.)Key Words * acidosis * fluorescence * Cl--HCO3

exchanger * Na+-H' antiport * cAMP

It is now well recognized that pHi in mammaliancardiac myocytes is controlled by various indepen-dent ion transporters, namely the Na+-H+ an-

tiport,'-4 the Na'-HCO3 symport,5'6 and the Cl-HCO3- exchanger.7'8 One should also mention thepossible presence of the Na+-dependent Cl--HCO3-exchanger, as observed in cultured chick heart cells.9Under normal conditions, the three Na+-dependent iontransporters would be responsible for the maintenanceof a pHi level above that expected from passive distri-bution of H' ions, whereas the Na+-independent C1--HCO3- exchanger acts as an acidifying mechanismthrough extrusion of HCO3- ions. Steady-state pHi istherefore the result of a balance between the activity ofthese transporters, transmembrane acid and base leaks,and the metabolic production of acids.

It is also well known that important neurohormonalregulation of pHi can occur through modulation of theseexchangers. For instance, a1-adrenergic receptor stimu-lation in cardiac myocytes engenders intracellular alka-linization by activation of the Na+-H+ exchanger.10-12Similarly, previous work from our laboratory13 demon-strated that the Cl--HCO3- exchanger is strongly stim-

Received February 28, 1994; accepted July 15, 1994.From INSERM U-390, Physiopathologie Cardiovasculaire,

CHU Arnaud de Villeneuve, Montpellier, France.Presented in part in abstract form (J Physiol Lond. 1993;

459:223P).Correspondence to M. Desilets, Department of Physiology,

Faculty of Medicine, University of Ottawa, 451 Smyth Road,Ottawa, Ontario, Canada K1H 8M5.

ulated by extracellular MgATP, thereby leading totransient acidosis. There is also increasing evidence that,3-catecholamines can modulate pHi regulation in car-diac ventricular myocytes. Early work on perused ratheart14'15 demonstrated that norepinephrine as well asdibutyryl cAMP and glucagon had little effect on basalpHi but significantly attenuated the pHi fall when per-fusing the paced hearts with hypercapnic solutions. Inthese experiments, pH, was estimated from measure-ments of [14C]dimethyloxazolidinedione distributionand therefore subjected to the inherent limitations ofthe method. Nevertheless, the results were remarkablein view of the expected cAMP-induced enhancement ofcontractile and metabolic activities that should lead, ifanything, to a decrease in pHi. In fact, it was recentlydemonstrated that in sheep Purkinje fibers superfusedwith HCO3--free solutions, isoproterenol (ISO) inducedan intracellular acidification that was nearly abolishedafter the inhibition of glycolysis and partly reduced inthe presence of Na+-H+ exchange blockers.16 The au-thors therefore concluded that ISO induced metabolicacidosis together with a partial inhibition of the Na+-H+exchange. On the other hand, such an effect of en-hanced metabolic activity could not be detected inguinea pig isolated ventricular myocytes, although asmall inhibition of the Na+-H+ antiport was also ob-served.12 Furthermore, this latter study demonstratedthat ,B-adrenergic agonists also caused an activation ofthe Na+-HCO3- cotransport, thereby leading to anopposite effect on acid extrusion.

by guest on March 2, 2014http://circres.ahajournals.org/Downloaded from



Desilets et al i-Adrenergic Regulation of pH; 863

The Cl--HCO3- exchanger is another potentially im-portant target for pHi regulation by ,B-adrenergic recep-tors. In other preparations, cAMP has been shown tomodulate the anion exchanger, although the effectsappear to vary extensively between tissues and species.Thus, the exchanger may be either stimulated,17"18 in-hibited,19'20 or unaffected2' by intracellular cAMP. How-ever, whether such modulatory effects occur in heartremains to be determined. The present study, which hasbeen designed to assess the effects of ISO on pHi ofventricular myocytes isolated from rat heart, does dem-onstrate an activation of the Cl-HCO3- exchanger. Themeasurements of pHi were carried out in single cellsloaded with the fluorescent indicator SNARF-1. Theyfurther show that this ISO-induced acidification occursthrough 3,1-adrenergic receptors, is mimicked by forsko-lin, and is substantially attenuated by adenosine.

Materials and MethodsCell Isolation and SNARF-1/AM Loading

Cardiac ventricular myocytes were isolated from maleWistar rats from collagenase perfusions, as previously de-scribed.22 After isolation, myocytes were first allowed to attachto small glass coverslips, previously coated with 10 gg laminin,by incubation for 1 hour at 37°C in HEPES-buffered solutionsupplemented with 0.5% bovine serum albumin (BSA). Theloading of the acetoxymethyl ester (AM) form of the pH-sensitive fluorescent dye SNARF-1 consisted of a 20- to30-minute incubation of the myocytes at room temperature inthe BSA solution containing 5 ,umol/L SNARF-i/AM previ-ously dissolved in dimethyl sulfoxide (DMSO) and 0.02%pluronic acid. The dye was then removed by replacing theincubation solution with fresh BSA solution. The cells werethen incubated for at least another 30-minute period at roomtemperature to ensure complete hydrolysis of the ester form ofthe dye in the cells. In some experiments, the cells were alsoloaded with the Ca2' chelator BAPTA to buffer intracellularCa'2 ions. In this case, loading was carried out immediatelyafter SNARF-i/AM washout in a manner identical to that forSNARF-1, with the loading solution containing 25 gmol/LBAPTA/AM.

pH Measurements in Single CellspHi was monitored as described previously.13 Briefly, a

coverslip containing the myocytes was transferred in a smallbath placed on the stage of an inverted epifluorescent micro-scope. Continuous superfusion was performed by positioningclose to the myocyte the end of a small capillary, whichreceived the output of one of seven converging tubings, eachconnected to 60-mL reservoirs. This allowed fast changes(within 5 seconds) of superfusion solution with minimal deadspace. The temperature of the solutions and microscope setupwas maintained at 33°C for all the experiments.

