The Journal of Immunology
Sustained Depolarization and ADP-Ribose Activate a Common Ionic Current in Rat Peritoneal Macrophages1 Brice Campo, Annmarie Surprenant, and R. Alan North2 Phagocytosis is associated with large changes in the membrane potential of macrophages, but the functional significance of this is unknown. Whole cell recordings were made from rat peritoneal macrophages. Sustained (>30 s) depolarization of the cells progressively activated a conductance that remained high (several nanoSeimens) for several tens of seconds. This current: 1) was linearly dependent on potential between ⴚ100 and ⴙ50 mV; 2) reversed close to 0 mV in a physiological external solution; 3) could also be carried in part by N-methyl-D-glucamine (PNMDG/PNa 0.7), chloride (PCl/PNa 0.5), or calcium (PCa/PNa 1.3); and 4) was blocked by intracellular ATP (5 mM) or ADP (10 mM) and by extracellular lanthanum (half-maximal concentration 1 mM). A current with all the same properties was recorded in cells when the intracellular solution contained ADP-ribose (10 –300 M) or -NAD (1 mM) (but not any other nucleotide analogs tested). The results suggest that prolonged depolarization leads to an increased intracellular level of ADP-ribose, which in turn activates this nonselective conductance(s). The Journal of Immunology, 2003, 170: 1167–1173.
T
he membrane potential of macrophages undergoes a prolonged depolarization following phagocytosis of an immune complex (1). This depolarization follows an increase in superoxide by a variable period, but typically tens of seconds, and it is subsequently maintained for several minutes. The underlying current shows little rectification and reverses close to 0 mV. The current is blocked by an inhibitor of NADPH-oxidase, and mimicked by xanthine and xanthine oxidase, leading to the conclusion that the depolarization results from free radical release within the cell following particle engulfment. This finding is in keeping with a body of work in a range of cells, which indicates that oxidative stressors, such as H2O2, can activate inward currents (2, 3). The capacity of such currents to deliver calcium to the cell has been considered to contribute to subsequent oxidative damage. Although intracellular calcium plays a key role in several aspects of macrophage function, including activation, migration, and proliferation as well as generation of superoxide anion and stimulation of NO production, calcium influx is not immediately required for phagocytosis (4 – 6). It is unlikely that the well-studied store-operated or capacitative calcium entry current, which is present in all nonexcitable cells, could underlie the current associated with phagocytosis in macrophage because it shows a characteristically strong rectification, the currents are very small (only a few pA at ⫺60 mV), and it is relatively impermeable to sodium under normal conditions (6 –11). Recently, a current has been described in the human macrophage U937 cell line that is activated by extracellular H2O2, or intracellular -NAD free acid (NAD⫹;3
Institute of Molecular Physiology, University of Sheffield, Sheffield, United Kingdom Received for publication July 9, 2002. Accepted for publication November 18, 2002. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1
This work was supported by the Wellcome Trust.
2
Address correspondence and reprint requests to Dr. R. Alan North, Institute of Molecular Physiology, Alfred Denny Building, Western Bank, University of Sheffield, Sheffield S10 2TN, U.K. E-mail address:
[email protected] Abbreviations used in this paper: NAD⫹, -NAD free acid; NADH, -NAD disodium salt; NADP⫹, -NAD phosphate disodium salt; NADPH, -NADPH disodium salt; NMDG, (N-methyl-D-glucamine); SBFI, 1,3-benzenedicarboxylic acid, 4,4⬘-[1,4,10trioxa-7,13-diazacyclopentadecane-7,13-diylbis(5-methoxy-6,2-benzofurandiyl)]bis-, tetra-ammonium salt; SKF96365, (1-[b-[3-(4-methoxyphenyl)propoxy]-4-methoxyphenethyl]-1H-imidazole hydrochloride.
1 mM) (12, 13) and might therefore underlie the current following phagocytosis described by Holevinsky and Nelson (1). This current is also activated by intracellular ADP-ribose at concentrations (10 –300 M) lower than those required for NAD⫹. It corresponds in some of its properties to the currents observed with heterologous expression of TRPM2 subunits (formerly called LTRPC2; for nomenclature, see Ref. 14) (12, 13, 15). These results are consistent with the interpretation that the prolonged depolarization that follows phagocytosis described by Holevinsky and Nelson (1) results from an inward current, such as that activated by NAD⫹ in U937 cells (12, 13). However, during the course of voltage-clamp recordings from rat peritoneal macrophages, we noticed that a sustained inward current could actually be activated by prolonged depolarization. In this work, we describe some properties of this current, and compare these with the current activated by NAD⫹ and ADP-ribose in the same cells. Both currents showed the same permeability to sodium, calcium, N-methylD-glucamine (NMDG), chloride, and gluconate ions, and sensitivity to block by lanthanum. The current activated by intracellular ADP-ribose occluded the depolarization-activated current. Thus, we conclude it is likely that these two stimuli activate the same conductance.
Materials and Methods Culture of rat peritoneal macrophages Rat peritoneal cells were obtained from male Wistar rats (10 wk old) killed by cervical dislocation. The peritoneal cavity was lavaged with PBS (Life Technologies Cell Culture, Invitrogen, Paisley, U.K.). After centrifugation, the macrophage fraction was separated from the mast cells by fluorescence-activated cell sorting. Two populations of cells were separated by forward and side scatter. Cells were then collected by centrifugation and resuspended in DMEM (Life Technologies Cell Culture) containing 10% (v/v) FCS (Sigma-Aldrich, Poole, U.K.), 1% glutamine (Calbiochem, La Jolla, CA), and 1% antibiotics (Life Technologies Cell Culture, Invitrogen). Cells were plated onto glass coverslips and kept at 37°C in a humidified atmosphere containing 5% CO2. The cultured macrophages were used in patch-clamp experiments 1–3 days after plating. Cells were identified as macrophages using immunohistochemistry staining with the macrophage specifics Abs ED1–10 and ED-2 (Serotec, Oxford, U.K.).
