Identification of endogenous outward currents in the human embryonic kidney (HEK 293) cell line

Journal of Neuroscience Methods 81 (1998) 73 – 83

Identification of endogenous outward currents in the human embryonic kidney (HEK 293) cell line Guoyun Zhu, Yan Zhang, Haoxing Xu, Chun Jiang * Department of Biology, Georgia State Uni6ersity, 24 Peachtree Center A6enue, Atlanta, GA 30303 -4010, USA Received 3 July 1997; received in revised form 5 January 1998; accepted 6 January 1998

Abstract Human embryonic kidney cells (HEK 293) are widely used as an expression system in studies of ion channels. However, their endogenous ionic currents remain largely unidentified. To characterize these currents, we performed patch clamp experiments on this expression system. In whole-cell voltage clamp mode, the HEK 293 cells showed mainly outward currents using physiological concentrations of Na + and K + and symmetric concentrations of Cl − (150 mM) across the plasma membranes. K + currents contributed to a small portion of these outward currents, since a shift of the reversal potentials of only 20 mV was seen with a change of extracellular K + concentration from 3 to 150 mM. In contrast, the reversal potential shifted  25 mV when extracellular Cl − was reduced to 50 mM, indicating that most of the outward currents are carried by Cl − . In inside-out patches, several distinct Cl − currents were identified. They were: (1) 350 pS Cl − current, which was voltage-activated and had a moderate outward rectification; (2) 240 pS Cl − current with a weak outward rectification; and (3) 55 pS Cl − current, which was voltage-activated, sensitive to DIDS, and showed a strong outward rectification. Activation of these Cl − currents did not require an elevation of free Ca2 + level in the cytosol. Besides these three currents, we observed two other Cl − currents with much smaller conductances (25 and 16 pS, respectively). Two different K + currents were seen in the HEK 293 cells, with one of them (125 pS) showing inward rectification and the other (70 pS) outward rectification. Moreover, a 50 pS cation channel was recorded in these cells. The presence of a variety of ion channels in the HEK 293 cells suggests that a great precaution needs to be taken when this expression system is used in studies of several similar ion channels. © 1998 Elsevier Science B.V. All rights reserved. Keywords: Cl − current; K + current; Cation current; Patch clamp; Single channel; Expression system

1. Introduction Our understanding of neuronal membrane electrophysiological behaviors has been dramatically increased over the past decade. This is largely attributed to the introduction of patch clamp and molecular genetic technologies, especially cloning, expression and mutagenesis of ion channels. Several expression systems have

Abbre6iations: EGTA, ethylene glycol-bis(b-aminoethylether)N,N,N%,N%-tetraacetic acid; Gluconate, gluconic acid; HEPES, N-(2-hydroxyethyl)piperazine-N%-(2-ethanesulfonic acid); NMDG, N-methylD-glucamine; MSF, methanesulfonic acid; TEA, tetraethyl ammonium. * Corresponding author. Tel.: + 1 404 6510913; fax: + 1 404 6512509; e-mail: [email protected] 0165-0270/98/$19.00 © 1998 Elsevier Science B.V. All rights reserved. PII S0165-0270(98)00019-3

played a critical role in this achievement, including Xenopus oocytes, insect Sf9 cells, COS cells, Chinese hamster ovary (CHO) cells, and human embryonic kidney cells (HEK 293). While the Xenopus oocyte expression system provides only a transient expression of foreign channels, a stable expression of these proteins can be produced in some of the other expression systems such as the HEK 293 cells, which has a clear advantage not only in saving labor costs but also in minimizing variations between each expression experiments (Claudio, 1992). Most importantly, experiments of co-expression of several foreign genes can be designed based on a stable expression of one of these genes first in the cells followed by a transient or stable expression of other genes. Besides, the HEK 293 cells