Fluorescence measurements were carried out by illuminat-ing a single cell with the light of a 100-W xenon lamp filteredat 514 nm and focused through a x40 immersion objective.Emitted light was separated at 605 nm by a dichroic mirror,and the two beams were filtered at 580 and 640 nm beforedetection by their respective photomultipliers. Analog signalsfrom the photomultipliers were filtered at 10 Hz beforeacquisition by a tape recorder and a computer, which allowedthe on-line calculation of the ratio of the two emitted wave-lengths (640/580). Results presented in the present study areall chart recordings of this ratio. Conversion of the 640/580ratio into pHi values was done from calibration curves estab-lished regularly from SNARE-1-loaded myocytes exposed to

ElectrophysiologySimultaneous recordings of transmembrane currents in

SNARF-1-loaded myocytes were performed by the discon-tinuous voltage-clamp technique, whereby the recordingelectrode was switched at frequency of %2 to 3 KHz from avoltage-recording to a current-passing mode, as describedelsewhere.23 The electrodes were small suction pipettes withresistances of 40 to 50 Mfl and filled with the followingsolution (mmol/L): KCI 12, potassium aspartate 130, MgCl24, EGTA 0.5, and HEPES 5, pH adjusted to 7.2 with KOH.These small electrodes were used to minimize perturbationof the intracellular medium, and stable pHi values as well asCa21 currents could be recorded for at least 20 minutes afteraccess to the cell interior. Current and voltage tracings werelow-pass-filtered at 300 Hz before their display on a chartrecorder.

SolutionsThe control superfusion solution contained (mmol/L) NaCl

120, KCI 5.7, NaHCO3 4.4, NaH2PO4 1.2, CaCl2 1.8, MgCl2 1.7,HEPES 20, and glucose 10, pH adjusted to 7.4 at 33°C withNaOH. The solution was buffered with 20 mmol/L HEPES toensure proper buffering capacity and hence to avoid possibleextracellular accumulation or depletion of acids. Bicarbonatewas added at a concentration of 4.4 mmol/L to permit normalfunctioning of the Cl--HCO3- exchanger, whose half-maximalactivation occurs at -1 mmol/L HCO3 .7 CaCl2 was omittedfor the experiments performed with nominally Ca 2k-free solu-tion. In the experiments involving Cl--free solutions, HCO3-concentration was raised from 4.4 to 7.5 mmol/L to ensurecomplete activation of the ClW-HCO3- exchanger. For thoseexperiments, the control Cl--containing solution had the fol-lowing composition (mmol/L): NaCl 120, KHCO3 5.7,NaH2PO4 1.2, CaCO3 1.8, MgSO4 1.7, HEPES 20, and glucose10, pH 7.4. In the corresponding Cl -free solution, NaCl wasreplaced with 120 mmol/L sodium gluconate. (±)ISO HCI,DL-propranolol HCl, adenosine, and dibutyryl cAMP wereadded from stock solutions. Solutions of CGP 20712 A and ICI118,551 were made the day of the experiments. Forskolin wasprepared as 10-mmol/L stock in ethanol. Amiloride, 2-NN-hexamethylene amiloride (HMA), and DIDS were prepared as100-mmol/L stock solutions in DMSO. Probenecid was dis-solved in 0.5N NaOH at a concentration of 100 mmol/L. Allthese chemicals were purchased from Sigma Chemical Co,with the exception of HMA (Research Biochemical Inc). CGP20712 A and ICI 118,551 were gifts from Dr A. Mugelli.SNARF-i/AM and BAPTA/AM were purchased from Molec-ular Probes.

ResultsFig 1 demonstrates the effects of 1 gmol/L ISO on

pHi in a SNARF-1-loaded ventricular myocyte whoseCa2' current was simultaneously recorded. Exposure tothe f8-agonist caused the expected rapid and largeincrease of this slow inward current, together with asmall inward shift of the steady-state current measuredat the end of the 200-millisecond pulse. In addition, ISOinduced a decrease of pHi. However, the time course ofthis acidification was much slower than that of theaugmentation of the transient inward current. Botheffects were reversible on washout, with similar timecourses of their development. Unstimulated cells exhib-ited similar acidification when exposed to ISO, and thissimpler protocol, which does not require the measure-ment of transmembrane currents, was used for theremaining part of the present study. On average, expo-sure to 1 gmol/L ISO induced a sustained decrease ofpH, of 0.07+0.01 pH unit (n=19). Most of the experi-ments were performed with this concentration of thenigericin and standard buffers.13




864 Circulation Research Vol 75, No 5 November 1994

A7.25 Iso

PH 17.0J

15 S

nA o]-1

nA

200 msFIG 1. Effect of isoproterenol (ISO) on pH, and Ca2+ current.Tracings are simultaneous chart recordings of pH, (upper trac-ing) and transmembrane currents (middle tracing) in a SNARF-1-loaded single cell. Membrane potential was held at -40 mV,and Ca2+ currents were elicited every second with 200-millisec-ond depolarizing pulses to 0 mV. Addition of 1 ,umol/L ISOcaused an acidification of the myocyte together with an increaseof the transient inward currents. Both effects were reversible, asdemonstrated by the right portion of the two tracings, which weretaken 5 minutes after washout of ISO. The lower tracingsrepresent expanded current recordings taken before (left) and 2minutes after the addition of ISO (middle) and 6 minutes afterwashout of the 8-agonist (right).