HEK293 cells expressing P2X2/3 receptors
3
Copyright © 2003 by The American Association of Immunologists, Inc.
HEK293 cells were transfected, as previously described (16). Briefly, cells were electroporated (Gene Pulser; Bio-Rad, Hercules, CA) in the presence of 30 mg of plasmid containing both P2X2 and P2X3 receptor cDNAs. Following electroporation, cells were plated and selected for neomycin resistance, subcloned by FACS sorting, and resuspended in standard 0022-1767/03/$02.00
1168 growth medium (Dulbecco’s) supplemented with 10% heat-inactivated FCS and 2 mM L-glutamine. Cells were grown to confluency and plated on coverslips, and recordings were made 2– 48 h later.
Whole cell recording Ion currents were studied using the whole cell configuration of the patchclamp technique, as described by Hamill et al. (17). Glass electrodes (Harvard Apparatus, Edenbridge, U.K.) ranged from 6 to 8 M⍀ in resistance and were fire polished. The indifferent electrode was an Ag-AgCl electrode connected to the bath through a short 3 M KCl agar bridge. All the experiments were performed at room temperature using an EPC9 patch-clamp system (HEKA, Lambrecht, Germany). The cells were kept in an extracellular solution containing (mM): NaCl, 147; KCl, 2; MgCl2, 1; CaCl2, 2; HEPES, 10; and glucose, 13. The pH was adjusted to 7.3 by adding NaOH (4 mM). The standard intracellular solution contained (mM): NaCl, 148; HEPES, 10; EGTA, 10. We used intracellular sodium rather than potassium in most experiments so that voltage-dependent potassium currents did not complicate our estimates of reversal potential. We used a high EGTA concentration so as to reduce any involvement of store-operated conductances in those cases in which we had calcium-free extracellular solutions. The pH was adjusted to 7.3 by adding NaOH (24 mM). Drug application and solution exchanges were conducted using a fast flow RSC 200 system (Biologic Science Instruments, Claix, France). Cells were held at ⫺80 mV, except where stated, and current-voltage plots were obtained from ramp commands (⫺130 to 50 mV in 500 ms). Permeability experiments were conducted using standard intracellular solution and extracellular solution of different composition. The low NaCl solution contained (mM): NaCl, 40; HEPES, 10; CaCl2, 0.3 mM; glucose, 236 (pH brought to 7.3 with 4.2 mM NaOH). The NMDG solution contained (mM): NMDG, base 154; HEPES, 10; glucose, 23 (pH brought to 7.3 with HCl; final Cl concentration 146 mM). The gluconate solution contained (mM): Na-gluconate, 147; HEPES, 10; glucose, 33 (pH brought to 7.3 with 4 mM NaOH). The isotonic calcium solution contained (mM): CaCl2, 110; Ca(OH)2, 2; HEPES, 10; glucose, 12 (pH 7.3). Series resistance was compensated by ⬃75%. Liquid junction potentials were calculated directly from ionic mobilities (18). Measured potentials were corrected for the junction potential existing at the recording pipette before seal formation (pipette solution ⫺1.9 mV with respect to bath solution) and for any change in junction potential at the indifferent electrode when the bath solution was changed (⬍4 mV in all cases).
Intracellular sodium measurement Cells were loaded with 1,3-benzenedicarboxylic acid, 4,4⬘-[1,4,10-trioxa7,13-diazacyclopentadecane-7,13-diylbis(5-methoxy-6,2-benzofurandiyl)]bis-, tetra-ammonium salt (SBFI, 10 mM; Molecular Probes, Eugene, OR), and 0.02% pluronic acid for 45 min at 37°C. Ratiometric fluorescence measurements were performed using an Axiovert 100 microscope (Zeiss, Oberkochen, Germany). Cells were excited at 340 and 380 nm using a computer-driven monochromator, and the emission fluorescence (510 nm) was collected through a bandpass filter onto a charge-coupled device camera. Data were analyzed with Photonics software (Planegg, Germany).
Measurements of permeability We first measured the permeability of chloride relative to that of sodium (PCl/PNa) in bi-ionic conditions. PCl/PNa was then calculated from PCl/PNa ⫽ ((1 ⫺ exp(⫺x)) ⫻ ([Na]o ⫺ [Na]i exp(x)))/(([Cl]i exp(⫺x) ⫺ [Cl]o) ⫻ (1 ⫺ exp(x))) (equation 1), where x ⫽ Er F/RT (extracellular 0.3 mM calcium ignored), where Er is the reversal potential, R is the Gas constant, F is the Faraday, and T is the absolute temperature. We used this value of PCl/PNa in subsequent calculations in which three permeant ions were present. The expressions used were: PNMDG/PNa ⫽ ((([Na]i exp(x) ⫺ [Na]o)/(1 ⫺ exp(x))) ⫹ PCl/PNa ⫻ (([Cl]i exp(⫺x) ⫺ [Cl]o)/(1 ⫺ exp(⫺x))))/(([NMDG]o ⫺ [NMDG]i exp(x))/(1 ⫺ exp(x))) (equation 2); Pgluc/PNa ⫽ ((([Na]i exp(x) ⫺ [Na]o)/(1 ⫺ exp(x))) ⫹ PCl/PNa ⫻ (([Cl]i exp(⫺x) ⫺ [Cl]o)/(1 ⫺ exp(⫺x))))/(([gluconate]o ⫺ [gluconate]i exp(⫺x))/ (1 ⫺ exp(⫺x))) (equation 3); and PCa/PNa ⫽ (([Na]i exp(x)/(1 ⫺ exp(x))) ⫺ PCl/PNa ⫻ (([Cl]o ⫺ [Cl]iexp(⫺x))/(1 ⫺ exp(⫺x))))/(4 [Ca]o/(1 ⫺ exp(2x)) (equation 4). Although not shown in these equations, we have multiplied ion concentrations by the following activity coefficients: Na⫹, 0.8; K⫹, 0.8; Cl⫺, 0.8; NMDG⫹, 0.8; gluconate⫺, 0.8; Ca2⫹, 0.28 (19).