G. Zhu et al. / Journal of Neuroscience Methods 81 (1998) 73–83

74 Table 1 Solutions for patch clamp Solution

K+

Na+

Ca2+

A B C D E F G H I J

3 3 3 3 3 150 3 3 3 3

140

1 1 1 1.003

5 5 5

NMDG

1 1 1

Cs+

145 140 140 140 94

140 40 40

TEA

20

26 100 100

Cl−

Mg2+ Gluconate

147 152 152 152 152 150 50 47 47 47

1 1 1 1 1 0.2 0.5 1 1 1

have two other advantages in ion channel expression. (1) They are mammalian cells. Hence, the transcription, translation and post-translation modification machineries for expressing mammalian ion channels are conserved. (2) They are ideal in expression of certain species of ion channels that are normally expressed only in excitable cells, since these ion channels may not exist in the HEK 293 cells (Ukomadu et al., 1992; Brust et al., 1993; Critz et al., 1993; Cummins et al., 1993; Buisson et al., 1996; Kohr and Seeburg, 1996; Swanson et al., 1997). Therefore, the HEK 293 cells become one of the most useful expression systems in ion channel studies Hoo et al., 1994; Korpi et al., 1994; Fletcher et al., 1995; Gopalakrishnan et al., 1995; Moran and Conti, 1995; Gopalakrishnan et al., 1996; Raymond et al., 1996; Tseng-Crank et al., 1996). To express exogenous ion channels, the most fundamental information that should first be known is what species of native ionic currents are expressed in the HEK 293 cells. If the endogenous ion currents are identical or similar to those to be expressed, other expression systems may therefore need to be considered. Thus, this opens a question as to whether all or most of the endogenous ion currents in the HEK 293 cells can be identified, in addition to a few of the native ion channels that have previously been observed. One of these known endogenous currents is the TTX-sensitive Na + channel that was occasionally observed in these cells, and the expression density of this Na + channel is 5–10 pA/pF Ukomadu et al., 1992; Cummins et al., 1993). Another is an endogenous Ca2 + channel, ISR-HEK, that is activated by depolarization to − 40 mV and is sensitive to dihydropyridine antagonists Berjukow et al., 1996). The expression density of this Ca2 + channel is 0.24 – 0.39 pA/pF Berjukow et al., 1996). In addition, there is a 48 pS Ca2 + -dependent cation channel, though it has not been systematically studied Ye et al., 1996). Although the inward currents have been characterized in the HEK 293 cells, endogenous outward currents remain to be identified. In order to characterize the outward currents that are endoge-

ADP

0.2 95 100 100

EGTA

MSF

HEPES

Glucose

100

10 10 10 10 10 10 10 10 10 10

10 10 10 10 10 10 10 10 10 10

1 1 1 1

nously expressed in the HEK 293 cells, we designed these patch-clamp experiments. Our results have shown that a variety of ion channels are normally expressed in this expression system. 2. Methods Human embryonic kidney HEK 293) cells were purchased from American Type Culture Collection CRL1573, Batch c F-13543, and F-14633, ATCC, Rockville, MD). Cells were grown in MEM-E medium with 10% fetal bovine serum and penicillin/streptomycin added, and incubated at 37°C with 5% CO2 in atmosphere. The cells were cultured as monolayer in flask and routinely split twice weekly. For patch-clamp experiments, a few drops of the cell suspension were added to a petri dish, and then incubated at 37°C for at least 2 h before experiments. Patch-clamp experiments were performed at room temperature  24°C). Fire-polished patch pipettes 2–6 MV) were made from thick-wall borosilicate glass capillaries using a Sutter P-97 puller Sutter Instrument, Novato, CA). Whole cell currents were studied in the voltage-clamp mode, and single channel currents were recorded from inside-out patches Hamill et al., 1981) using an Axo-patch 200B amplifier Axon Instruments, Foster City, CA). Current records were low-pass filtered Bessel, 4-pole filter, 3 dB at 2 kHz), digitized with pClamp 6.03 software Axon Instruments), and stored on computer disk for the subsequent data analysis. Jiang and Haddad, 1994; Jiang et al., 1994; Jiang and Haddad, 1997). After formation of gigaohm seal, series resistance was monitored with the Axo-patch 200B amplifier. Further experiments were done when the series resistance was B 10 MV in whole-cell voltage clamp. To study voltage-activated currents, linear leak subtraction was performed on whole-cell currents by applying five hyperpolarizing voltage pulses from a holding potential of −60 mV prior to each test pulse. Junction potential between bath and pipette solutions was appropriately nulled before seal formation.

G. Zhu et al. / Journal of Neuroscience Methods 81 (1998) 73–83

75

Fig. 1. Outward currents recorded from an HEK 293 cell in whole-cell voltage clamp mode. A) Membrane potential Vm) of the cell was held at −60 mV and depolarized with a series of step pulses from − 50 to 100 mV with 10 mV increments. Solution A was used in the bath and solution F in the pipette. B) Whole-cell currents recorded using ramp voltage commands from −100 to 100 mV. In physiological levels of Na + and K + solution A in bath and solution F in pipette), the currents showed a reversal potential of − 20.5 mV solid lines). When extracellular K + concentration increased from 3 to 150 mM, the reversal potential shifted to 0 mV dash line). This was followed by a complete recovery after washout solid line). A decrease in extracellular Cl − concentration from 150 mM to 47 mV changed the reversal potential by 25 mV to 4.4 mV, dot line). Note that each current trace was obtained from an average of six current records.