P-agonist. Nevertheless, preliminary work demon-strated that although exposure to 10 nmolIL ISO failedto induce a detectable change of pHi, a concentration of100 nmol/L was sufficient to produce a near-maximaleffect, the average diminution being 0.06±+0.01 pH unit(n=4). These experiments on quiescent cells furtherindicated that an acidification can occur in the absenceof an ISO-induced increase in cytosolic free-Ca2' con-

centration. In fact, measurements from four individualmyocytes preloaded with the Ca2' chelator BAPTA andmaintained in nominally Ca2+-free solution also exhib-ited a similar decrease of pHi when exposed to 1 gmol/LISO, the average diminution being equal to 0.09+±0.03pH unit.

Effects of P-Antagonists, Forskolin, and Adenosineon ISO-Induced Changes of pHiThe specificity of the acidifying effect of ISO was

tested by the use of propranolol and by the /31-antago-nist CGP 20712 A24 and the A2-antagonist ICI 118,551.25As shown in Fig 2, preexposure of the myocytes to either10 ,umol/L propranolol (panel A) or 0.1 ,umol/L CGP20712 A (panel B) entirely prevented the ISO-inducedacidification. On the other hand, this acidification wasnot significantly altered by 0.1 ,umol/L ICI 118,551 (datanot shown), since the average ISO-induced decrease ofpHi was equal to 0.06±+0.01 under those conditions(n=6). The inhibition by propranolol and CGP 20712 Awas completely reversible, such that the pHi decreaseinduced by ISO after washout of these antagonists wasnot significantly different from that measured undercontrol conditions (ISO-induced decrease of 0.06±+0.1and 0.07±0.01 pH unit after washout of propranololand CGP 20712 A, respectively, with five independentmeasurements in each case). It should also be men-tioned that the experiments involving CGP 20712 A andICI 118,551 were carried out in the presence of 20imnol/L HMA to further block the Na+-H+ antiport. As

discussed below, the acidification caused by ISO underthose conditions allows one to rule out the contributionof the Na+-H+ exchanger in this process.

lSo725]

pH]7.15

Propranolol180

2 min

B

7.1 CGP

PH.9 ISO ISO

1 min

FIG 2. Tracings showing inhibition by propranolol and CGP20712 A (CGP) of isoproterenol (ISO)-induced decrease of pH1.A, The myocyte was first exposed to 1 gmol/L ISO, which causeda decrease of steady-state pH1 from 7.25 to 7.17. About 5minutes after washout of ISO, 10 gmol/L propranolol was addedto the solution, followed by a second exposure to ISO. Althoughnot shown, readdition of ISO 5 minutes after removal of propran-olol induced a decrease of pH1 of 0.06 pH unit. B, The myocytewas first exposed to 0.1 gmol/L CGP, which prevented theISO-induced change in pHi. Reexposure to ISO in the absenceof CGP caused a decrease of pH from 6.98 to 6.90. In thisexperiment, 2-N,N-hexamethylene amiloride (20 gtmol/L) waspresent throughout. Measurements are from two different cells.Note the different time and pH scales between the two panels.The relatively large noise in the tracing of panel B was due to theabsence of an analog filter normally used before data analysis.

Fig 3A demonstrates that exposure of a myocyte to 10,umol/L forskolin induces a pHi diminution similar tothat observed with ISO. On average, the alkaloid causeda reversible decrease of pH, of 0.11±0.02 (n=9). Fur-thermore, adenosine substantially attenuated this re-sponse to ISO, as illustrated in Fig 3B. Thus, addition ofISO in the presence of 100 pmol/L adenosine (and 20,Lmol/L HMA) produced only a small acidification that,however, could be increased on washout of adenosine.Conversely, although the addition of adenosine alonedid not affect significantly steady-state pH1, it caused animportant and reversible alkalinization during ISO ex-posure. This inhibitory effect of adenosine was observedin all eight cells studied. On average, the ISO-induceddecrease of pHi was 0.03+±0.01 in the presence ofadenosine, a value significantly smaller (P<.05) than the

A72

PH7.0

A

Ade

For8kolin

1 min

Ade

p7H -S \

1 m;nFIG 3. Tracings showing the effect of forskolin and adenosine(Ade) on isoproterenol (ISO)-induced decrease of pH,. A, Themyocyte was exposed to 10 ,umol/L forskolin during the timeindicated by the horizontal bar. B, The myocyte was first ex-posed to 100 ,umollL Ade, which did not cause a detectablechange of pH1. Addition of 1 ,umol/L ISO caused a decrease ofpHj from 7.26 to 7.20. Removal of adenosine in the presence ofISO induced an additional pH decrease to 7.02, which wasreversed on readdition of Ade. Tracings are from two differentcells.




De'silets et al f-Adrenergic Regulation of pH; 865

A HMA7.4 O ----------------------

pH7.2- 1 min

B7.4

pH]

7.2J

ProbenecidISO lSO

1 min

FIG 4. Probenecid, but not 2-N,N-hexamethylene amiloride(HMA), inhibits the isoproterenol (ISO)-induced decrease ofsteady-state pHi. A, Tracing shows the acidifying effect of 1gmol/L ISO in the presence of 20 ,umol/L HMA. Note the smallHMA-induced decrease of pH1. B, Tracings show pH, in a singlecell preexposed for 10 minutes to 100 ,umol/L probenecid (lefttracing) and 5 minutes after its removal (right tracing). Addition of1 gmol/L ISO during exposure to probenecid had very little effecton pH, compared with that measured after washout of the drug.As for HMA, probenecid induced a decrease of pHi that wasreadily noticeable from the difference between the steady-statelevels of the two tracings.

decrease of 0.10±0.02 pH unit observed after washoutof the purine. Moreover, when adenosine was addedwhile washing ISO, the recovery of cell pH, toward itsresting value was significantly accelerated (data notshown). Overall, these results directly point to theexpected involvement of cAMP in the acidifying effectof ISO. On the other hand, we failed to observe a

significant change of pH, in myocytes exposed to dibu-tyryl cAMP. Nevertheless, and as presented below, itdid influence the Cl--dependent changes of pH, in amanner similar to that observed with ISO.

