Pflugers Arch - Eur J Physiol (2002) 444:816–820DOI 10.1007/s00424-002-0891-0

O R I G I N A L A R T I C L E

Yanina A. Assef · Basilio A. Kotsias

An outwardly rectifying anion channel in human leukaemic K562 cells

Received: 26 April 2002 / Accepted: 13 June 2002 / Published online: 2 August 2002� Springer-Verlag 2002

Abstract In this study, an outwardly rectifying anionchannel was characterized in the cell line K562 obtainedfrom a chronic human leukaemia. Ion channel activitywas recorded in the cell-detached (inside-out) configura-tion with standard patch-clamp technology. In most of theK562 cells studied, the channel exhibited low spontane-ous activity, an outwardly rectifying current/voltagerelationship and single-channel conductances of 19 pSand 40 pS for inwards and outwards currents respectively.The channel had a low permeability for gluconate with arelative permeability Pgluconate/PCl of 0.14 and wasblocked by glibenclamide (50 �M) or diphenylamine-2-carboxylate (DPC, 1 mM) added to the cytoplasmic sideof the patch. These results are characteristic of theoutwardly rectifying Cl channel (ORCC) found in othertypes of cells.

Keywords ORCC · K562 cells · Leukaemia cells · Ionchannels · Patch clamp · Glibenclamide · DPC

Introduction

Ion channels in blood cells are involved in various cellfunctions including control of membrane potential,differentiation, T-cell activation and cell-volume regula-tion [1]. Studies have shown the presence of differenttypes of ion channels in blood cells, including inwardsand outwards rectifying K channels, mechanosensitivecation channels, Cl and Na channels and a Ca-dependentNa channel regulated by the actin cytoskeleton [2, 3, 4, 5,6].

The K562 cell line is derived from a patient withchronic myeloid leukaemia, during his blastic crisis [7].The K562 cells are arrested in very early stages of

development and provide a unique population of primitivehuman myeloid leukaemia cells that can be induced todifferentiate along the erythroid, granulocytic, macro-phage and megakaryocytic lineages in response to variousagents. Once they differentiate to mature erythrocyticforms, they begin to synthesize haemoglobin and otherproteins, with an increase in iron uptake. K562 cells areused as target cells for natural killer lymphocytes [8].

In this paper we described an anion channel in K562cells compatible with the outwardly rectifying Cl chan-nels (ORCC) found in other types of cells [9, 10, 11, 12,13, 14, 15, 16, 17]. An abstract with preliminary resultshas appeared elsewhere [18].

Material and methods

K562 cells (American Type Culture Collection, USA) growspontaneously in suspension and are sustained (1�106 cells/ml) inculture medium RPMI supplemented with 10% decomplementedbovine fetal serum, at 37 �C in humid air (5% CO2). The cells areseen as round bodies, with a high nucleus/cytoplasm ratio and adiameter of 13–23 �m.

The activity of ion channels in K562 cells was recorded in thecell-detached (inside-out) configuration using standard patch-clamptechnology and employing an amplifier (PC501, Warner Instru-ments, Hamden, Conn., USA) with a 10-GW feedback resistor.Electrical signals were filtered at 2 kHz and digitized at 6 kHzusing standard software (pCLAMP v. 6, Axon Instruments, UnionCity, Calif., USA). Currents and potentials are described using thephysiological convention, according to which negative potentialsare negative with respect to the cytoplasmic side and inwardscurrents (downwards deflections in the current traces) correspondto outwards anion flow. Potentials were corrected for junctionpotentials at the bath reference electrode with the various testsolutions using software written by Dr Peter Barry (University ofNew South Wales, Australia). All experiments were done at roomtemperature (19–24 �C).