For whole-cell recording, one pair of several bath and pipette solutions shown in Table 1 was selected depending on experimental protocols. For single channel studies, identical solutions were applied to the bath and recording pipette. Osmolarity of all solutions was measured and adjusted to 300 mosmol Vapro 5520 Osmometer) and pH to 7.4. A perfusion system was used to apply solutions in parallel to cells or patches at a rate of  1 ml/min. This perfusion system allows a fast change of solutions surrounding the cell or patch membrane in  200 ms. Single-channel data were analyzed by further filtering 0 – 1000 Hz) with a Gaussian filter. Events shorter than 200 ms were ignored. The unitary conductance was calculated by measuring slope current amplitudes in the I/V plot using a slope voltage command. The channel open-state probability popen) was calculated as described previously Sigworth and Sine, 1987; Jiang and Haddad, 1994; Jiang et al., 1994; Jiang and Haddad, 1997): 1) For stretches of data having a total duration of ]20 s) obtained using the Fetchex software, the time, tj, spent at current levels corresponding to j =0, 1, 2,…N channels open was first measured. The popen was then ex-





pressed as a function of tj :popen = N j = 1tj j /TN,where N is the number of channels active in the patch and T is the duration of recordings. 2) For segmented data obtained using the Clampex software, stimulating artifact and leak currents were first removed by subtracting these currents from each traces of records, and 40–50 records were then averaged. popen values were calculated as:popen = I/ni where I is the amplitude of averaged

macroscopic currents, n the number of channels active in the patch, and i the amplitude of single channel current. Chemicals were all purchased from Sigma St. Louis, MO). Data are presented as means9 SE, standard error, n= number of patches) and differences in means were tested with the Student’s t-test and accepted as significant if p5 0.05.

3. Results

3.1. Whole cell currents In voltage clamp mode with solution F in recording pipettes and solution A in the bath, only outward currents were recorded in the HEK 293 cells. The average amplitude of the currents was 0.87 9 0.07 nA n= 60), and the peak current density 40.89 3.3 pA/pF without leak subtraction Fig. 1A). Most of these currents were voltage-dependent, since the amplitude of these currents was reduced by only 18.8%, from 0.859 0.05 nA without leak subtraction to 0.699 0.03 nA with the leak subtraction pB 0.05, n= 23) using a voltage command of 100 mV from a holding potential of − 60 mV. When free Ca2 + in the bath solution was reduced from 1 mM to 10 nM, the amplitude of these currents decreased by 14 9 9% p \0.05, n =3), suggesting that the majority of these currents are Ca2 + independent.

76

G. Zhu et al. / Journal of Neuroscience Methods 81 (1998) 73–83

When ramp voltage commands were applied to the cell from −100 to 100 mV or from 100 to − 100 mV at a holding potential of − 60 mV, these currents showed a reversal potential of − 20 mV using a physiological level of Na + and K + and symmetric concentrations of Cl − across the plasma membranes Fig. 1B). A rise in extracellular K + concentration from 3 to 150 mM Solution F) shifted the reversal potential to  0 mV Fig. 1B), suggesting that K + channels are expressed in this cell line but are not the main contributors to these outward currents. A decrease in extracellular Cl − concentration from 150 to 47 mM Solution H) displaced the reversal potential by 25 mV Fig. 1B). When 400 mM 4,4%-diisothiocyanatostilbene2,2%-disulfonic acid DIDS, a Cl − channel blocker) was added to the extracellular solution, the current amplitude was inhibited by 4598% p B0.05; n = 7). These results suggest that most of the outward currents appear to be carried by Cl − .

3.2. Single channel currents To characterize specific ionic currents, single channel activity was studied in inside-out patches. In these experiments, patch recordings were first made with solution F in pipette and solution A in bath. Patches were excised from the cell 1 – 2 min after G-seal formation, and 20–30 records were immediately made using ramp voltage commands. The internal bath) solution was then changed to solution F, and subsequent experiments were carried out under this condition i.e. the same solution applied to both internal and external surfaces of the membranes) except the experiments aimed at testing ion selectivity. This solution solution F) containing 150 mM K + and 150 mM Cl − was chosen to show K + and Cl − currents, because endogenous inward currents have previously been studied in these cells Ukomadu et al., 1992; Berjukow et al., 1996). About 50% 150) patches were studied in such a way. The other 50%  150) patches were recorded using solutions B, C, D and E, in which both Na + and K + were largely or completely removed so that we were able to focus on only Cl − currents.