Effect of Inhibitors of Na+-H+ Exchange and Cl-Transport Systems on ISO-Induced Changes of pHiAs illustrated in Figs 2 and 3, the acidifying effect of

ISO was not prevented by inclusion of HMA, an

amiloride derivative with substantially enhanced po-tency to inhibit the Na+-H+ antiport.26 This is demon-strated again in Fig 4A. On the other hand, preexposureto probenecid, an inhibitor of the Na+-independentCW-HCO3- antiporter,27 virtually abolished the effect ofISO, as illustrated in Fig 4B. The average ISO-inducedchanges of pH1 in the presence of these two blockers andalso in the presence of the less specific blockers

7.3 ".We

AA

7.10 -8C

pH

7A NH4CI

72- JHMAObnelPH 1180

7D- C

6.- NH 1R

cano a dse That bthe es aendprobenecrcid of t

Hr m a amilo.Th tri e, pr nev s rpeentedtheacidif caton

changes ofpHd during and after exposure to 20 mmol/L NH4Cs. Ademonstrates the effect of ISO when added during pHarecoveryafter washout of Na4C1; B and C, the effects of ISO when appliedafter the addition of 20 simol/L 2-NN-hexamethylene amilorideor 100 ,mmol/L probenecid, respectively. Tracings are fromdifferent cells.

amiloride and DIDS are compiled in the Table. Thus, itcan be seen that both DIDS and probenecid, but notHMA and amiloride, prevented the acidification in-duced by the ,B-agonist. Similarly, forskolin failed tocause a decrease of pHi in two myocytes preexposed to200 ,umol/L DIDS (data not shown). It can also benoted from the Table and Fig 4 that the blockers causedthemselves a small but significant decrease of pHi.To further test whether ISO could affect the Na+-H+

antiport and to confirm that HMA did cause an inhibi-tion of the Na+-H+ exchange, standard protocols involv-ing NH4C1 exposures were carried out. The acidosisafter NH4C1 washout should activate the Na+-H+ ex-

changer, and the time course of recovery should partlyreflect the activity of this acid extrusion system.28 Anexample of such an experiment is shown in Fig 5, which

Summary of Effect of 1 /imol/L Isoproterenol on pH, Under Various Conditions

Conditions Initial pH, After Treatment ISO-Induced ApHi n m

Control 7.20±0.02 ... -0.07+0.01* 1 9 11

Amiloride, 1 mmol/L 7.19±0.01 7.15±0.01 -0.05±0.01* 8 2

HMA, 20 1moi/L 7.28±0.01 7.22±0.01 -0.07±0.02* 10 3

DIDS, 200 ,umol/L 7.09±0.01 7.02±0.01 -0.015±0.004 6 2

Probenecid, 100 gmol/L 7.33±0.03 7.21 ±0.03 -0.010±0.003 9 3

0 [Cl-] 7.21 ±0.05 7.49±0.02 +0.010±0.005 4 3

ISO indicates isoproterenol; HMA, 2-N,N-hexamethylene amiloride; n, number of cells; and m, number of tissues.Values are mean±SEM.

Initial pHi values were measured immediately before treatment with the various blockers or Cl--free solution (0[Cl-]). ISO-induced ApHi was determined as the difference between steady-state pH, measured during and beforeISO exposure.*Changes of pHi significantly different from 0 (P<.01).





Cl-free

O i0

180 180 ISO

A725]

pH725.

3 min

FIG 6. Tracing showing the effect of isoproterenol (ISO) and Cl-removal on pH;. The myocyte was first exposed to 1 ,umol/L ISO,which caused the usual diminution of pH1. Replacement ofextracellular Cl- with gluconate induced a slow but large in-crease of pHi. The acidifying effect of ISO was not observedunder these conditions. On the other hand, the recovery fromalkalosis after readdition of Cl- ions was accelerated by theP-agonist.

compares the effects of ISO, HMA, and probenecid on

pHi recovery after washout of NH4Cl applied at 20mmol/L during 90 to 120 seconds. It can be seen fromthe upper panel that addition of ISO during the pH,recovery appeared to momentarily reverse this recov-ery. This could be interpreted as an inhibition of theNa+-H+ antiport. However, addition of HMA, whichalso almost completely inhibited the pHi recovery, didnot prevent this acidification, thereby indicating thatISO and HMA act on different mechanisms. Fig 5Cillustrates the effect of probenecid under those condi-tions. The drug also induced a transient decrease of pHiand reduced the slope of its return toward steady-statelevel. Contrary to the effect of HMA, however, itinhibited the acidifying effect of ISO. In fact, it can benoticed from the recording that the ,-agonist induced a

small but clear increase of pHi under these conditions.Such an increase has been observed in four of the sixindependent recordings. To test the involvement of theNa+-K-2Cl- cotransport in ISO-induced acidification,two cells were superfused with 10 ,umol/L bumetanidebefore ISO exposure. In both cells, ISO still triggered a

decrease of pH, (-0.08±0.02 pH unit) at a magnitudesimilar to that found in the absence of the drug.

