During the Respiratory Burst, Do Phagocytes Need Proton Channels or Potassium Channels, or Both?

See allHide authors and affiliations

Science's STKE  18 May 2004:
Vol. 2004, Issue 233, pp. pe21
DOI: 10.1126/stke.2332004pe21


The NADPH (reduced form of nicotinamide adenine dinucleotide phosphate) oxidase enzyme complex, a crucial component of innate immunity, produces superoxide anion (O2), which is a precursor to many reactive oxygen species. NADPH oxidase produces O2 by transferring electrons from intracellular NADPH across the membrane to extracellular (or phagosomal) oxygen and is thus electrogenic. It is widely believed that electroneutrality is preserved by proton flux through voltage-gated proton channels. A series of recent papers have challenged several key aspects of this view of the "respiratory burst." The most recent study solidifies the proposal that O2- and other reactive oxygen species produced by phagocytes are not toxic to microbes under physiological conditions. Further, an essential role for high-conductance, Ca2+-activated K+ (maxi-K+) channels in microbe killing is proposed. Finally, the results cast doubt on the widely held view that H+ efflux through voltage-gated proton channels (i) is the main mechanism of charge compensation, and (ii) is essential to continuous O2- production by the NADPH oxidase. My analysis of the new data and of a large body of data in the literature indicates that the proposed role of maxi-K+ channels in the respiratory burst is not yet credibly established. H+ efflux through proton channels thus remains the most viable mechanism for charge compensation and continuous O2 production. The important question of the toxicity of reactive oxygen species in phagocytes and in other cells, which has long been simply taken for granted, is a widespread assumption that deserves critical study.

The Respiratory Burst

Phagocytosis is an important defense mechanism of innate immunity, in which a macrophage, neutrophil, or eosinophil encounters a microbial invader, engulfs it, and kills it. NADPH oxidase mediates the "respiratory burst," a large(100-fold) increase in oxygen consumption that occurs during phagocytosis and results in microbe death. This multicomponent enzyme (Fig. 1) produces superoxide anion (O2), which is a precursor to various reactive oxygen species (ROS). NADPH oxidase function is clearly essential to innate immunity, because a hereditary defect in any of the major NADPH oxidase components results in chronic granulomatous disease (CGD). Without medical intervention, CGD patients typically die in childhood of severe recurrent infections (1). The idea that ROS themselves are lethal to bacteria has been challenged by Segal’s group (2, 3). In their recent study, Ahluwalia et al. (4) not only conclude that ROS contribute little to microbe killing, but also cast doubt on the widespread belief that ROS contribute to various human diseases.

Fig. 1.

The "respiratory burst" in a phagocyte is triggered when a bacterium is phagocytosed. During the phagocytosis of bacteria by macrophages and neutrophils, the phagosome membrane pinches off and the microbe is endocytosed along with a small volume of extracellular fluid. Eosinophils attack helminths too large to endocytose, and NADPH oxidase assembles in their plasma membrane. The main components of NADPH oxidase are shown in yellow. The cytoplasmic components are at a remote location inside resting phagocytes; stimulation results in their migration to the membrane, where the entire complex assembles. Electrons are removed from NADPH in the cytoplasm and transferred through the gp91phox component (which includes flavin adenine dinucleotide and two hemes) across the membrane, where they reduce extracellular (or intraphagosomal) O2 to O2. Protons left behind in the cell are extruded through voltage-gated proton channels (red). Some of the ROS derived from O2 are indicated. Spontaneous or superoxide dismutase–catalyzed disproportionation of O2 produces hydrogen peroxide (H2O2), which may be converted to HOCl (hypochlorous acid, or household bleach) by myeloperoxidase (MPO). (A) Traditional view of the respiratory burst with charge compensation by proton channels. A perfect match of one proton per electron results in no change in membrane potential, intracellular pH (pHi), or external pH (pHo) and little change in ionic strength (2, 28). Because proton channels are separate molecules and for the most part operate independently of NADPH oxidase, perfect 1:1 stoichiometry is not obligatory. The large depolarization that occurs during the respiratory burst in intact neutrophils and eosinophils (41, 42) is likely the most important factor that causes proton channels to open (11), although both pHi and pHo tend to change in a direction that causes proton channels to open. The fact that depolarization occurs demonstrates unequivocally that proton efflux initially lags behind electron efflux, as was first pointed out by Henderson et al. (5, 28). (B) If any fraction of the total charge compensation were mediated by K+ efflux, pHi would fall, pHo (or phagosomal pH) would increase, and the osmolality of the phagosomal contents would increase. In this model (2, 4), the elevated pH and osmolality of the phagosomal contents are crucial to activating proteolytic enzymes that actually kill bacteria, as opposed to ROS, which are said to be inert (3). An extreme version of this model obviates the contribution of proton channels and ascribes the entire charge compensation function to maxi-K+ channels (4). [Adapted by permission from (11).]