3.2.1. 350 pS Cl − channel In excised inside-out patches, a large-conductance current was observed in 15 patches from 300 cells 1 patch per cell). This current had single channel conductance of 34595 pS n =15) with moderate outward rectification Fig. 2C, Fig. 3A1). When Cl − concentration on the internal surface of the membrane was reduced to 47 mM solution H Cl − was replaced by 100 mM gluconate), a shift of reversal potential of − 23.39 1 mV n=3) was seen with a decrease in the unitary conductance to 230925 pS n = 3) Fig. 3A2). This channel showed strong voltage dependence. Chan-

nel activation started at − 80 mV and reached a maximal level at − 20 mV with a 50% maximal activation at − 49 mV Fig. 2A,B). Activation of this current did not require the presence of Ca2 + in the cytosol, because 1) this current was seen with  10 nM free Ca2 + in the intracellular solution, and 2) an increase in free Ca2 + level to 1 mM did not raise channel activity. Thus, these results indicate that this large-conductance current is recorded from a voltage-activated, Ca2 + -independent Cl − channel.

3.2.2. 240 pS Cl − channels Under the same experimental conditions, another current was observed. This current had a slope conductance of 240 99 pS n =10) with a modest outward rectification Fig. 3B1) using solutions of symmetric Cl − concentrations on both sides of membranes. When Cl − concentration on the internal surface of the membrane was reduced to 47 mM solution G or I), the reversal potential moved to −25 mV, and the unitary conductance decreased to 212 915 pS n= 6) Fig. 3B2). Channel activity was not affected by a change in free Ca2 + concentration from 1 mM to 10 nM in the intracellular solution. Its voltage-dependence was not examined because a rapid rundown was seen in this current. These indicate that this channel is highly Cl − -selective and Ca2 + -independent. 3.2.3. 55 pS Cl − channel With equal concentrations of Cl − applied to both sides of the membrane, a small conductance current was observed in 12 patches from  300 cells Fig. 4). Single channel conductance of this current was 5696 pS n= 12) with a strong outward rectification Fig. 4A1). When Cl − concentration on the internal surface was reduced to  50 mM solution G, H or I), the reversal potential of this current shifted  25 mV to the negative direction, and the unitary conductance decreased to 319 3 pS n= 10) Fig. 4B2). This current was also voltage-activated. Channel activity started to increase at a membrane potential Vm) of − 10 mV and reached a maximal activation at 40 mV. The Vm for 50% of the maximal activation was at 17 mV Fig. 4B). Baseline channel open state probability popen) was 0.605 at a Vm of 80 mV. The channel activity was inhibited strongly popen was reduced to 0.029) when the patch was exposed to 200 mM DIDS. The inhibition of popen by DIDS was restored completely after washout popen = 0.702), but single channel conductance only showed a partial recovery Fig. 4C). When free Ca2 + was reduced from 1 mM to 10 nM, popen did not show any significant change. These results therefore demonstrate another Ca2 + -independent Cl − channel. In addition to these three distinct Cl − currents, we observed two other currents that had much smaller conductances that were 25 and 16 pS, respectively.

Fig. 2. Large conductance current recorded in an inside-out patch from an HEK 293 cell. A) Single channel current was recorded in an inside-out patch. The membrane potential Vm) was held at 0 mV, and step potentials were given to a level indicated on the left side with both internal and external solutions containing equal concentrations of Cl − 150 mM). Solid lines indicate channel openings and the dash lines closures. B) Voltage-dependent activation of this current. Under the same experimental condition as in A, channel activity popen, normalized to its maximum level) starts to increase at : − 80 mV. A full activation is reached when Vm is \20 mV. The relation of popen to Vm can be described with the Boltzmann expression: popen = 1/{1+ exp [Kv − Vm)/k]}, where popen is the open-state probability that has been normalized to its maximal level, Vm the membrane potential, Kv = −49 mV Vm at 50% of popen), and k = 10 mV the Boltzmann coefficient). C) Current –voltage I–V) relation of this large conductance current. Each point was taken at a corresponding Vm in A. The dot line represents a slope conductance of 370 pS.