Cl- Dependence of ISO-Induced Decrease of pHiThese observations suggest that alterations of Cl--

dependent transports, but not of the Na+-HW antiport,underlie the ISO-induced increase of proton concentra-tion. This conclusion is supported by experiments in-volving removal of extracellular Cl-. As depicted in Fig6, equimolar replacement of extracellular Cl- with thelarge anion gluconate induced a slow rise of pH, thatreached near steady state within 10 to 15 minutes.Preliminary experiments showed that the absence ofHCO3- in the external medium or 200 ,umol/L DIDSfully blocks such an alkalinization. Under Cl--free con-ditions, an ISO-induced decrease of pH, was not ob-served and yet could be reproduced during the recoveryfollowing readdition of extracellular Cl-. The failure ofISO to decrease steady-state pH1 reached after incuba-tion in Cl--free medium is also indicated in the Table.On the other hand, application of the 3-agonist duringthe initial increase of pHi, ie, shortly after Cl- removal,enhanced the rate of pH1 rise. This is better demon-strated by the experiment illustrated in Fig 7, whichreports the effects of brief exposures to the Cl--freesolution in the presence or absence of ISO. This record-ing clearly demonstrates that the increase of pH, was

BdbcAMP 2 min

72- a b c

pH

7.0 _

FIG 7. Effect of isoproterenol (ISO) and dibutyryl cAMP (dbcAMP)on the initial rate of pH, rise induced by removal of extracellular ClIions. The upper tracing depicts the changes of pH; during briefexposures of the myocyte to Cl--free solution (indicated by thesmall horizontal bars) occurring before, during, and 5 minutes afterwashout of 1 gmol/L ISO. A similar protocol was used to determinethe influence of 100 ,umol/L dbcAMP (lower panel), except thatexposures to Cl--free solutions were repeated at longer intervals toallow for the slow action of the cAMP analogue. Thus, tracings athrough c were sequentially obtained before and 5 minutes afterthe addition of dbcAMP and 10 minutes after its washout, respec-tively. The amplitude of the pH, increases; hence, the rate of risewas enhanced by both ISO and dbcAMP.

reversibly accelerated in the presence of the 8-agonist.Fig 7 also shows the effect of dibutyryl cAMP underthose conditions. As mentioned above, it failed to mimicthe ISO-induced decrease of steady-state pHi. On theother hand, it can be seen that prolonged exposure todibutyryl cAMP did cause an augmentation of the rateof pHi rise initiated by removal of extracellular Cl-. Thestimulatory effect of ISO and dibutyryl cAMP is alsopresented in Fig 8, which illustrates in a graph theaverage values of pHi changes plotted as a function oftime after removal of extracellular Cl-. Both ISO(P<.01) and dibutyryl cAMP (P<.03) significantly en-hanced the rate of change of pHi, as determined fromtwo-way ANOVA, with five different cells for eachtreatment. In fact, this stimulatory effect was observedin all cells studied. Furthermore, both treatments in-duced an ;=65% increase in the rate of pHi rise afterremoval of external Cl-, as measured 90 seconds afterCl- washout. Thus, the change of pH measured at thattime increased from 0.050±0.009 to 0.083±0.013 withISO and from 0.047±0.014 to 0.077±0.021 for cellspreincubated at least 5 minutes in 100 ,mol/L dibutyrylcAMP.

DiscussionAcidifying Effect of ISO Occurs Through Increase ofcAMP After Activation of Pi-Adrenergic ReceptorsThe present results demonstrate that stimulation of

f-adrenergic receptors in rat cardiomyocytes induces anintracellular acidification that results from an activationof the Na+-independent ClW-HCO3- exchanger. The factthat this decrease of pHi was completely prevented bythe highly specific 81-antagonist CGP 20712 A but notby the ,82-antagonist ICI 118,551 further supports thenotion that ISO-induced acidification is entirely medi-ated by the 831-adrenergic receptor subtype.29The ISO-induced acidification is likely to occur

through an increase of intracellular cAMP content, asdemonstrated by three different approaches. First, theamplitude and time course of this acidification could bemimicked with the use of forskolin, a direct activator ofadenylyl cyclase.30 Second, the decrease of pHi could be

7.6pH

7.4-

7.2.

lSO




Desilets et al P-Adrenergic Regulation of pH; 867

0.10-

0.05

Z

0

0.10

0.05

m

m-

A

Time (8)

Time (a)

FIG 8. Graphs comparing effects of isoproterenol and dibutyrylcAMP on the average increases of pH1 induced by removal ofextracellular Cl-. Time zero was taken as the time of removal ofCl- ions, and changes of pHi (,ApH) were calculated from thedifference between the pHi value at a given time and thatmeasured before Cl- removal. The increase of pHi was signifi-cantly augmented (P<.05) in the presence of both isoproterenolat 1 ,umol/L (upper graph) and dibutyryl cAMP at 100 gmol/L(lower graph). For both graphs, data points represent theaverage values obtained from five individual myocytes, withvertical bars indicating 1 SEM. * indicates control condition;stimulation with isoproterenol or dibutyryl cAMP.

substantially attenuated by adenosine. It is indeed wellrecognized that one of the major effects of adenosine onventricular cells is its inhibition of adenylyl cyclaseactivity and the consequent decrease of cellular cAMPlevel in preparations exposed to ,B-adrenergic agonists.31Our observation that adenosine did not affect basal pH,under control conditions, as also reported previously,'3but caused an increase of pH, after exposure to ISO isconsistent with this mechanism. This is analogous to thenegative inotropic action of adenosine that occurs onlyon cardiac muscles that have been previously exposed toagents that induced an increase of intracellular cAMPcontent.3' Finally, the involvement of cAMP was alsodemonstrated by the use of dibutyryl cAMP, whichstimulated the Cl -HCO3- exchanger by working in thereverse mode after removal of external Cl- (Figs 7 and8). The fact that we failed to detect a significant effect ofthis permeant cAMP analogue on steady-state pHi can

be explained by its slow diffusion and, as discussedbelow, points to the involvement of secondary mecha-nisms with opposite effects that could also be modulatedby intracellular cAMP.