Data analysis The results shown in the figures are mean ⫾ SEM. Curve fits of the pooled data were performed using Kaleidagraph (Synergy Software, Reading, PA) and Excel (Microsoft, Redmond, WA) software. The lanthanum inhibition curve was fitted with I/Imax ⫽ 100 (IC50n/(IC50n ⫹ [La]n)), where I is the
IONIC CURRENTS IN MACROPHAGES current at a given lanthanum concentration expressed as percentage of the current in the absence of lanthanum (Imax), IC50 is the lanthanum concentration at which I ⫽ 0.5 ⫻ Imax, and n is the Hill coefficient. Concentrationcurrent curves were fitted with I/Imax ⫽ 100 ([ADP-ribose]n/([ADP-ribose]n ⫹ EC50n)), where I is the current activated by a given concentration of ADP-ribose and expressed as percentage of the maximal currents (Imax), EC50 is the ADP-ribose concentration evoking half of the maximal current, and n is the Hill coefficient.
Chemicals ADP, ADP-glucose, ADP-ribose, cyclic ADP-ribose, AMP, ATP, EGTA, lanthanum chloride, D-ribose, D-ribose 5-phosphate, NMDG, xanthine, and xanthine oxidase were purchased from Sigma-Aldrich. NAD⫹, -NAD disodium salt (NADH), -NAD phosphate disodium salt (NADP⫹), -NADPH disodium salt (NADPH), and 1-[b-[3-(4-methoxyphenyl)propoxy]-4-methoxyphenethyl]-1H-imidazole (hydrochloride; SKF96365) were purchased from Calbiochem. SBFI was from Molecular Probes.
Results Macrophages had a capacitance of 11.4 ⫾ 0.4 pF (n ⫽ 110). The cell diameter was 9.1 ⫾ 0.2 m (n ⫽ 30). Step depolarizations from a holding potential of ⫺80 mV (10 mV increments to ⫹40 mV, 500 ms duration) revealed no time-dependent currents when a sodium chloride-based internal solution was used. When a potassium chloride-based solution was used (150 mM KCl, 10 mM HEPES, 10 mM EGTA), the resting potential of the cells was ⫺30.9 ⫾ 0.7 mV (n ⫽ 11). Current activated by depolarization Cells held at ⫺80 mV showed no change in holding current or conductance during 5–10 min of whole cell recording. However, if the cell was held at ⫹10 mV for several tens of seconds and then returned to ⫺80 mV, a large inward current was observed (Fig. 1A). This current quickly (⬍5 s) reached its peak amplitude (several hundred pA) and then declined during the next few minutes (Fig. 1A; in 10 cells the amplitude after 2 min was 19 ⫾ 4% of the peak amplitude). The duration of the depolarization required to evoke the subsequent inward current was not studied in detail; however, a depolarization to ⫹10 mV for only 10 s was not effective (n ⫽ 5). After an interval of several minutes, a second depolarization again elicited a large inward current with similar properties (n ⫽ 10). The inward current was associated with a large increase in cell conductance. The current-voltage relation of the cell remained essentially linear, and reversed close to 0 mV (⫺3.1 ⫾ 0.24 mV, n ⫽ 18) in normal external solution (Fig. 1, B and C) (Table I). The time course of the conductance increase was followed by holding the cell at ⫹10 mV, but repeatedly applying brief ramp voltage commands (⫺130 to ⫹50 mV in 500 ms; 5 s intervals). Fig. 1C shows the time course of the resulting conductance increase. The current (at ⫺80 mV) grew more slowly and was much smaller with a sustained holding potential of ⫺40 mV (Fig. 1D), and it was larger with depolarization to ⫹40 mV. In low chloride extracellular solution (40 mM NaCl), the current at ⫺80 mV was reduced to 51 ⫾ 16% (n ⫽ 3) of its control value; the residual current reversed polarity at ⫺13.1 ⫾ 1.6 mV (n ⫽ 4) (Fig. 2A); from equation 1 this corresponds to PCl/PNa of ⬃0.5 (Table I). When extracellular sodium was substituted by NMDG, the current evoked by depolarization was reduced to 79 ⫾ 2% (n ⫽ 6) of its control value (measured at ⫺80 mV). The residual current reversed polarity at ⫺5.4 ⫾ 0.45 mV (n ⫽ 7) (Table I; Fig. 2A), which indicates that NMDG is significantly permeable (PNMDG/ PNa 0.74; from equation 2). When extracellular gluconate replaced chloride, the current at ⫺80 mV was reduced to 91 ⫾ 4% (n ⫽ 4), and when extracellular calcium replaced sodium, the inhibition was to 48 ⫾ 6% (n ⫽ 6) of control (Fig. 2B). By measuring the reversal potential of the residual current (Table I), it was evident
The Journal of Immunology
1169
FIGURE 1. Current activated by depolarization of cell membrane. A, Example in one cell of current activated by sustained depolarization (⫹10 mV). When the holding potential is returned to ⫺80 mV after 30 s depolarization, a large inward current develops and then progressively declines during the next minutes. The current remaining after 90 s is quickly and completely blocked by lanthanum (10 M). B, Current-voltage plots at the times indicated in A, before, during, and after activation of the current. C, Current-voltage plots recorded at different times after moving to a holding potential of ⫹10 mV. A voltage ramp (⫺130 to 50 mV in 500 ms) was applied at 5 s intervals (time indicated beside selected traces). D, Time course of development of current at ⫺80 mV, taken from experiments such as those in C. A large current was evoked by holding at ⫹10 mV (F); this was smaller when the potential was held at ⫺40 mV (Œ), and did not occur with a holding potential of ⫺80 mV (E). Points are mean amplitude (⫾SEM) for eight (⫹10 mV), three (⫺40 mV), and five (⫺80 mV) cells.