G. Zhu et al. / Journal of Neuroscience Methods 81 (1998) 73–83 77

78

G. Zhu et al. / Journal of Neuroscience Methods 81 (1998) 73–83

Fig. 3. Conductance, rectification and Cl − selectivity of two different single channel currents. Single channel currents were recorded from inside-out patches from two different HEK 293 cells. Membrane potential Vm) was held at 0 mV, and ramp command potentials were given from − 100 to 100 mV. A1, A2: 370 pS Cl − current from the same patch in Fig. 2 A1. When internal and external solutions contained equal concentrations of Cl − 150 mM), this current showed a moderate outward rectification with a reversal potential at  0 mV. The dot line indicates a slope conductance of 370 pS. A2: When Cl − concentration in the internal solution was reduced to 47 mM substituted with 100mM gluconate), the reversal potential shifted to − 24.4 mV arrow) and the unitary conductance was reduced to 180 pS. B1, B2: The 220 pS Cl − current from another patch. B1. When internal and external solutions contained equal concentrations of Cl- 150 mM), this current showed a weak outward rectification with a reversal potential at 4 mV. The dot line indicates a slope conductance of 220 pS. B2: When Cl − level in the internal solution was reduced to 50 mM with 100 mM gluconate added), the reversal potential was displaced to −24.0 mV arrow), and the unitary conductance reduced to about 150 pS.

None of them showed an voltage range from − 100 currents were recorded with of membranes, it is likely currents.

evident rectification in a to 100 mV. Since these solutions B on both sides that they are also Cl −

3.2.4. Potassium channels Since our internal and external solutions contain 150 mM K + and 150 mM Cl − Solution F) in about half of the recordings, there is a high likelihood to record from K + channel as from Cl − channels. In contrast to Cl −

currents that were generally seen in one out of every 5–6 patches, we only observed K + currents in four patches from a total of  150 recordings. Two distinct K + currents were seen in these four patches. One of them had single channel conductance of 125 pS with inward rectification Fig. 5A1). This channel was highly selective to K + , since its reversal potential migrated to near 80 mV when K + concentration on the internal surface was reduced from 150 Solution F) to 3 mM solution A) Fig. 5A2). The other current showed a single channel conduc-

G. Zhu et al. / Journal of Neuroscience Methods 81 (1998) 73–83

79

Fig. 4. Characterization of the 55 pS Cl − channel. A1, A2: A single channel current was recorded in an inside-out patch. Ramp command potentials were given from −100 to 100 mV at a holding potential of 0 mV. A1: When symmetric concentrations of Cl − 150 mM) were applied to both sides of the membrane, this current showed an outward rectification with a reversal potential at 6 mV. The dot line indicates a slope conductance of 55 pS. A2: When Cl − concentration in the internal solution dropped to 47 mM, the reversal potential of this current shifted to near − 18 mV arrow), and the conductance of the current became 25 pS. B: Channel activity of this current is a function of Vm. Single channel current was recorded from another inside-out patch with equal concentrations of Cl − 150 mM) in both internal and external solutions. Channel activity popen, normalized to its maximum level) started to increase at : − 10 mV. Full activation was reached when Vm was \40 mV. The relation of popen to Vm can be described with the Boltzmann expression, with Kv = −17 mV and k =12 mV. C: Effect of 4,4%-Diisothiocyanatostilbene-2,2% disulfonic acid DIDS) on channel activity of the 55 pS Cl − current. Single channel current was recorded from an inside-out patch with membrane potential held at 80 mV and the same solution applied to both sides of the membrane. This channel had baseline popen of 0.605 top two traces). Application of DIDS 200 mM) to the cytosolic side of the membrane caused a marked inhibition of channel activity popen = 0.029, middle two traces). This inhibitory effect on popen was reversible, and channel activity was restored after washout of DIDS popen =0.702, bottom two traces). However, the unitary conductance showed an only partial recovery after washout. Note the lower traces in each group were expended from the top ones between two arrows. Solid lines indicate channel openings and the dash lines closures.

tance of 679 3 pS n = 3) with outward rectification Fig. 5B1). This current was also K + selective, although its K + selectivity was not as high as the previous K + current, i.e. when K + concentration on the internal surface was reduced to 3 mM Solution A), the reversal potential was displaced to near 60 mV and the unitary conductance decreased to 429 8 pS n =3) Fig. 5B2).

3.2.5. Cation channel Besides these Cl − and K + currents, we observed a current that had a single channel conductance of 5392 pS with outward rectification in 15 patches Fig. 6). Using solution F external) and solution A internal), this current had a reversal potential  0 mV, indicating that it is not a K + current. This reversal potential was

