Department of Physiology, Semmelweis University, H-1444 Budapest, P.O. Box 259, Hungary
Correspondence: Dr. Miklós Geiszt, Bldg. 10, Rm. 11N106, National Institutes of Health, Bethesda, MD 20892. E-mail: mgeiszt{at}nih.gov
|
|
|---|
Key Words: neutrophils NADPH oxidase proton channel calcium membrane potential
|
|
|---|
![]() View larger version (20K): [in a new window] |
Figure 1. The active NADPH oxidase is composed of five subunits. The enzyme
complex drives electrons from NADPH to molecular oxygen, resulting in
strong depolarization of the plasma membrane.
|
|
|
|---|
One early observation is that despite the common finding in all forms of CGD, i.e., the absence or serious reduction of superoxide production, the p47phox-deficient form of CGD usually follows a less severe clinical course than cytochrome b558-deficient forms of the disease [6 , 7 ]. Residual superoxide production observed in p47phox-deficient CGD neutrophils [8 , 9 ] may be responsible for the above-mentioned difference, however a role for other factors, including an additional function of cytochrome b558, cannot be excluded.
Superoxide anions themselves are only weakly anti-pathogenic, and for efficient killing of pathogens, first, superoxide has to be converted to more toxic metabolites. One of the key enzymes involved in this metabolic process is myeloperoxidase (MPO), which catalyzes the formation of HOCl from H2O2. It is interesting that MPO deficiencya more frequent condition than CGD (prevalence is 1:2000)usually manifests only in mild infections or remains essentially asymptomatic [10 ]. Although increased superoxide production or processing of superoxide through alternative pathways (e.g., nitroso compounds) may partially compensate for the absence of MPO, this apparent contradiction remains to be clarified.
A strong argument supporting the central role of superoxide in phagocytic killing is the observation that CGD patients are more sensitive to catalase-positive organisms than to species without this enzyme. It has been proposed that spontaneous H2O2 formation in catalase-negative pathogens may compensate for the defective NADPH oxidase function during phagocytosis, whereas catalase-positive organisms lack this supportive mechanism because they rapidly metabolise H2O2. Although this explanation is generally accepted, its accuracy has never been tested. Moreover, recently Chang et al. [11 ] showed that the virulence of Aspergillus nidulans in p47phox knockout mice was not decreased when the animals were infected by catalase-deficient mutants of the fungus. These observations are particularly important from a clinical standpoint, because Aspergillus species are the most frequent fungal pathogens affecting CGD patients [12 ]. Hopefully, in the near future, similar experiments will focus on the role of catalase enzymes in bacteria.
|
|
|---|
Some more recent findings suggest that intracellular/intraphagosomal pH regulation of CGD neutrophils might be abnormal not only because of the absent superoxide production but also because of altered H+ transport through the plasma membrane. Superoxide production is accompanied normally by a robust intracellular H+ production resulting from the oxidation of NADPH to NADP+ + H+. Neutrophil granulocytes possess three pathways for the removal of protons from the intracellular milieu: Na+-H+ exchange, H+-adenosinetriphosphatase (ATPase), and electrogenic proton transport [14 ]. This latter mechanism was first described by Henderson et al. [15 , 16 ]. They suggested that a proton-conducting channel might be responsible for the transfer of protons through the plasma membrane during activation of superoxide production. The proton transport was activated by low amounts of arachidonic acid and inhibited by divalent cations such as Zn2+ or Cd2+. Subsequent work from our laboratory demonstrated that phorbol 12-myristate 13-acetate (PMA), the most effective activator of superoxide production, is also a potent stimulus of this proton conductance [17 ]. The proton-conducting pathway of phagocytes was characterized by patch-clamp measurements, proving its electrogenic feature [18 19 20 ], however the molecular nature of the conductance remained elusive. Henderson et al. [21 ] could show that gp91phox functions as a proton channel when expressed in Chinese hamster ovary (CHO) cells, and even an N-terminal truncated form of the protein seems to be functional [22 ]. On the contrary, it was also demonstrated in patch-clamp experiments that gp91phox-deficient CGD monocytes still express a proton-conducting pathway [23 ], however PMA activation seemed to be disturbed in the CGD cells [24 ]. Furthermore, arachidonate activatable Zn2+- and Cd2+-sensitive, but PMA-resistant, electrogenic H+ transport was detected in T lymphocytes, where gp91phox protein was absent [25 ]. In a recent work, DeCoursey et al. [26 ] studied NADPH oxidase-related proton and electron currents in PMA-stimulated human neutrophils. The authors did not find any correlation between the PMA-induced electron current and proton conductance; therefore, they concluded that NADPH oxidase probably did not mediate the observed proton current. They also suggested that PMA stimulation probably modulates pre-existing proton channels rather than introducing new ones into the plasma membrane. On the contrary, a study by Banfi et al. [27 ] showed that two separate, proton-conducting pathways exist in phagocytes. The "basal" H+ conductance opens when the membrane potential becomes more positive than the H+ equilibrium potential and shows definite outward rectification, whereas the oxidase-associated conductance opens upon activation of the enzyme (e.g., by PMA) at a more negative potential than the H+ equilibrium potential and allows the movement of protons in both directions, following their electrochemical gradient.