that the conductance activated by prolonged depolarization is significantly permeable to chloride and gluconate, NMDG, and calcium ions. In each of these ion substitutions, the current-voltage relation remained essentially linear (Fig. 2). We conducted control measurements of the permeability of P2X receptors expressed in HEK293 cells, using the same intracellular and extracellular solutions. We used the agonist ␣meATP (30 M) to activate the inward current in cells expressing P2X2/3 subunits, which was typically ⬃1 nA in amplitude (20). We observed that the agonist-induced current reversed polarity at 0.8 ⫾ 0.5 mV (n ⫽ 23) in control extracellular solution, ⫺33 ⫾ 1.4 mV (n ⫽ 4) in 40 mM NaCl, and ⫺86.5 ⫾ 0.4 mV (n ⫽ 4) in NMDG solution. These reversal potentials correspond (from equations 1 and 2) to relative permeabilities of PCl/PNa ⬍ 0.02 and PNMDG/PNa ⬍ 0.03 (see Ref. 21).
The current induced by depolarization was substantially blocked by lanthanum (10 M) (Fig. 1); the IC50 for this action was ⬃1 M (Fig. 3A). The current was also significantly inhibited (76 ⫾ 10%, n ⫽ 4) by SKF96365 (100 M); at 10 M SKF96365 inhibited the current in some (75 ⫾ 8%, n ⫽ 4), but not other (n ⫽ 3) cells. SKF96365 has been widely used to block nonselective cation channels and store-operated calcium entry in nonexcitable cells since its introduction by Merritt et al. (22). The current evoked by depolarization was also greatly reduced by including in the pipette either ATP (5 mM, inhibited by 89 ⫾ 5%, n ⫽ 4) or ADP (10 mM, inhibited by 97 ⫾ 4%, n ⫽ 5) (Fig. 3B). Current activated by ADP-ribose An inward current progressively developed over tens of seconds in cells held at ⫺80 mV, when the recording electrode contained
Table I. Reversal potentials and relative permeabilities for two macrophage currentsa Depolarization
Control Low chloride NMDG Gluconate Isotonic calcium a
ADP-Ribose
Remaining Current (%)
Erev (mV)
PX/PNa
Remaining Current (%)
Erev (mV)
PX/PNa
51 ⫾ 16 (3) 79 ⫾ 2 (6) 91 ⫾ 4 (4) 48 ⫾ 6 (6)
⫺3.1 ⫾ 0.24 (18) ⫺13.1 ⫾ 1.6 (4) ⫺5.4 ⫾ 0.45 (7) ⫺8.8 ⫾ 0.6 (6) 0.20 ⫾ 0.50 (6)
0.47 0.74 0.96 1.33
69 ⫾ 2 (4) 75 ⫾ 2 (4) 84 ⫾ 4 (4) 42 ⫾ 6 (3)
⫺2.7 ⫾ 0.65 (11) ⫺13.5 ⫾ 1.2 (4) ⫺4.8 ⫾ 0.35 (4) ⫺8.3 ⫾ 0.34 (4) 0.36 ⫾ 1.6 (5)
0.45 0.84 0.90 1.15
Mean ⫾ SEM for number of cells in parentheses.
1170
FIGURE 2. Relative permeability of conductance activated by prolonged depolarization. A, Current-voltage plots (ramp from ⫺130 to 50 mV in 500 ms) in control solution, and in low (40 mM) sodium chloride, and NMDG-chloride solutions. Reversal potentials are ⫹2, ⫺11, and ⫺5 mV (see enlarged inset). B, Current-voltage plot from a different cell shows no change in reversal potential when external solution is changed to isotonic calcium chloride. Reversal potentials in this cell are 0 and ⫺2 mV for control and isotonic calcium, respectively.