ISO-Induced Acidification Is Not ofIntracellular OriginThe observation that Cl- transport blockers as well as

Cl- washout completely inhibited the ISO-induced acid-ification strongly suggests that this effect occurs throughan alteration of sarcolemmal Cl- transport mechanisms.As such, possible sources of intracellular acidificationincluding adrenergic stimulation of metabolic activity or

induced changes of intracellular buffering capacity ap-pear unlikely to represent major acidifying mechanismsunder our experimental conditions. The former possi-bility has been recently suggested by Guo et al,16 whoshowed that ISO-induced acidosis in quiescent sheepPurkinje fibers was almost completely prevented byinhibition of glycolysis. Catecholamines are indeed wellknown to stimulate cardiac glycogenolysis32 and glycol-ysis,33 although the consequent effect on pHi remains tobe established. As reviewed by Dennis et al,34 themetabolism of glucose or glycogen should not lead to alarge increase of intracellular proton concentrationunder normal aerobic conditions. Furthermore, meta-bolic acidosis is likely to be minimal for quiescent cellsin which the energy consumption normally associatedwith adrenergic-induced positive inotropy should begreatly reduced. In any case, conciliation of our resultswith the concept of significant enhancement of meta-bolic acid production would require one to invoke either(1) the fact that inhibition of transsarcolemmal Cl-movement would somehow interfere with the adrener-gic modulation of metabolic activity or (2) the presenceof some Cl--independent mechanisms that would ex-actly counterbalance, and hence mask, this increase ofmetabolic acid production. Similarly, the Cl- depen-dence of the ISO effects indicates that changes ofbuffering capacity, conceivably leading to proton re-lease from intracellular sources, do not represent themajor cause of decreased pHi. In that regard, the resultsdirectly demonstrated that variations of intracellularfree-Ca'+ concentration, which can cause release of H'ions from intracellular stores,35 did not contribute sig-nificantly to the ISO-induced acidification, since similareffects were observed in BAPTA-loaded myocytesmaintained in nominally Ca2`-free solution.

ISO Modulates Activity of Cl-HCO3- ExchangerOur results strongly suggest that the major mecha-

nism underlying p-agonist-induced acidosis in rat ven-tricular myocytes is a stimulation of the Cl -HCO3-exchanger. That this acidification was blocked by bothDIDS and probenecid is already a strong indication ofthe involvement of this exchanger. It should be men-tioned that the latter compound was used because of itsreported specificity as an inhibitor of the Na+-indepen-dent exchanger in Vero cells.27 However, its acidifyingeffect points to other alterations of proton movement,since selective inhibition of the Na+-independent C1--HCO3- exchanger should induce an alkalosis. In thatregard, probenecid did not prove to be more specificthan DIDS, which had similar effects on pH1. Theseresults were confirmed by the absence of ISO-inducedacidosis in myocytes incubated in Cl--free medium.Furthermore, the augmentation by ISO of the pHi rateof rise after Cl- replacement with gluconate directlysupports the concept of a stimulation of the Cl--HCO3exchanger. The alkalinization induced by removal ofexternal Cl- should mainly reflect the reversal of thisexchanger36 and consequently reversal of the ISO effecton pHi. It is interesting to note that the steady-state pHichange induced by Cl- removal was found to be 0.28 pHunit above basal value (Table). If it is assumed that thecellular buffering capacity is of the order of 40 mmol/Lper pH unit,13 that the HCO3- entering the cell israpidly converted to C02, thereby acting as a strong





base, and that there is a 1:1 exchange ratio between Cl-and HCO3-, an intracellular Cl- concentration of 12mmol/L would then need to be exchanged with extra-cellular HCO3- to account for this pHi augmentation.Because intracellular Cl- concentration in ventricularcells is of the order of 25 mmol/L,8 this would indicatethat about half of the Cl- efflux after its removal fromthe extracellular milieu would occur through the Cl-HCO3- exchange system. The increase of pHi observedon washout of extracellular Cl- could also be due to atransient stimulation of the Na+-dependent ClW-HCO3-exchanger, inasmuch as reversal of the transmembraneCl- gradient should favor the outward movement of Cl-ions. As such, the enhanced rate of alkalinization ob-served in cells exposed to ISO could be due to astimulation of this exchanger. The presence of thisexchanger in mammalian heart has been questioned byVaughan-Jones and collaborators,5,6 who demonstratedthe activity of a Cl--independent Na+-HCO3- cotrans-porter. Nevertheless, it was also shown that this Na+-dependent HCO3- entry was stimulated by ISO inguinea pig cardiac myocytes,12 and our results do notexclude this possibility. In fact, the observation thatprobenecid can reverse the effect of ISO on the rate ofpHi recovery after NH4Cl washout (Fig 4C) directlypointed to the activation by ISO of some acid extrusionmechanism(s). A parallel stimulation of alkalinizingsystems could also explain, at least partly, the apparentlack of effect of dibutyryl cAMP on steady-state pHi ifone assumes that their activation at normal pH, occursmore slowly than that of the Na+-independent Cl--HCO3- exchanger. Contrary to ISO and forskolin,which induce a rapid rise of intracellular cAMP levels,the rate-limiting factor with dibutyryl cAMP would beits intracellular diffusion, thereby leading to a slow butconcomitant stimulation of both Cl -HCO3 exchangemechanisms. In other words, ,B-adrenergic stimulationof both Na+-independent and Na+-dependent Cl-HCO3- exchangers can occur in rat ventricular myo-cytes, but the former one would initially prevail underour experimental conditions. This phenomenon wouldbe similar to the action of arginine vasopressin inmesangial cells,36 which stimulates the three acid-basetransporters but whose overall effect is an intracellularacidification.Recent works from other laboratories suggested that

the Na+-H+ antiport activity in guinea pig ventricularmyocytes'2 as well as in sheep Purkinje fibers16 might bereduced by ISO. Such an inhibition could account forour observed acidification. However, the fact that theISO-induced acidosis was larger than that caused bysaturating concentrations of amiloride or HMA and itspersistence in the presence of these two blockers readilysuggest that inhibition of the Na+-H+ antiport does notplay a major role in this decrease of pHi. It should bementioned that most of our measurements were done atrelatively high pHi levels. Under those conditions, theantiport should be largely inoperative,4 as demonstratedby the small acidifying effect of the blockers. However,we found, if anything, some evidence for a stimulationof the antiport, since at lower pHi values after NH4C1washout, ISO accelerates pHi recovery in the presenceof probenecid (Fig SC).