Solutions

All concentrations are given in mM. The pipette solution contained:NaCl 140, KCl 5, CaCl2 2.5, MgCl2 1, HEPES 10 (pH 7.4), and thebath solution was the same as in the pipette. For the substitutionexperiments, NaCl in the bath was replaced by Na-gluconate or N-

Y.A. Assef · B.A. Kotsias ())Instituto de Investigaciones M�dicas Alfredo Lanari,Universidad de Buenos Aires, C. de Malvinas 3150,1427 Buenos Aires, Argentinae-mail: kotsias@mail.retina.arFax: +54-11-45238947

methyl-d-glucamine (NMDG)-Cl. Glibenclamide and diphenyl-amine-2-carboxylate (DPC) (Sigma, St. Louis, Mo., USA) werediluted in DMSO and added directly to the chamber containing thecells in solution. The final concentration of DMSO had no affectper se on the ionic activity (trials in our laboratory with up to 0.2%DMSO on various cell types; data not shown).

Statistical methods

Data are given as means€SD. Student’s t-test or ANOVA wereperformed. Differences were considered significant when P<0.05.

Results

ORCC channels can be observed after application of abrief depolarizing [19] or hyperpolarizing [20] pulses.The channel is characterized by a disparity between theamplitudes of inwards and outwards unitary current insymmetrical solution. In our hands, after excision of thepatch, a brief hyperpolarizing pulse usually elicitedchannel activity that remained steady for 20–30 min(Fig. 1A). Among the recorded channels, the outwardlyrectifying channels were observed in 1/3 of the patches.To prevent run down or inactivation of the channels,20 �M forskolin, 100 �M cAMP and 1 mM Mg-ATP wereadded to the bath [21]. Another channel with a linearconductance of about 12 pS was seen in some patches.

The current/voltage (I/V) relationship allowed theconductance to be estimated. Slope conductance wasdetermined by linear regression over the voltage rangesfor inwards (–100 to –40 mV) and outwards currents (40to 100 mV). This is shown in Fig. 1B. The correspondingsingle-channel conductances were about 19 and 40 pS,illustrating the rectifying property of the channel.

Owing to the fact that the solution used is the same forboth sides of the membrane (140 mM NaCl), current maybe carried by incoming Na+ or by outgoing Cl–. Replacingbath Na+ by NMDG did not change the conductance orthe reversal potential Er. The inwards (–100 to –40 mV)and outwards conductances (40 to 100 mV) and the Erwere: 140 NaCl solution (control) 39.8€7.8 and19.1€17.3 pS and –1.7€2.4 mV (n=10); 140 NMDG-Cl31.3€5.2 and 23.9€19.0 pS and –4.2€9.2 mV (n=5, n.s.).To demonstrate the anionic selectivity of this channel,NaCl was replaced by Na-gluconate in the bath solutionin contact with the normal intracellular membranesurface. The conductances and the reversal potentials in140 Na-gluconate were: 36.5€4.7 and 22.0€2.9 pS and–33.6€11.7 mV (n=6). The permeability ratios (PX/PCl)were calculated from the shifts of the Er using themodified Goldman-Hodgkin-Katz equation:

DEr ¼ ErðXÞ � ErðClÞ ¼ 58 � logPx½X�i þ PCl½Cl�i

PCl Cl½ �owhere DEr is the shift in Er induced by replacing 140 mMof the internal Cl– (bath solution) by the anion X(gluconate); Er(Cl) is the reversal potential in the presenceof Cl– and R, T and F have their usual meaning. The shiftin the Er to more negative potentials (P<0.01) is

indicative of a lower permeability for gluconate comparedwith Cl–, with a ratio Pgluconate/PCl of 0.14 after correctionfor the junction potential at the bath reference electrode.

We assessed the effect of glibenclamide [20, 21, 22].The drug, added to the cytoplasmic side of the patchreduced the activity at all potentials. Figure 2A shows arepresentative record in which channel activity in thepatch held at 60 mV and –60 mV is reduced by 50 �Mglibenclamide, whereby glibenclamide had no effect onsingle-channel current amplitude but increased the open-channel noise. For the calculation of open probability(Po), records with only one level were analysed. The Pofor this channel had a small dependence on membranepotential, being greater at positive potentials. Glibencla-mide decreased Po by increasing the frequency of open-to-closed transitions; this effect was more marked atnegative cell potentials (Fig. 2B, n=6).