The discovery that NADPH oxidase is electrogenic (5) was pivotal to understanding its mechanism. The enzyme moves electrons from NADPH inside the cell across the cell membrane to reduce extracellular oxygen to superoxide (Fig. 1A). A practical consequence is that one can detect NADPH oxidase activity directly and quantitatively in real time as an electrical current. NADPH oxidase–generated electron current has been recorded in cells dialyzed with artificial cytoplasm in the whole-cell patch-clamp configuration (6, 7) and in intact phagocytes studied using the perforated-patch approach (8, 9). The latter approach has the advantage of being physiological in the sense that the cell is allowed to keep its own cytoplasm. Because the cytoplasm is retained, intracellular signaling pathways that are lost in whole-cell configuration are preserved and the respiratory burst can be triggered and studied at the single-cell level.

Electron translocation by NADPH oxidase will produce rapid, massive depolarization unless it is balanced by positive current flow across the membrane. This physical requirement, based on Maxwell’s equations, was clearly stated in 1952 by Hodgkin et al. in regard to sodium and potassium currents in squid axons (10):

Itotal = CM(dV/dt) + Iionic

Applying this equation to the respiratory burst, the total electron current Itotal must equal the sum of the balancing ionic current Iionic (which could be outward cation or inward anion flux) and the membrane capacity CM term that gives the rate of membrane depolarization. During respiratory burst, the membrane potential of an eosinophil would depolarize at a rate of 18 V/s if no charge compensation mechanism existed (11). Within 20 ms, the membrane potential would reach +200 mV, which would prevent electron translocation, effectively shutting down NADPH oxidase function (12). Shortly thereafter, voltages that cause membrane rupture would be reached. It has been generally believed that practically all of the required charge compensation is mediated by H+ efflux through proton channels (1, 5, 1215) (Fig. 1A). Segal’s group reported that ~6% of the charge compensation is mediated by K+ efflux (2). When even a small fraction of charge compensation is mediated by K+, the phagosome becomes hypertonic and its pH increases (Fig. 1B). In their model, both changes enhance bacterial killing by proteolytic enzymes that are released into the phagosome (2). Recently, Segal’s group extended this proposal to suggest that most or all of the charge compensation is mediated by K+ efflux, a scenario that challenges the necessity and even the existence of proton channels in phagocytes (4).

What Ion Channels Are Expressed in Phagocytes?

The electrophysiological data of Ahluwalia et al. (4) differ markedly from those obtained in previous studies of neutrophils and eosinophils. They report "small" currents of 250 pA at +140 mV in resting neutrophils, and describe development of "a large outwardly rectifying current" [p. 854 in (4)] upon stimulation with the phorbol ester, phorbol myristate acetate (PMA). The selectivity of the currents in resting or stimulated cells was not studied; however, they were reversibly inhibited by iberiotoxin, an inhibitor of "maxi-K+" channels (high-conductance, voltage- and calcium-activated K+ channels). On the basis of this pharmacology, Ahluwalia et al. (4) conclude that maxi-K+ channels are responsible for the currents seen in neutrophils and in eosinophils. In neither cell were PMA-activated currents inhibited by 3 mM Zn2+, a classical inhibitor of proton channels (16). If one views these data in isolation, one would conclude that the main and possibly only important ionic conductance in phagocytes is maxi-K+ current. Iberiotoxin and paxilline (another maxi-K+ channel blocker) inhibited Rb+ fluxes, phagosomal alkalinization, and killing of several pathogens (Staphylococcus aureus, Candida albicans, and Serratia marcescens). The implication is that maxi-K+ channels and not proton channels are necessary for effective microbe killing.