ADP-ribose (Fig. 4A). The current-voltage relation remained linear throughout this period, and reversed close to 0 mV (⫺2.7 ⫾ 0.6 mV, n ⫽ 11) (Fig. 4B). With 1 M ADP-ribose, the development of the current was fitted by an exponential with a time constant of ⬃1 min (Fig. 4C). The lowest concentration of ADP-ribose that was effective was 100 nM, and maximum value of the current was reached with ⬃100 M; the EC50 was ⬃10 M (Fig. 4D). Changing the level of intracellular calcium buffering did not obviously alter the effectiveness of ADP-ribose to activate this current. In our control, strongly buffered solution (10 mM EGTA, 0 mM calcium, ⬃1 nM [Ca]i) ADP-ribose (300 M) elicited a current of 703 ⫾ 119 pA (n ⫽ 3); with a higher level of intracellular calcium (11 mM EGTA, 2 mM calcium, ⬃100 nM [Ca]i), this was 811 ⫾ 149 pA (n ⫽ 5). No similar current developed when the recording electrode contained ADP (100 M), ADP-glucose (100 M), cyclic ADP-ribose (100 M), AMP (100 M), ATP magnesium salt (100 M), D-ribose (100 M), D-ribose 5-phosphate (100 M), NADH (300 M or 1 mM), NADP⫹ (300 M), or NADPH (300 M) (n ⫽ 3 cells in all cases). However, NAD⫹ at a concentration of 1 mM (n ⫽ 4) evoked a current similar to that observed with ADP-ribose, although 100 and 300 M were ineffective. The peak current elicited by NAD⫹ (1 mM) was 758 ⫾ 428 pA (n ⫽ 4), which is not different from that seen with ADP-ribose (Fig. 4). The relative permeability of the current induced by ADP-ribose (300 M) was examined by measuring the reversal potential in different extracellular solutions. From the values given in Table I, it can be seen that the relative permeability to chloride, NMDG,
IONIC CURRENTS IN MACROPHAGES
FIGURE 3. Similar inhibition of depolarization activated and ADP-ribose-activated currents. A, Dose-response curves for inhibition by lanthanum of the currents in macrophages evoked by depolarization (F, n ⫽ 3) and ADP-ribose (E, n ⫽ 4). Recordings made at ⫺80 mV. (No difference between depolarization and ADP-ribose: ANOVA, p ⬎ 0.05.) B, Currents evoked by depolarization (filled bars) and by intracellular ADP-ribose (open bars) have similar sensitivities to inhibition by intracellular ATP (5 mM) and ADP (10 mM). (No difference between depolarization and ADPribose: ANOVA, p ⬎ 0.05.)
gluconate, and calcium ions did not differ markedly from those determined for the currents elicited by depolarization. The current evoked by ADP-ribose was also very sensitive to block by extracellular lanthanum (Fig. 3A), with an IC50 of ⬃10 M. It was also blocked by SKF96365 (100 M, 77 ⫾ 4%, 5 of 5 cells tested; 10 M, 70 ⫾ 7%, 7 of 12 cells tested) and by intracellular either ATP (5 mM, inhibited by 66 ⫾ 11%, n ⫽ 5) or ADP (10 mM, inhibited by 76 ⫾ 3%, n ⫽ 4) (Fig. 3B). These relative inhibitions were not different from those observed for the current activated by depolarization (ANOVA, p ⬎ 0.05). A membrane current with similar properties was also elicited by applying xanthine (0.1 mM) and xanthine oxidase (0.05 U/ml), to generate superoxide anions; this current was completely blocked by 100 M lanthanum (Fig. 5A). Application of xanthine alone had no effect. In parallel experiments, we used intracellular sodium imaging as an index of depolarization of undisturbed macrophages. In these cells, H2O2 (10 mM) evoked a progressive increase in intracellular sodium concentration ([Na]i) (Fig. 5B). The time course of the [Na]i increase was similar to that of the inward current observed with xanthine/xanthine oxidase, or depolarization, and the increase in [Na]i was also strongly reduced by lanthanum (1 mM) (Fig. 5B). Occlusion of the two currents The similar properties of the two currents led us to hypothesize that sustained depolarization and intracellular ADP-ribose were
The Journal of Immunology
1171
FIGURE 4. Current activated by intracellular ADP-ribose. A, A large inward current progressively develops during the first 50 s of recording with a pipette containing 300 M ADP-ribose. At the periods indicated, the superfusion solution was changed to NMDG, isotonic calcium, or lanthanum. B, Current-voltage plots showing the progressive increase in conductance after beginning whole cell recording with an electrode containing ADP-ribose (300 M). Voltage ramps were ⫺130 to ⫹50 mV in 500 ms, repeated every 5 s. Times indicated beside selected traces. C, Time course of development of current with different ADP-ribose concentrations. Control, no ADP-ribose (E, n ⫽ 4); F, 100 nM (n ⫽ 6); 䡺, 1 M (n ⫽ 3); f, 100 M (n ⫽ 6); and ‚, 300 M (n ⫽ 6). The currents are normalized to the value at 240 s in each particular cell; these values were 66 ⫾ 25 pA (4), 309 ⫾ 143 pA (6), 356 ⫾ 145 pA (3), 949 ⫾ 390 pA (6), and 1521 ⫾ 432 pA (6) for control, 100 nM, 1 M, 100 M, and 300 M, respectively. The numbers beside each trace are the time constants of fitted exponentials. Holding potential ⫺80 mV. D, Concentration-response curve for current measured at 240 s as a function of intracellular ADP-ribose concentration (n ⫽ 3– 6 cells for each point).
activating the same conductance. We recorded from 10 cells with electrodes containing ADP-ribose (300 M); after 4 min, when the inward current reached its steady value, we applied a depolarization to ⫹10 mV for an additional 4 min. This depolarization evoked no additional current. In time-matched controls, the depolarization elicited large inward currents recorded at ⫺80 mV (Fig. 6).