Open- and closed-time histograms in the absence orpresence of glibenclamide were constructed for patchesheld at 60 mV. For this, event-duration histograms wereconstructed from 30-s records in which only a single openlevel was observed. An exponential function was fittedusing the least-squares routine of pCLAMP. A minimumresolution of 1 ms was imposed on the data by ignoringall events shorter than this time [23]. Addition ofglibenclamide reduced the open time constant from(control) 21.4€7.2 to 3.6€0.7 ms (glibenclamide)(P<0.002). The closed time constants were: control1.7€0.5 ms; glibenclamide 2.4€0.2 ms (P<0.03).

Fig. 1 A Single-channel currents in an excised, inside-out, humanleukaemic K562 cell patch held at different positive or negativepotentials in symmetrical 140 mM NaCl solution. Dotted linesindicate the closed state of the channels. Each trace contains 400 msof recording. B Current/voltage (I/V) relationship for in excised,inside-out patches with symmetrical NaCl solution in both pipetteand bath (n=10). Means€SD

817

Figure 3 shows histograms of current amplitude in theabsence or presence of 50 �M glibenclamide. In thepresence of glibenclamide the flickery block skews theopen level distribution towards the closed level.

The effect of 1 mM DPC upon the channel activity wastested. As with glibenclamide, the presence of 1 mM DPCreduced the channel activity in the patch (Fig. 4A). Singleconductance was not changed by DPC (Fig. 4B, n=4).

The bath [Ca2+] was 2.5 mM, equal to that used in thepipette. Due to the existence of Ca2+-sensitive Cl channelsand ORCC channels regulated by Ca2+ [22, 24] wemeasured Po and current amplitude in experiments withthe intracellular [Ca2+] (bath solution) reduced to about100 �M (nominally Ca2+-free solution, [22]) or to 0 (0Ca2+ plus 0.2 mM EGTA). Neither Po nor amplitude wereaffected by changes in intracellular [Ca2+]. Thus, in threeexperiments in which patches were held at 60 mV, Po for2.5 mM, 100 �M and 0 Ca2+ was 0.73€0.06, 0.64€0.11and 0.60€0.17 (n.s., ANOVA). The corresponding am-plitudes were: 1.58€0.37, 1.70€0.24 and 1.75€0.40 pA(n.s., ANOVA).

Discussion

We employed the patch-clamp technique to characterizethe ion channels in K562 cells. Our experiments demon-strated the presence of a rectifying anion channel with alower permeability to gluconate than to Cl– and inhibitedby glibenclamide and DPC.

Fig. 4 A Single channel currents in excised, inside-out patches inthe absence (upper row) or presence (lower row) of 1 mMdiphenylamine-2-carboxylate (DPC). Patch held at 80 mV. Dottedlines indicate the closed state of the channels. B I/V relationship inthe absence (filled circles) or presence (open circles) of 1 mMDPC. Means€SD, n=4

Fig. 2 A Representative recordings showing the effect of 50 �Mglibenclamide on the activity of single channels at voltages of–60 mV (upper row) and 60 mV (lower row). The effect ofglibenclamide is also shown in the expanded tracing (bottom). Thepatch was bathed in symmetrical 140 mM NaCl solutions and thedrug was added to the cytoplasmic side of the patch. The dottedlines indicate the closed state of the channels. B Relationshipbetween open probability (Po) and voltage in the absence (filledcircles) or presence (open circles) of 50 �M glibenclamide whenthe patch was bathed in symmetrical 140 mM NaCl solutions.Means€SD, n=6

Fig. 3 Analysis of distribution of current amplitudes for a patchheld at 60 mV in the absence (upper panel) or presence (lowerpanel) of 50 �M glibenclamide

818

The outwardly rectifying Cl channel (ORCC) is alsocalled the intermediate-conductance outwardly rectifyingCl channel (ORIC, ICOR). Although its molecularidentity and its physiological role are not well known,the ORCC is distributed widely in epithelial and non-epithelial cells (see introduction) and may play a role incell volume homeostasis. Gadolinium is an opener of theORCC in human B-lymphocytes and HeLa cells andresults in cell swelling [25], suggesting that thesechannels play a role in cell volume homeostasis. Thisproperty makes it possible for blood cells to regulate cellvolume during passage through the renal microcirculationand interstitial spaces that experience great changes inosmolarity [1]. In this regard, the Cl– transport propertiesof K562 cells are significant and the [Cl–] in K562 cells isabout 84 mM [26], a high concentration compared withother cell types and one that – as attained in matureerythrocytes – helps explain the relatively low restingmembrane potential. In addition, ORCC may be relevantfor cystic fibrosis therapy because it could function as analternative Cl channel [19, 27]. Healthy cells express bothCFTR and the ORCC channels whereas cystic fibrosisaffects the expression or the properties of both proteins[19, 28].