Among many previous electrophysiological studies of human eosinophils and neutrophils, none has reported outward currents (other than proton currents) that are within an order of magnitude of the 1000-pA putative maxi-K+ currents reported in neutrophils by Ahluwalia et al. (4). In human neutrophils, Krause and Welsh (17) reported Ca2+-activated K+ and Cl currents that combined were <50 pA at +90 mV. Volume-regulated Cl currents in neutrophils studied by Stoddard et al. (18) were ~70 pA at +80 mV. In previous studies of human eosinophils, outward K+ currents were small or absent: Using KCl-containing pipette solutions, Tare et al. (19) and Gordienko et al. (20) saw little outward current (<10 pA). Schwingshackl et al. (21) reported that after PMA enhancement, the total current was ~65 pA at +80 mV, including both Cl and K+ currents. Calcium-activated K+ channels reported in stimulated eosinophils by Saito et al. (22) were not voltage-gated, and their conductance was only 10 to 22 pS, one-tenth that of maxi-K+ channels. Finally, H+ current was the only detectable outward current in PMA-stimulated human eosinophils with active NADPH oxidase in perforated-patch studies with K+-containing pipette solutions (23).

Ahluwalia et al. (4) report that Zn2+, a classical inhibitor of voltage-gated proton channels (24), "at concentrations three orders of magnitude greater than those causing almost complete blockage of proton channels, was also without effect on the currents from neutrophils and eosinophils" [p. 854 in (4)]. In contrast, 42 electrophysiological studies of phagocytes and related leukocytes (eosinophils, neutrophils, macrophages, microglia, mast cells, osteoclasts, basophils, lymphocytes, and related cell lines) have described Zn2+-sensitive, voltage-gated proton current (11). H+ currents in human neutrophils are small (25) and difficult to detect, but the H+ current density in human eosinophils (~200 pA/pF) is the largest known in any cell (20, 11, 26); thus, the failure of Ahluwalia et al. (4) to detect H+ currents is puzzling.

Possible Artifacts in the Standard Superoxide Assay

An important piece of evidence supporting the charge compensation model depicted in Fig. 1A is the inhibition of NADPH oxidase function by H+ channel inhibitors such as Cd2+ and Zn2+ (12, 2732). Ahluwalia et al. (4) challenge this evidence, which is based on a widely used assay in which O2 produced by NADPH oxidase reduces cytochrome c, resulting in a color change that is measured spectroscopically. Using a cell-free assay in which O2 was generated by xanthine plus xanthine oxidase, they showed that 3 mM Zn2+ or Cd2+ accelerated the spontaneous conversion of O2 into H2O2 (dismutation). The implication is that these metals do not inhibit NADPH oxidase activity in phagocytes by inhibiting proton channels, but instead interfere with the measurement of O2. Several previous studies have shown that these metals can interact with various ROS, although interpretation of the data can be complex (3335). If we accept that metals may catalyze the transfer of electrons from O2 to H2O2, then the question is whether and to what extent this mechanism contributes to the apparent inhibition of NADPH oxidase activity in phagocytes.

Three observations indicate that direct catalytic conversion of O2 to H2O2 is not responsible for the observed inhibition of NADPH oxidase activity by metals. First, in previous studies, Zn2+ and Cd2+, at concentrations that inhibit O2 generation by phagocytes (12, 30, 32), did not interfere with the cytochrome c assay, whether O2 was generated by photoreduction of riboflavin using methionine as the electron donor (30) or by the xanthine plus xanthine oxidase method (33). Second, in the presence of Zn2+ concentrations that largely abolish O2 generation (measured as cytochrome c reduction) in phagocytes, addition of the protonophore CCCP (carbonyl cyanide m-chlorophenyl hydrazone) restored part or all of the response (12, 29). Third, COS-7 cells transfected with the four main components of NADPH oxidase produced O2 (measured as cytochrome c reduction) in the presence of ZnCl2 at concentrations up to 1 to 3 mM—conditions in which O2 production is inhibited by 90% in eosinophils and neutrophils (12). In summary, although metals may scavenge ROS under some circumstances, the reported inhibition of NADPH oxidase function by Zn2+ and Cd2+ (12, 2832) is most likely mediated through their effects on proton channels.