Discussion We found that sustained depolarization of rat peritoneal macrophages activates a large inward current. The current is carried by a conductance that is nonselective for cations, and that also has quite significant permeability to chloride and gluconate ions. The permeability to chloride ions was surprising, and for this reason we compared the results directly with those for the permeability of another cation-selective membrane channel. In the low chloride solutions that we have used ([Na]i ⫽ 172 mM, [Cl]i ⫽ 148 mM, [Na]o ⫽ 44 mM, [Cl]o ⫽ 41 mM), the observed reversal potential for the P2X2/3 receptor corresponds to a very low chloride permeability (PCl/PNa of ⬃0.02). The estimates of permeability to NMDG, calcium, and gluconate should be considered as less reliable than those for PCl/PNa. This is because in those cases there are three permeant ions present, and the estimate involves using the value previously determined for PCl/PNa. Likewise, the reversal potential observed with extracellular NMDG indicates that the P2X2/3 receptor is essentially impermeant (PNMDG/ PNa ⬍ 0.03), whereas both the currents presently studied in macrophages have significant permeability to NMDG. From experiments of the type described, it is difficult to exclude the possibility that the sustained depolarization is activating more than one set of ionic channels; for example, one permeable to anions and another permeable to cations. All that can be said in this case is that each has a very similar, and linear, current-voltage relation and each has a similar sensitivity
to block by lanthanum. We shall return later to discuss the question of how a sustained depolarization might lead to the activation of such an ionic conductance(s). There are many similarities between these permeation properties and those reported by Holevinsky and Nelson (1) for the current following phagocytosis of fluorescein-labeled immune complexes recorded from macrophages derived from human blood monocytes. That was a large current (typically ⬃1 nA), similar in amplitude to the currents that we presently describe; this is in marked contrast to the very small (⬃50 pA) current elicited by thapsigargin in the same cells (10). The underlying conductance had PNMDG/PNa of 0.69 in the work of Holevinsky and Nelson (1), which is close to our present estimate. Furthermore, they also reported that the underlying channels had significant anion permeability, and that glutamate appeared to be more permeable than chloride. That current was also completely blocked by 1 mM lanthanum, although the effects of lower concentrations were not examined. Holevinsky and Nelson (1) found that a current with similar properties could be activated by xanthine plus xanthine oxidase, although not by xanthine alone, and they suggested that the release of free radicals following phagocytosis activated the underlying conductance. Holevinsky and Nelson (1) did not investigate whether a sustained depolarization could elicit such a current, as we have reported; they did report that, under current clamp, the current led to a sustained depolarization of the cells. It has recently become clear that oxidants such as H2O2 can indeed activate a cation conductance in certain immune cells (12). In the monocytic cell line U937, as well as Jurkat T lymphocytes and the eosinophil line EOL1, this can be mimicked by intracellular NAD⫹, but not by NADH (12, 13). This implies that NAD⫹, formed by the action of free radicals on intracellular
1172
IONIC CURRENTS IN MACROPHAGES
FIGURE 5. Current evoked by superoxide anions. A, Superfusion with xanthine (1 mM) and xanthine oxidase (0.05 U/ml) evoked an inward current that developed over several tens of seconds (n ⫽ 6). This was rapidly and completely blocked by lanthanum (100 M). After washout of lanthanum, depolarization evoked the usual current in the same cells. B, Changes in intracellular sodium were measured as SBFI fluorescence. Application of H2O2 induced a slow rise in [Na]i (E) that was strongly reduced by lanthanum (1 mM; f). Control cells (F, no H2O2 added) showed only a small rise in [Na]i during the same time period. Values are mean ⫾ SEM for 10 cells.
FIGURE 6. Occlusion of currents activated by depolarization and by ADP-ribose. A, Development of the current in 10 macrophages held at ⫺80 mV for 240 s (F, filled bar); recording electrode contained ADP-ribose (300 M). At 240 s, the holding potential was changed to ⫹10 mV (E, open bars); the current at ⫺80 mV was determined from voltage command ramps. Depolarization to ⫹10 mV did not result in the development of any additional current. B, Time-matched control recordings in which the electrode did not contain ADP-ribose. Depolarization to ⫹10 mV beginning at 240 s resulted in the development of an inward current (recorded at ⫺80 mV) similar to that illustrated in Fig. 1.