In conclusion, the present experimental data strongly

myocytes is the result of a stimulation of the ClW-HCO3-exchanger. The underlying mechanisms remain to beelucidated, although a Ca'+-dependent pathway hasbeen ruled out. Alternatively, this stimulation couldoccur indirectly through perturbations of Cl- gradientafter alterations by ISO of transmembrane Cl- pathwaysparallel to the Cl-HCO3- exchanger. Thus, Na+-depen-dent Cl- cotransport systems have been shown to befunctional in ventricular muscle8 and could be under,B-adrenergic control. The now well-characterizedcAMP-dependent Cl- conductance37 could also playsuch a role. Although its presence in rat ventricularmyocytes has been questioned,38 our observation of an

ISO-induced inward shift of the steady-state current atO mV is consistent with such a current. These hypothet-ical alterations of transmembrane Cl- movement couldexplain the ISO-induced acidosis if they were to inducea rapid decrease of intracellular Cl- concentration,39thereby increasing net HCO3- extrusion through activa-tion of the Na+-independent Cl--HCO3- exchangerand/or inhibition of the Na+-dependent ClW-HCO3-exchanger. On the other hand, this presumed diminu-tion of internal Cl- content should induce, if anything, a

diminution of the rate of pHi rise elicited on washout ofexternal Cl- , an effect opposite our observation. Inother words, it would appear unlikely that a change ofintracellular Cl- activity could explain both the ISO-induced decrease of steady-state pHi and its enhancingeffect on pHi rise in zero Cl- concentration. Accord-ingly, these results strongly argue in favor of a directcontrol by ISO of the Cl--HCO3- exchanger.At present, we can also only speculate on the physi-

ological significance of the ClW-HCO3- stimulation by,B-adrenergic agonists. It is conceivable that enhancedextrusion of HCO3- ions would alleviate the intracellu-lar CO2 accumulation that occurs under adrenergicstimulation of the heart and/or allow the neutralizationof the concomitant acid accumulation in the fourfoldnarrower intercellular space. Conversely, the possibleparallel stimulation of the Na+-dependent HCO3- influxand/or the Na+-H+ antiport would confer to the workingmyocytes an enhanced effective buffering capacity,thereby allowing a more precise control of pHi for cellsexposed to catecholamines. The latter possibility wouldbe in line with results from pH measurements in intactheart,14'15 which indicated that norepinephrine as wellas dibutyryl cAMP increased the effective bufferingcapacity of the heart exposed to hypercapnic conditions.

AcknowledgmentDr Desilets was a visiting scientist supported by

MRC-INSERM.

References1. Piwnica-Worms D, Lieberman M. Microfluorimetric monitoring of

pHi in cultured heart cells: Na+/H+ exchange.Am J Physiol. 1983;244:C422-C428.

2. Frelin C, Vigne P, Lazdunski M. The role of the Na+/HI exchangesystem in cardiac cells in relation to the control of the internal Na'concentration: a molecular basis for the antagonistic effect ofouabain and amiloride on the heart. J Biol Chem. 1984;259:8880-8885.

3. Ellis D, MacLeod KT. Sodium-dependent control of intracellularpH in Purkinje fibres of sheep heart. J Physiol (Lond). 1985;359:81-105.

4. Vaughan-Jones RD. Regulation of intracellular pH in cardiacsuggest that the ISO-induced acidosis in rat ventricular muscle. In: Ciba Foundation Symposium 139. Proton Passage




Dtsilets et al /-Adrenergic Regulation of pHi 869

Across Cell Membranes. Chichester, England: John Wiley & SonsInc; 1988:23-46.

5. Dart C, Vaughan-Jones RV. Na+-HCO3 symport in the sheepcardiac Purkinje fibre. J Physiol (Lond). 1992;451:365-385.

6. Lagadic-Gossmann D, Buckler KJ, Vaughan-Jones RD. Role ofbicarbonate in pH recovery from intracellular acidosis in theguinea-pig ventricular myocyte. JPhysiol (Lond). 1992;458:361-384.

7. Vaughan-Jones RD. An investigation of chloride-bicarbonateexchange in the sheep cardiac Purkinje fibre. J Physiol (Lond).1986;379:377-406.

8. Baumgarten CM, Duncan SWN. Regulation of Cl- activity inventricular muscle: Cl-/HCO3- exchange and Na+-dependent Cl-cotransport. In: Dhalla NS, Pierce GN, Beamish RE, eds. HeartFunction and Metabolism. Boston, Mass: Martinus Nijhoff Pub-lishing; 1987:117-131.

9. Liu S, Piwnica-Worms D, Lieberman M. Intracellular pH regu-lation in cultured embryonic chick heart cells: Na+-dependentCl- /HCO3- exchange. J Gen PhysioL 1990;96:1247-1269.

10. Iwakura K, Hori M, Watanabe Y, Kitabatake A, Cragoe E,Yoshida H, Kamada T. a1t-Adrenoceptor stimulation increasesintracellular pH and Ca2+ in cardiomyocytes through Na+/H' andNal /Ca2+ exchange. Eur J Pharmacol 1990;186:28 - 40.

11. Terzic A, Puc6at M, Cl6ment 0, Scamps F, Vassort G. ai-Adren-ergic effects on intracellular pH and calcium and on myofilamentsin single rat cardiac cells. J Physiol (Lond). 1992;447:275-292.

12. Lagadic-Gossmann D, Vaughan-Jones RD. Coupling of dual acidextrusion in the guinea-pig isolated ventricular myocyte to al- and3-adrenoceptors. J Physiol (Lond). 1993;464:49-73.