80

G. Zhu et al. / Journal of Neuroscience Methods 81 (1998) 73–83

Fig. 5. K + selective currents in HEK 293 cells. Single channel currents were recorded in inside-out patches with ramp command potentials from − 100 to 100 mV at a Vm of 0 mV. A1, A2: 125 pS K + current recorded from an HEK 293 cell. A1: Using symmetric concentrations of K + 150 mM), this current showed an inward rectification with a reversal potential at 0 mV. The dot line indicates a slope conductance of 125 pS. A2: When K + in the internal solution was reduced to 3 mM K + substituted with 140 mM Na + ), the reversal potential shifted to near 80 mV arrow) and the unitary conductance was reduced to  100 pS. B1, B2: A smaller K + current recorded from another HEK 293 cell. B1: When internal and external solutions contained equal concentrations of K + 150 mM), this current showed an outward rectification with a reversal potential at 0 mV. The dot line indicates a slope conductance of 60 pS. B2: When K + in the internal solution is reduced to 3 mM, the reversal potential of this current shifted to 60 mV arrow) with no change in the unitary conductance.

not affected by a decrease of Cl − concentration from 150 to 50 mM on internal sides of membranes without changing cation concentrations solution H). Substitution of 100 mM NaCl with 100 mM cesium gluconate or cesium methanesulfonate solution I or J) in the internal solution did not change the reversal potential either, although it slightly reduced the unitary conductance.

These indicate that this channel is cation-selective. These experiments were done using two batches of HEK 293 cells F-13543, and F-14633, ATCC) with up to 30 passages. Our experience was that the expression density of these endogenous Cl − , K + , and cationic channels was independent of batches of cells and the passage numbers.

G. Zhu et al. / Journal of Neuroscience Methods 81 (1998) 73–83

4. Discussion This is the first systematic study of endogenous outward currents in the HEK 293 cells, in which we have shown that a variety of ionic currents are expressed in these cells. These include five Cl − currents, two K + currents and one cation current. Their single channel properties and voltage sensitivity are clearly different from each other, indicating that each of them consists of a distinct group of ion channels. The presence of endogenous ionic currents in the HEK 293 cells have been noticed in several previous studies Ukomadu et al., 1992; Attali et al., 1993; Cummins et al., 1993; Berjukow et al., 1996; Ye et al., 1996). Detailed characterization of the outward currents has been done in our current studies. First, we have found that a great proportion of these endogenous currents in

Fig. 6. Cation current in an HEK 293 cell. A single channel current was recorded in an inside-out patch using a holding potential of 0 mV and ramp command potentials from − 100 to 100 mV. A: When solution A internal) and solution F external) were used, this current showed an outward rectification with a reversal potential at 0 mV. The dot line indicates a slope conductance of 60 pS. B: Substitution of 100 mM Cl − with the same concentration of gluconate in the internal solution without changing cation concentrations had no effect on both reversal potential and unitary conductance.

81

HEK 293 cells is voltage-activated, although the reason for the presence of these currents in these epithelial cells is not understood. At single channel level, two Cl − currents have been shown, in our current studies, to be voltage-gated with a midpoint activation of −45 and 17 mV, respectively. Therefore, the observation of the voltage-dependence of single channel currents is consistent with our whole-cell data. Second, most of the endogenous currents show an outward rectification in inside-out patches, which fits well to our data in wholecell recordings showing also mainly outward rectification in these cells. Despite the weakness or absence of rectification in the 25 pS and 16 pS Cl − channels, their contribution to the whole-cell currents seems small in the sense that they have very small conductance with a low appearance in our single channel patch recordings. One of the K + currents we observed has a strong inward rectification. This indicates that though the number is small, there are inward rectifier K + currents, especially when the external K + is high. Third, besides the 125 pS inward rectifier K + current, we have recorded another K + current that has an outward rectification in three out of 150 patches. This low incidence in recordings from K + currents suggests that K + channel density is relatively low in the HEK 293 cells, which is also consistent with our whole-cell recordings. However, the rare appearance of K + currents in inside-out patches cannot overthrow the possibility that there may be other K + currents with conductances undetectable in our recording system. Even though these extremely small K + channels may exist, their contribution to the whole-cell currents still cannot be overwhelming. This is again based on our observations of the low density of overall K + currents in whole-cell recordings. Fourth, one of the other interesting findings from our current studies is that none of these ionic currents is Ca2 + -dependent. We did not see any evident increase in channel activity popen) in all these identified currents when free Ca2 + concentration increased from 10 nM to 1 mM. Therefore, our data indicate that several Ca2 + -independent Cl − and K + channels are endogenously expressed in the HEK 293 cells. Currents observed in these studies were classified according their single channel conductance, ion selectivity or sensitivity to pharmacological blockers. The 350 pS Cl − channels seem close to the Maxi-Cl − channels, both having high Cl − selectivity, large conductance  350 pS) and outward rectification Mclarnon and Kim, 1991; Becq et al., 1992; Brown et al., 1993; Marcus et al., 1993; Kajita et al., 1995). The properties of 55 pS Cl − channels in HEK293 such as outward rectification, conductance and voltage-dependence are similar to those of epithelial Cl − channels Gogelein, 1988; Vaughan and French, 1989). The 250 pS Cl − channels were also identified from canine tracheal ep-