How Is Charge Compensated?

If not protons, then what compensates charge? NADPH oxidase is electrogenic: It deposits the equivalent of 4 M electrons into the phagosome during the respiratory burst (2). If this electron flux were not compensated, the phagosome membrane in a neutrophil would depolarize by 6400 V during the respiratory burst (11), roughly 25,000 times the amount of depolarization required to shut off NADPH oxidase completely (12). K+ influx into the phagosome may compensate ~6% of the required charge (2); however, full compensation by K+ would result in 4 M K+ in the phagosome, with intolerable osmotic consequences. Chloride efflux from the phagosome could compensate <4% of the electron current before depleting phagosomal Cl concentration to nil. Ahluwalia et al. (4) reported that Zn2+ had no effect on the currents in PMA-activated neutrophils and eosinophils, but that iberiotoxin completely abolished all currents in stimulated neutrophils without affecting NADPH oxidase activity. These data indicate that the only detectable conductance in phagocytes during the respiratory burst is maxi-K+, yet inhibiting the putative maxi-K+ channel did not inhibit O2 production by neutrophils (4). The results of Ahluwalia et al. (4) thus seem to have eliminated all possible routes of compensating the charge translocated during NADPH oxidase activity. Should we conclude that charge compensation is unnecessary?

Charge compensation is necessary (12), and in all likelihood it is mediated by proton flux. Other than protons, no ion is capable of compensating the massive electron translocation. In contrast, the proton can combine with O2 and, after peregrinating in the guise of various ROS such as H2O2 and HOCl, can essentially disappear as membrane-permeable H2O. Voltage-gated proton channels are observed universally in phagocytes and related cells (11). The level of H+ channel expression is roughly proportional to the amplitude of the respiratory burst in various phagocytes, and in all cells there are at least 10 times as many H+ channels as necessary to compensate maximal NADPH oxidase activity completely (11, 36). Thus, kinetic competence has been demonstrated. Finally, protons are extruded from phagocytes during the respiratory burst stoichiometrically with O2 consumed (3740). In fact, the original discovery that NADPH oxidase was electrogenic (5) was based on pH changes that indicated massive H+ efflux proton efflux during the respiratory burst (5, 28, 41). These observations all indicate that the bulk of charge compensation is mediated by protons.

The Huge Putative K+ Conductance Would Prevent the Depolarization Observed in Phagocytes During the Respiratory Burst

To compensate the electrogenic activity of NADPH oxidase, cation current must match the electron current. The electron efflux generated by NADPH oxidase (adjusted to body temperature) corresponds to 8 to 16 pA in neutrophils (25) and 30 to 40 pA in eosinophils (9, 23). During the respiratory burst, the membrane depolarizes to +58 mV in human neutrophils (42) and +17 mV in human eosinophils (43). The observed depolarization of 120 mV or 50 mV, respectively, from the resting potential is sufficient to activate precisely the H+ efflux necessary for full compensation in both cells. In contrast, the putative maxi-K+ currents reported by Ahluwalia et al. (4) are so large that proton currents, which are ubiquitous in phagocytes (11), were not even detected. Activation of such a large K+ conductance in a living cell would clamp the membrane potential within 10 to 20 mV of the potassium equilibrium potential (EK, roughly –70 mV), thus preventing the massive, sustained depolarization that has been observed to occur during the respiratory burst in intact phagocytes (42, 43). Indeed, this depolarization must occur in order to activate the H+ efflux required for charge compensation (11).

In conclusion, the original Segal model (2)—in which a small fraction of charge is compensated by K+, resulting in biologically important increases in pH and ionic strength of the phagosome—is intriguing and worth pursuing, although several questions have been raised (15, 44). The recently proposed exclusive involvement of maxi-K+ channels in microbe killing by phagocytes (4) is inconsistent with a large body of existing data, and hence must be viewed with skepticism. H+ efflux through proton channels thus remains the most viable mechanism for the charge compensation that is necessary to ensure continuous O2 production in phagocytes during the respiratory burst. Finally, the toxicity of ROS in phagocytes and in other cells, which has long been taken for granted, remains a widespread assumption that deserves critical study.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
View Abstract

Navigate This Article