13. Puc6at M, Cl6ment 0, Vassort G. Extracellular MgATP activatesthe ClV/HCO3y exchanger in single rat cardiac cells. J Physiol(Lond). 1991;444:241-246.

14. Riegle KM, Clancy RL. Effect of norepinephrine on myocardialintracellular hydrogen ion concentration. Am J Physiol 1975;229:344-349.

15. Fenton RA, Gonzales NC, Clancy RL. The effect of dibutyrylcyclic AMP and glucagon on the myocardial cell pH. RespirPhysiol.1978;32:213-223.

16. Guo H, Wasserstrom JA, Rosenthal JE. Effect of catecholamineson intracellular pH in sheep cardiac Purkinje fibres. J Physiol(Lond). 1992;458:289-306.

17. Harada H, Kanai Y, Anzai M, Suketa Y. cAMP activatesCl-/HCO3- exchange for regulation of intracellular pH in renalepithelial cells. Biochim Biophys Acta. 1991;1092:404-407.

18. Yanaka A, Carter KJ, Goddard PJ, Silen W. Prostaglandin stim-ulates ClW-HCO3- exchange in amphibian oxynticopeptic cells.AmJPhysiol. 1992;262:G44-G49.

19. Green J, Kleeman CR. Role of calcium and cAMP messengersystems in intracellular pH regulation of osteoblastic cells. Am JPhysiol. 1992;262:C111-C121.

20. Vigne P, Breittmayer J-P, Frelin C, Lazdunski M. Dual control ofthe intracellular pH in aortic smooth muscles by a cAMP-sensitiveHCO3-/CP~ antiporter and a protein kinase C-sensitive Na+/H+antiporter. J Biol Chem. 1988;263:18023-18029.

21. Tonnessen TI, Aas AT, Ludt J, Blomhoff HK, Olsnes S. Regu-lation of Na+/H+ and Cl-/HCO3- antiports in Vero cells. J CellPhysioL. 1990;143:178-187.

22. Puceat M, Clement 0, Lechene P, Pelosin JM, Ventura-Clapier R,Vassort G. Neurohormonal control of calcium sensitivity of myo-filaments in rat single heart cells. Circ Res. 1990;67:517-524.

23. Eley DW, Korecky B, Fliss H, D6silets M. Calcium homeostasis inrabbit ventricular myocytes: disruption by hypochlorous acid andrestoration by dithiothreitol. Circ Res. 1991;69:1132-1138.

24. Dooley DJ, Bitteger H, Reymann NC. CGP 20712 A: a useful toolfor quantitating ,1- and P2-adrenoceptors. Eur J Pharmacol 1986;130:137-139.

25. Bilski AJ, Halliday SE, Fitzgerald JD, Wale JL. The pharmacology ofa f2-selective adrenoceptor antagonist (ICI 118,551). J CardiovascPharmacol. 1983;5:430-437.

26. Simchourtz L, Cragoe EJ. Inhibition of chemotactic factor-activated Na+-H+ exchange in human neutrophils by analogues ofamiloride: structure-activity relationship in the amiloride series.Mol Pharmacol 1986;30:112-120.

27. Madshus IH, Olsnes S. Selective inhibition of sodium-linked andsodium-independent bicarbonate/chloride antiport in Vero cells.J Biol Chem. 1987;262:7486-7491.

28. Boron WF, De Weer P. Intracellular pH transients in squid giantaxons caused by CO2, NH3 and metabolic inhibitors. J Gen Physiol1976;67:91-112.

29. Hancock AA, DeLean AL, Lefkowitz RJ. Quantitative resolutionof Beta-adrenergic receptor subtypes by selective ligand binding:application of a computerized model fitting technique. MolPharmacoL 1979;16:1-9.

30. Seamon KB, Padgett W, Daly JW. Forskolin: unique diterpeneactivator of adenylate cyclase in membranes and in intact cells.Proc NatlAcad Sci USA. 1981;78:3363-3367.

31. Belardinelli L, Linden J, Berne RM. The cardiac effects ofadenosine. Prog Cardiovasc Dis. 1989;32:73-97.

32. Williamson JR. Metabolic effects of epinephrine in the isolated,perfused rat heart. J Biol Chem. 1964;239:2721-2729.

33. Clark MG, Patten GS. Adrenergic regulation of glucose metabolismin rat heart: a calcium-dependent mechanism mediated by both a-and fB-adrenergic receptors. J Biol Chem. 1984;259:15204-15211.

34. Dennis SC, Gevers W, Opie LH. Protons in ischemia: where dothey come from: where do they go to? J Mol Cell Cardiol 1991;23:1077-1086.

35. Vaughan-Jones RD, Lederer WJ, Eisner DA. Ca2` ions can affectintracellular pH in mammalian cardiac muscle. Nature. 1983;301:522-524.

36. Ganz MB, Boyarsky G, Sterzel RB, Boron WF. Arginine vaso-pressin enhances pH, regulation in the presence of HCO3 bystimulating three acid-base transport systems. Nature. 1989;337:648-651.

37. Hume JR, Harvey RD. Chloride conductance pathways in heart.Am J Physiol 1991;261:C399-C412.

38. Dukes ID, Cleemann L, Morad M. Tedisamil blocks the transientand delayed rectifier K' currents in mammalian and cardiac andglial cells. J Pharmacol Exp Ther. 1990;254:560-569.

39. Nakaya H, Hattori Y, Tohse N, Shida S, Kanno M. ,-Adreno-ceptor-mediated depolarization of the resting membrane inguinea-pig papillary muscles: changes in intracellular Nal, K',and Cl- activities. Pflugers Arch. 1990;417:185-193.




chloride dependence of ph modulation by beta-adrenergic

Documents