82

G. Zhu et al. / Journal of Neuroscience Methods 81 (1998) 73–83

ithelial and human skeletal muscle cells, which are voltage-independent Fahlke et al., 1992; Shoemaker et al., 1986). A 70–85 pS K + channel has been identified in rat and rabbit kidney Bleich et al., 1990; Wang, 1995). The conductance is similar to that of K + channels that we observed in the HEK 293 cells with different rectification and expression density Bleich et al., 1990; Wang, 1995). Taken together, the endogenous K + , Cl − channels in HEK 293 cells share many similarities to those reported previously in kidney and several other tissues. Based on these results, we believe that the HEK 293 cells are an ideal expression system for Na + and K + channel studies. Their utilization in the expression of Cl − channels, however, may be limited to those that exhibit one or more different channel properties from these endogenous ionic currents, such as rectification, conductances and Ca2 + -dependence. Although identification of the expression of ion channels in the HEK 293 cells can also be done using other approaches such as Northern blotting, immuno-blotting and immunocytochemistry, patch clamp offers a unique solution for the manifestation of functional ion channels. Our current studies therefore contribute useful information to the ‘data base’ of this widely used expression system in ion channel studies.

Acknowledgements This work was supported by the NIH grant RO1 HL58410, the United Cerebral Palsy Foundation, and the American Lung Association of Connecticut.

References Attali B, Guillemare E, Lesage F, Honore E, Romey G, Lazdunski M, Barhanin J. The protein IsK is a dual activator of K + and Cl − channels. Nature 1993;365:850–2. Becq F, Fanjul M, Mahieu I, Berger Z, Gola M, Hollande E. Anion channels in a human pancreatic cancer cell line Capan-1) of ductal origin. Pflug Arch 1992;420:46–53. Berjukow S, Doring F, Froschmayr M, Grabner M, Glossmann H, Hering S. Endogenous calcium channels in human embryonic kidney HEK 293 cells. Br J Pharmacol 1996;118:748–54. Bleich M, Schlatter E, Greger R. The luminal K + channel of the thick ascending limb of Henle’s loop. Pflug Arch 1990;415:449 – 60. Brown PD, Greenwood SL, Robinson J, Boyd RD. Chloride channels of high conductance in the microvillous membrane of term human placenta. Placenta 1993;14:103–15. Brust P, Simerson FS, McCue AF, Deal CR, Schoonmaker S, Williams ME, Velicelebi G, Johnson EC, Harpold MM, Ellis SB. Human neuronal voltage-dependent calcium channels: Studies on subunit structure and role in channel assembly. Neuropharmacology 1993;32:1089 – 102. Buisson B, Gopalakrishnan M, Arneric SP, Sullivan JP, Bertrand D. Human a 4 b 2 neuronal nicotinic acetylcholine receptor in HEK 293 cells: A patch-clamp study. J Neurosci 1996;16:7880– 91.