CFTR and ORCC are distinct proteins with a regula-tory relationship [29]. Thus, it has been proposed thatCFTR controls the regulation of other channels andtransporters [21, 29, 30, 31, 32]. More specifically,evidence suggests that CFTR, a protein belonging to thefamily of ATP-binding cassette transporters, to which P-glycoprotein-170 (Pgp 170) and the sulphonylurea recep-tor also belong, regulates ORCC by an autocrine mech-anism involving CFTR-dependent ATP release [21].Accordingly, the expression of ORCC is enhanced inthe presence of CFTR [33, 34].

The channels described in this research were highlyselective for Cl– over Na+ and had a conductanceconsistent with previous reports on ORCC [12, 14, 20].Gluconate currents and Er were observed when most ofthe internal Cl– ions had been replaced by gluconate,although the gluconate/Cl permeability ratio for thesechannels was very low under bionic conditions. When Cl–

was replaced by gluconate the Er shifted to more negativepotentials and the inwards movement of Cl– (an outwardsor positive current) is facilitated by the lower permeabil-ity to the new anion. Thus, to stop the inwards flow of Cl–

a more negative internal potential is required.Although a detailed study of the regulation of ORCC

by Ca2+ was beyond the scope of this work, ourexperiments suggest that these channels are not Ca2+

sensitive, in contrast to those described in other tissues[22]. In our experiments Po and current amplitude werenot affected by changes in the intracellular [Ca2+].Moreover, these results suggest that the small voltagesensitivity of the ORCC (Fig. 2B) is not due to a Ca2+-dependent effect. ORCC channels are characterized bytheir sensitivity to block by intracellular disulphonicstilbenes (DIDS) [13, 20, 21, 35], blockade by thesulphonylurea hypoglycaemic agent glibenclamide [20,

21, 22] and by DPC, 5-nitro-2-(3-phenylpropy-lamino)benzoate (NPPB) [9, 13, 19, 21, 36, 37] andarachidonic acid [10, 14]. In insect cells, CFTR confersthe glibenclamide sensitivity to ORCC channel [38].These channels are also blocked by micromolar concen-trations of steroids, aldosterone antagonists [20] and by acytosolic inhibitor found in the cytosol of human placentaand epithelial cells [39]. The blocking effect of gliben-clamide and DPC in excised membrane patches seen inthe present study: a reduction in Po at all voltages and adecreased open time, supports these earlier observationssuggesting open channel blockade and the drug gainingaccess to its binding site in the channel from theintracellular side of the membrane.

In conclusion, our experiments have demonstrated thepresence of a rectifying anion channel with a conductanceof 19 and 40 pS for inwards and outwards currentsrespectively and which is inhibited by glibenclamide andDPC. This is most consistent with the presence of afunctional ORCC in K562 cells.

Acknowledgements We would like to thank Dr H. Guio for thegift of the cells and Dr C. Peracchia for the gift of the A/Dconverter and to the reviewer for his/her help. B.A. Kotsias is amember of the National Research Council of Argentina (CON-ICET). This work was supported by a grant from the Ministerio deSalud de La Naci�n.