NADH, mediates the effect. The molecular identity of the underlying channel has been approached in two ways. First, heterologous expression of TRPM2 cDNA results in the appearance of channels permeable to cations that can be activated by intracellular NAD⫹ (⫽1 mM) (12, 13). Second, treatment of U937 monocytes with antisense oligonucleotides that reduced the expression of TRPM2 (measured by Western blotting and immunocytochemistry) also reduced the calcium entry and cell death elicited by H2O2 (12). These experiments strongly suggest that one or more TRPM2 subunits are a key component of the channel expressed by these immune cell lines that is activated by oxidative damage (12, 13). In contrast, it has recently been found that H2O2 can activate heterologously expressed TRPM2 channels by a mechanism that appears not to involve changes in ADP-ribose or NAD levels (23). The homomeric channels formed by expression of TRPM2 subunits are directly activated also by ADP-ribose (13, 15), although not by several closely related analogs (cyclic ADP-ribose, NADH, NADP⫹, NADPH, ATP, and ADP) (12, 13, 15). The activation results from binding of the ADP-ribose to a Nudix domain in the C terminus of the protein, because it is prevented by point mutations of key residues in this motif (12, 15). The concentration of ADP-ribose causing 50% activation of the whole cell
current is ⬃30 –100 M (15). When activated by intracellular diffusion of ADP-ribose during the course of whole cell recordings, these TRPM2 currents show close to linear I/V relations and others properties that resemble the current first described by Wilding et al. (24) in ascidian oocytes. The current that we now report in rat peritoneal macrophages, which is activated by ADP-ribose, has some, but not all, of the same features. These include the effectiveness of ADP-ribose and NAD⫹ (although not a range of other structurally related compounds), the effective concentrations of ADP-ribose (EC50 ⬃30 M; Fig. 4D) and NAD⫹ (1 mM), and the block by ATP (5 mM) (13). We considered that an excess of ADP might inhibit the binding of ADP-ribose to its Nudix domain (25), and consistent with this we observed that intracellular ADP also blocked the currents activated by both ADP-ribose and depolarization. However, there also appear to be important differences between the properties of heterologously expressed TRPM2 channels and the macrophage currents. The most notable is the significant permeability to NMDG, chloride, and gluconate that we observed, whereas substitution of external cations by NMDG almost completely blocks the current through homomeric TRPM2 channels (13, 15). Furthermore, we observed a relatively high sensitivity to lanthanum (IC50 ⬇
The Journal of Immunology 1–10 M) on the depolarization-evoked and ADP-ribose-evoked current in peritoneal macrophages, but the sensitivity of TRPM2 currents to lanthanum has not yet been studied systematically. In contrast, other currents have been described with properties that more closely resemble the macrophage currents observed in the present study. For example, heterologous expression of PKD1 and PKD2, genes named for the polycystic kidney disease that arises when they are mutated in humans (26), gives rise to a similar conductance (27). Those channels have PNMDG/PNa of 0.51. The PCl/PNa cannot be calculated directly from the data of Hanaoka et al. (27) (they had three permeant ions present: cesium, sodium, and chloride); however, the 45 mV shift in reversal potential would indicate PCl/PNa of ⬃0.3. The lanthanum sensitivity of the PKD1/2 current (50% inhibition by ⬃50 M) is also not very different from that which we observed in peritoneal macrophages. It is not clear how polycystin channels could be activated by NAD⫹ and ADPribose, because their amino acid sequence does not contain a Nudix domain. Further work, such as recordings of single channel activity and antisense oligonucleotide approaches, will be needed to establish the molecular composition of the macrophage channel. The occlusion experiment illustrated in Fig. 5 strongly suggests that the two currents that we describe pass through the same channels (although it is also formally possible that intracellular ADPribose could in some way inhibit the effect of depolarization to activate a separate population of channels). This is consistent with the overall similarity in the properties of the currents with respect to permeability (Table I) and lanthanum blockade (Fig. 3). This raises the question deferred earlier: what might be the mechanism by which sustained depolarization activates this cell conductance? Speculations might include that depolarization evokes the release of ADP-ribose from intracellular stores, or in some way facilitates its synthesis from NAD⫹. In the latter regard, it is worth noting that the enzyme responsible for the formation of ADP-ribose from NAD⫹ (CD38, NAD-glycohydrolase) is itself membrane associated, with its enzymatic activity in the ectodomain (28, 29). It is possible that the movement of negatively charged ADP-ribose into the cell is in some way facilitated by the sustained applied depolarization. Our observation that the current elicited by depolarization and the current evoked by intracellular ADP-ribose had very similar sensitivities to blockade by intracellular ATP and ADP (Fig. 3B) is obviously consistent with the interpretation that ADPribose serves as the messenger for the depolarization. In this case, as with the experiments using xanthine and xanthine oxidase, it must be argued the involvement of ADP-ribose is local rather than distributed throughout the cytoplasm; this is because the responses were well maintained in cells during relatively long periods of intracellular dialysis with the contents of the recording pipette. In conclusion, an imposed depolarization of macrophages activates an inward current that is significantly permeable to large cations as well as anions, and that is blocked by low M concentrations of lanthanum. Under more physiological circumstances (i.e., without voltage-clamp), this current could amplify the depolarization by opening further channels. In this way, the effect of an initial stimulus that leads to the release of free radicals within the cell (such as phagocytosis) would be further enhanced. Such a positive feedback process, in which activation of the current leads to depolarization of the macrophage and depolarization leads to further activation, might contribute to the eventual cell death that readily follows exposure to exogenous oxidants.
Acknowledgments We thank Daniele Estoppey and Gareth Evans for assistance with cell culture, and Claire Fenech for advice on sodium imaging.