Claudio T. Stable expression of heterologous multisubunit protein complexes established by calcium phosphate- or lipid-mediated cotransfection. Methods Enzymol 1992;207:391 – 408. Critz SD, Wible BA, Lopez HS, Brown AM. Stable expression and regulation of a rat brain K + channel. J Neurochem 1993;60:1175 – 8. Cummins TR, Zhou J, Sigworth FJ, Ukomadu C, Stephan M, Ptacek LJ, Agnew W S. Functional consequences of a Na + channel mutation causing hyperkalemic periodic paralysis. Neuron 1993;10:667 – 78. Fahlke C, Zachar E, Rudel R. Single-channel recordings of chloride currents in cultured human skeletal muscle. Pflug Arch 1992;421:108 – 16. Fletcher E, Nutt SL, Hoo KH, Elliott CE, Korczak B, McWhinnie EA, Kamboj RK. Cloning, expression and pharmacological characterization of a human glutamate receptor: hGluR4. Recept Channels 1995;3:21 – 31. Gogelein H. Chloride channels in epithelia. Biochim Biophys Acta 1988;947:521 – 47. Gopalakrishnan M, Buisson B, Touma E, Giordano T, Campbell JE, Hu IC, Donnelyroberts D, Arneric SP, Bertrand D, Sullivan JP. Stable expression and pharmacological properties of the human a 7 nicotinic acetylcholine-receptor. Eur J Pharmacol Molecular Pharmacology Section) 1995;290:237 – 46. Gopalakrishnan M, Monteggia LM, Anderson DJ, Molinari EJ, Piattoni-Kaplan M, Donnelly-Roberts D, Arneric SP, Sullivan JP. Stable expression, pharmacologic properties and regulation of the human neuronal nicotinic acetylcholine a 4 b 2 receptor. J Pharmacol Exp Ther 1996;276:289 – 97. Hamill OP, Marty A, Neher E, Sakmann B, Sigworth FJ. Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pflu¨g Arch 1981;391:85 – 100. Hoo KH, Nutt SL, Fletcher E, Elliott CE, Korczak B, Deverill RM, Rampersad V, Fantaske RP, Kamboj RK. Functional expression and pharmacological characterization of the human EAA4 GluR6) glutamate-receptor— a kainate selective channel subunit. Recept Channels 1994;2:327 – 37. Jiang C, Haddad GG. A direct mechanism for sensing low oxygen levels by central neurons. Proc Natl Acad Sci USA 1994;91:7198– 201. Jiang C, Haddad GG. Modulation of K + channels by intracellular ATP in human neocortical neurons. J Neurophysiol 1997;77:93– 107. Jiang C, Sigworth FJ, Haddad GG. Oxygen deprivation activates an ATP-inhibitable K + channel in substantia nigra neurons. J Neurosci 1994;14:5590 – 602. Kajita H, Kotera T, Shirakata Y, Ueda S, Okuma M, Oda-Ohmae K, Takimoto M, Urade Y, Okada Y. A maxi Cl − channel coupled to endothelin B receptors in the basolateral membrane of guineapig parietal cells. J Physiol Lond) 1995;488:65 – 75. Kohr G, Seeburg PH. Subtype-specific regulation of recombinant NMDA receptor-channels by protein tyrosine kinases of the src family. J Physiol Lond) 1996;492:445 – 52. Korpi ER, Kuner T, Kristo P, Kohler M, Herb A, Luddens H, Seeburg PH. Small N-terminal deletion by splicing in cerebellar a 6 subunit abolishes GABAA receptor function. J Neurochem 1994;63:1167 – 70. Marcus DC, Takeuchi S, Wangemann P. Two types of chloride channel in the basolateral membrane of vestibular dark cells. Hear Res 1993;69:124 – 32. McLarnon JG, Kim SU. Ion channels in cultured adult human Schwann cells. Glia 1991;4:534 – 9. Moran O, Conti F. Properties of the Kv1.1 rat brain potassium channels expressed in mammalian cells: Temperature effects. Biochem Biophys Res Commun 1995;215:915 – 20.

G. Zhu et al. / Journal of Neuroscience Methods 81 (1998) 73–83 Raymond LA, Moshaver A, Tingley WG, Huganir RL. Glutamate receptor ion channel properties predict vulnerability to cytotoxicity in a transfected nonneuronal cell line. Mol Cell Neurosci 1996;7:102 – 15. Shoemaker RL, Frizzell RA, Dwyer TM, Farley JM. Single chloride channel currents from canine tracheal epithelial cells. Biochim Biophys Acta 1986;858:235–42. Sigworth FJ, Sine SM. Data transformations for improved display and fitting of single-channel dwell time histograms. Biophys J 1987;52:1047 – 54. Swanson GT, Kamboj S, Kcull-Candy SG. Single-channel properties of recombinant AMPA receptors depend on RNA editing, splice variation, and subunit composition. J Neurosci 1997;17:58 – 69. Tseng-Crank J, Godinot N, Johansen TE, Ahring PK, Strobaek D, Mertz R, Foster CD, Olesen SP, Reinhart PH. Cloning, expres-

.

83

sion, and distribution of a Ca2 + -activated K + channel b-subunit from human brain. Proc Natl Acad Sci USA 1996;93:9200–5. Ukomadu C, Zhou J, Sigworth FJ, Agnew WS. muI Na + channels expressed transiently in human embryonic kidney cells: Biochemical and biophysical properties. Neuron 1992;8:663 – 76. Vaughan PC, French AS. Non-ligand-activated chloride channels of skeletal muscle and epithelia. Prog Biophys Mol Biol 1989;54:59– 79. Wang WH. Regulation of hyperpolarization-activated K + channel in the lateral membrane of the cortical collecting duct. J Gen Physiol 1995;106:25 – 43. Ye C, Rogers K, Bai M, Quinn SJ, Brown EM, Vassilev PM. Agonists of the Ca-sensitive receptor CaR) activate nonselective cation channels in HEK 293 cells stably transfected with the human CaR. Biochem Biophys Res Commun 1996;226:572–9.