References

1. Cahalan MD, Wulff H, Chandy KG (2001) Molecular proper-ties and physiological roles of ion channels in the immunesystem. J Clin Immunol 21:235–252

2. Wieland SJ, Chou RH, Gong Q (1990) Macrophage-colony-stimulation factor modulates a differentiation-specific inward-rectificating potassium current in human leukemia cells. J CellPhysiol 142:643–651

3. Viana F, Van Acker K, De Greef C, Eggermont J, Raeymaek-ers L, Droogmans G, Nilius B (1995) Drug–transport andvolume–activated chloride channel functions in human eryth-roleukemia cells: relation to expression level of P-glycoprotein.J Membr Biol 145:87–98

4. Maximov AV, Vedernikova EA, Hinssen H, Khaitlina SY,Negulyaev YA (1997) Ca-dependent regulation of Na+-selec-tive channels via actin cytoskeleton modification in leukemiacells. FEBS Lett 412:94–96

5. Negulyaev YA, Khaitlina SY, Hinssen H, Shumilina EV,Vedernikova EA (2000) Sodium channel activity in leukemiacells is directly controlled by actin polymerization. J Biol Chem275:40933–40937

6. Staruschenko AV, Vedernikova EA (2002). Mechanosensitivecation channels in human leukaemia cells: calcium permeationand blocking effect. J Physiol (Lond) 541:81–90

7. Lozzio BL, Lozzio CB, Bamberger EG, Feliu AS (1981) Amultipotential leukemia cell line (K-562) of human origin. ProcSoc Exp Biol Med 166:546–550

8. L�bbert M, Koeffler HP (1988) Myeloid cell lines: tools forstudying differentiation of normal and abnormal hematopoieticcells. Blood Rev 2:121–133

9. Hayslett JP, G�gelein H, Kunzelmann K, Greger R (1987)Characteristics of apical chloride channels in human colon cells(HT29). Pflugers Arch 410:487–494

10. Hwang TC, Guggino SE, Guggino WB (1990) Direct modu-lation of secretory chloride channels by arachidonic and other

819

cis-unsaturated fatty acids. Proc Natl Acad Sci USA 87:5706–5709

11. Jorissen M, Vereecke J, Carmeliet E, Van der Berghe H,Cassiman JJ (1990) Outward-rectifying chloride channels incultured adult and fetal human nasal epithelial cells. J MembrBiol 117:123–130

12. Frizzell RA, Halm DR (1990) Chloride channels in epithelialcells. Curr Top Membr 37:247–282

13. Tilmann M, Kunzelmann K, Fr�be U, Cabantchik I, Lang HJ,Englert HC, Greger R (1991) Different types of blockers of theintermediate-conductance outwardly rectifying chloride chan-nel in epithelia. Pflugers Arch 418:556–563

14. Anderson MP, Sheppard DN, Berger HA, Welsh MJ (1992)Chloride channel in the apical membrane of normal and cysticfibrosis airway and intestinal epithelia. Am J Physiol 263:Ll–L14

15. Abrahamse SL, Bridges RJ, De Jong MD, Bindels RJ, VanOs CH (1996) Apical membrane isolation of surface and cryptcells from rabbit distal colon. Anal Biochem 242:112–122

16. Butt AG, Hamilton KL (1998) Ion channels in isolated mousejejunal crypts. Pflugers Arch 435:528–538

17. Catacuzzeno L, Trequattrini C, Petris A, Franciolini F (1999)Bimodal kinetics of a chloride channel from human fibroblasts.J Membr Biol 170:165–172

18. Assef Y, Kusnier C, Damiano A, Ibarra C, Kotsias BA (2000)Presencia de MDR (multidroga resistencia) y del regulador dela conductancia de la fibrosis qustica (CFTR) en c�lulas K562(abstract). Medicina (Buenos Aires) 60:793

19. Egan M, Flotte T, Afione S, Solow R, Zeltlin PL, Carter BJ,Guggino WB (1992) Defective regulation of outwardly recti-fying Cl– channels by protein kinase A corrected by insertion ofCFTR. Nature 358:581–584

20. Rabe A, Fr�mter E (2000) Micromolar concentrations ofsteroids and of aldosterone antagonists inhibit the outwardlyrectifying chloride channel with different kinetics. PflugersArch 439:559–566

21. Schwiebert EM, Egan ME, Hwang TH, Fulmer SB, Allen SS,Cutting GR, Guggino WB (1995) CFTR regulates outwardlyrectifying chloride channels through an autocrine mechanisminvolving ATP. Cell 81:1063–1073