1173
References 1. Holevinsky, K. O., and D. J. Nelson. 1995. Simultaneous detection of free radical release and membrane current during phagocytosis. J. Biol. Chem. 270:8328. 2. Herson, P. S., K. Lee, R. D. Pinnock, J. Hughes, and M. L. Ashford. 1999. Hydrogen peroxide induces intracellular calcium overload by activation of a nonselective cation channel in an insulin-secreting cell line. J. Biol. Chem. 274:833. 3. Coyle, J. T., and P. Puttfarcken. 1993. Oxidative stress, glutamate, and neurodegenerative disorders. Science 262:689. 4. Edberg, J. C., C. T. Lin, D. Lau, J. C. Unkeless, and R. P. Kimberly. 1995. The Ca2⫹ dependence of human Fc␥ receptor-initiated phagocytosis. J. Biol. Chem. 270:22301. 5. Krause, K. H., K. P. Campbell, M. J. Welsh, and D. P. Lew. 1990. The calcium signal and neutrophil activation. Clin. Biochem. 23:159. 6. Li, S. W., J. Westwick, and C. T. Poll. 2002. Receptor-operated Ca2⫹ influx channels in leukocytes: a therapeutic target? Trends Pharmacol. Sci. 23:63. 7. Lewis, R. S., and M. D. Cahalan. 1995. Potassium and calcium channels in lymphocytes. Annu. Rev. Immunol. 13:623. 8. Parekh, A. B., and R. Penner. 1997. Store depletion and calcium influx. Physiol. Rev. 77:901. 9. Putney, J. W., Jr., L. M. Broad, F. J. Braun, J. P. Lievremont, and G. S. Bird. 2001. Mechanisms of capacitative calcium entry. J. Cell Sci. 114:2223. 10. Malayev, A., and D. J. Nelson. 1995. Extracellular pH modulates the Ca2⫹ current activated by depletion of intracellular Ca2⫹ stores in human macrophages. J. Membr. Biol. 146:101. 11. Floto, R. A., M. P. Mahaut-Smith, J. M Allen, and B. Somasundaram. 1996. Differentiation of the human monocytic cell line U937 results in an up-regulation of the calcium release-activated current, ICRAC. J. Physiol. 495:331. 12. Hara, Y., M. Wakamori, M. Ishii, E. Maeno, M. Nishida, T. Yoshida, H. Yamada, S. Shimizu, E. Mori, J. Kudoh, et al. 2002. LTRPC2 Ca2⫹-permeable channel activated by changes in redox status confers susceptibility to cell death. Molec. Cell 9:163. 13. Sano, Y., K. Inamura, A. Miyake, S. Mochizuki, H., Yokoi, H. Matsushime, and K. Furuichi. 2001. Immunocyte Ca2⫹ influx system mediated by LTRPC2. Science 293:132. 14. Montell, C., L. Birnbaumer, V. Flockerzi, R. J. Bindels, E. A. Bruford, M. J. Caterina, D. E. Clapham, C. Harteneck, S. Heller, D. Julius, et al. 2002. A unified nomenclature for the superfamily of TRP cation channels. Molec. Cell 9:229. 15. Perraud, A. L., A. Fleig, C. A. Dunn, L. A. Bagley, P. Launay, C. Schmitz, A. J. Stokes, Q. Zhu, M. J. Bessman, R. Penner, et al. 2001. ADP-ribose gating of the calcium-permeable LTRPC2 channel revealed by Nudix motif homology. Nature 411:595. 16. Kawashima, E., D. Estoppey, C. Virginio, D. Fahmi, S. Rees, A. Surprenant, and R. A. North. 1998. A novel and efficient method for the stable expression of heteromeric ion channels in mammalian cells. Recept. Channels 5:53. 17. Hamill, O. P., A. Marty, E. Neher, B. Sakmann, and F. J. Sigworth. 1981. Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pflugers Arch. 391:85. 18. Barry, P. H., and J. W. Lynch. 1991. Liquid junction potentials and small cell effects in patch-clamp analysis. J. Membr. Biol. 121:101. 19. Kielland, J. 1937. Indivdual activity coefficients of ions in aqueous solutions. J. Am. Chem. Soc. 59:1675. 20. Lewis, C., S. Neidhart, C. Holy, R. A. North, G. Buell, and A. Surprenant. 1995. Coexpression of P2X2 and P2X3 receptor subunits can account for ATP-gated currents in sensory neurons. Nature 377:432. 21. Evans, R. J., C. Lewis, C. Virginio, K. Lundstrom, G. Buell, A. Surprenant, and R. A. North. 1996. Ionic permeability of, and divalent cation effects on, two ATP-gated cation channels (P2X receptors) expressed in mammalian cells. J. Physiol. 497:41.3. 22. Merritt, J. E., S. A. McCarthy, M. P. Davies, and K. E. Moores. 1990. Use of fluo-3 to measure cytosolic Ca2⫹ in platelets and neutrophils: loading cells with the dye, calibration of traces, measurements in the presence of plasma, and buffering of cytosolic Ca2⫹. Biochem. J. 269:513. 23. Wehage, E., J. Eisfeld, I. Heiner, E. Jungling, C. Zitt, and A. Luckhoff. 2002. Activation of the cation channel long transient receptor potential channel 2 (LTRPC2) by hydrogen peroxide: a splice variant reveals a mode of activation independent of ADP-ribose. J. Biol. Chem. 277:23150. 24. Wilding, M., G. L. Russo, A. Galione, M. Marino, and B. Dale. 1998. ADPribose gates the fertilization channel in ascidian oocytes. Am. J. Physiol. 275: C1277. 25. Gabelli, S. B., M. A. Bianchet, M. J. Bessman, and L. M. Amzel. 2001. The structure of ADP-ribose pyrophosphatase reveals the structural basis for the versatility of the Nudix family. Nat. Struct. Biol. 8:467. 26. Reeders, S. T., M. H. Breuning, K. E. Davies, R. D. Nicholls, A. P. Jarman, D. R. Higgs, P. L. Pearson, and D. J. Weatherall. 1985. A highly polymorphic DNA marker linked to adult polycystic kidney disease on chromosome 16. Nature 317:542. 27. Hanaoka, K., F. Qian, A. Boletta, A. K. Bhunia, K. Plontek, L. Tsiokas, V. P. Sukhatme, W. B. Guggino, and G. G. Germino. 2000. Co-assembly of polycystin-1 and -2 produces unique cation-permeable currents. Nature 408:990. 28. Shubinsky, G., and M. Schlesinger. 1997. The CD38 lymphocyte differentiation marker: new insight into its ectoenzymatic activity and its role as a signal transducer. Immunity 7:315. 29. Deaglio, S., K. Mehta, and F. Malavasi. 2001. Human CD38: a (r)evolutionary story of enzymes and receptors. Leuk. Res. 25:1.