22. Stewart GS, Glanville M, Aziz O, Simmons NL, Gray MA(2001). Regulation of an outwardly rectifying chloride con-ductance in renal epithelial cells by external and internalcalcium. J Membr Biol 180:49–64

23. Colquhoun D (1987) Practical analysis of single channelrecords. In: Standen NB, Gray PTA, Whitaker MJ (eds)Microelectrode techniques. The Plymouth Workshop Hand-book. Company of Biologists, Cambridge, pp 83–135

24. Fuller CM, Ji HL, Tousson A, Elble RC, Pauli BU, Benos DJ(2001) Ca2+-activated Cl– channels: a newly emerging aniontransport family. Pflugers Arch 443 [Suppl1]:S107–S110

25. Thinnes FP, G�tz W, Hellmann KP, Hellmann T, Merker R,Kiafard Z, Eben-Brunnen J, Schwarzer CH, Hilschmann N(2001) Gadolinium as an opener of the outwardly rectifying Cl–

channel (ORCC). Is there relevance for cystic fibrosis therapy?Pflugers Arch 443 [Suppl1]:S111–S116

26. Law F-Y, Steinfeld R, Knauf PA (1983) K562 cell anionexchange differs markedly from that of mature red blood cells.Am J Physiol 244:C68–C74

27. Welsh MJ, Li M, McCann JD (1989) Activation of normal andcystic fibrosis Cl channels by voltage, temperature, and trypsin.J Clin Invest 84:2002–2007

28. Gabriel SE, Clarke LL, Boucher RC, Stutts MJ (1993) CFTRand outward rectifying chloride channels are distinct proteinswith a regulatory relationship. Nature 363:263–268

29. Schwiebert EM, Benos DJ, Egan ME, Stutts J, Guggino WB(1999) CFTR is a conductance regulator as well as a chloridechannel. Physiol Rev 79:S145–S166

30. Kunzelmann K, Schreiber R (1999) CFTR, a regulator ofchannels. J Membr Biol 168:1–8

31. Lee MG, Wigley WC, Zeng W, Noel LE, Marino CR (1999)Regulation of Cl–/HCO3

–-exchange by cystic fibrosis trans-membrane conductance regulator expressed in NIH 3T3 andHEK 293 cells. J Biol Chem 274:3414–3421

32. Cantiello HF (2001) Role of actin filament organization inCFTR activation. Pflugers Arch 443:S75-S80

33. Grubb BR, Vick RN, Boucher RC (1994) Hyperabsorption ofNa+ and raised Ca2+ mediated Cl– secretion in nasal epithelia ofCF mice. Am J Physiol 266:C1478–C1483

34. Jovov B, Ismailov II, Berdiev BK, Fuller CM, Sorscher EJ,Dedman JR, Kaetzel MA, Benos DJ (1995) Interaction betweencystic fibrosis transmembrane conductance regulator andoutwardly rectified chloride channels. J Biol Chem270:29194–29200

35. Linsdell P, Hanrahan JW (1996) Disulphonic stilbene block ofcystic fibrosis transmembrane conductance regulator Cl– chan-nels expressed in a mammalian cell line, and its regulation by acritical pore residue. J Physiol (Lond) 496:687–693

36. Bridges RJ, Worrel RT, Frizzell RA, Benos DJ (1989) Stilbenedisulfonate blockade of colonic secretory Cl– channels in planarlipid bilayer. Am J Physiol 256:C209–C219

37. Singh AK, Afink GB, Venglarik CJ, Wang RP, Bridges RJ(1991) Colonic Cl channel blockade by three classes ofcompounds. Am J Physiol 261:C51–C63

38. Julien M, Verrier B, Cerutti M, Chappe V, Gola M, De-vauchelle G, Becq F (1999) Cystic fibrosis transmembraneconductance regulator (CFTR) confers glibenclamide sensitiv-ity to outwardly rectifying chloride channel (ORCC) in Hi-5insect cells. J Membr Biol 168:229–239

39. Hansen CP, Roch B, Kunzelmann K, Kubitz R, Greger R(1993) Small and intermediate conductance chloride channelsin HT29 cells. Pflugers Arch 424:456–464

820