Concerning the structural requirements for proton transport by gp91phox, site-directed mutagenesis experiments by Henderson and Meech [28 ] highlighted the importance of three histidine residues at positions 111, 115, and 119 in the third transmembrane domain of gp91phox. One of these histidines, His115, is also implicated in heme binding of gp91phox. The authors proposed a mechanism for proton transport through gp91phox, which includes the protonation-deprotonation of the histidine residue at 115, and residues at positions 111 and 119 are responsible for voltage-gating.
The importance of the missing H+-channel function in gp91phox-deficient CGD is unknown, however it might be an additional factor in the previously described difference between p47phox- and gp91phox-deficient CGD phenotypes.
The significance of gp91phox as a regulator of pH homeostasis was strengthened recently by the discovery of the mitogenic oxidase (Mox1), a homologue of gp91phox [29 ]. Mox1 is 56% identical to gp91phox and contains the structural elements, which are important in the electron transport activity of its phagocytic counterpart. Mox1 is expressed in the colon, prostate, uterus, and vascular smooth muscle and increases superoxide production and cell proliferation when overexpressed in NIH3T3 fibroblasts. Transfection of aortic smooth-muscle cells with antisense Mox1 decreased superoxide production and inhibited serum-dependent growth of the cells.
It is interesting that an alternatively spliced variant of the Mox1 transcript was also described [30 ], encoding a protein named NOH-1S (NADPH oxidase homolog 1, short variant), which is a 191 amino acid protein, containing the first three transmembrane domains of Mox1 but having a different fourth transmembrane region followed by a short C-terminal sequence. The three conserved, histidine residues, which are implicated in the proton channel function of gp91phox, are located at the same position in NOH-1S (positions 111, 115, and 119), however in the absence of the fifth transmembrane domain, His115 probably does not bind heme but might serve solely as proton translocator. When NOH-1S is expressed in HEK-293 cells, it shows voltage-dependent H+ channel activity, which is sensitive to Zn2+. It is interesting that aside from the tissues where Mox1 is expressed, the NOH-1S message was also detected in human leukaemia (HL-60) cells and leukocytes. The fact that the H+ channel function can be dissociated from the oxidase activity in the form of a separate protein also highlights the importance of the H+ channel activity of gp91phox.
Recently another novel homologue of gp91phox, Renox (Renal oxidase), was isolated and found to be expressed at high levels in kidney tubular cells [31 ]. Renox is a 578 amino acid-long protein showing 40% identity and 57% similarity to gp91phox. Renox shares all the structural features of gp91phox and Mox1, which are considered to be essential for NADPH oxidase function. NIH-3T3 fibroblasts overexpressing Renox showed increased superoxide production and in contrast to Mox1-transfected fibroblasts, showed reduced proliferation rate and signs of cellular senescence. Based on its localization and enzymatic function, Renox is a likely candidate for oxygen-sensor function in the kidney and might be responsible for the regulation of erythropoietin synthesis. When considering a possible proton channel function of Renox, it is noteworthy that Renox contains a phenylalanine residue in the position corresponding to His111 of gp91phox, although the other histidine residues are tightly conserved. Based on this structural difference, Renox offers an exciting experimental tool in the analysis of structural requirements for the proton transport function of gp91phox and its homologues.
|
|
|---|
These findings argued strongly for an electrogenic, NADPH oxidase, however this property was not directly assessed until recently. The first direct evidence comes from the work of Schrenzel et al. [34 ], who were able to detect the electron current generated by the NADPH oxidase by patch-clamp measurements. Working on human eosinophils, they detected an inward current, which was inhibited by DPI and was absent in CGD eosinophils. Furthermore, they could demonstrate the absolute dependence of the current on the availability of NADPH on the cytosolic side. They also described the activation of the current by N-formyl-methionyl-leucyl-phenylalanine (fMLP), a well-known stimulus of the NADPH oxidase. It is interesting that the current was inward over the entire tested voltage range (-100 mV+60 mV), indicating an exceptionally strong charge-separator capability of the oxidase. Previously, the oxidase-generated depolarization was measured only by fluorescent techniques, which do not allow accurate calibration. However, even with these traditional techniques, it seemed possible that granulocytes depolarize to positive membrane-potential values during the activation of the oxidase. According to the electrophysiological measurements of Schrenzel et al. [34 ], this value is +25 mV+30 mV, which is comparable with the membrane potential response of excitable cells producing action potentials. The most recent experiments carried out with an improved fluorescent technique showed that the membrane potential of PMA-activated neutrophils could reach +58 mV [35 ].
|
|
|---|
When neutrophils are stimulated by various receptor agonists, an increase of intracellular calcium concentration ([Ca2+]i) occurs. This Ca2+ transient has an important role in the transduction of extracellular signals, leading to the stimulation of different effector responses (e.g., degranulation, superoxide production) of the cells [37 ]. The agonist-induced Ca2+ signal of neutrophils is composed of two components: a rapid, transient increase of [Ca2+]i as a result of Ca2+ release from the internal stores and a more sustained elevation of [Ca2+]i because of Ca2+ influx from the extracellular space. Neutrophil granulocytes do not express voltage-dependent Ca2+ channels, but the inositol triphosphate (IP3)-mediated emptying of internal Ca2+ stores induces Ca2+ influx from the extracellular space [38 ]. This mechanism of calcium entry is referred to as capacitative or store-operated Ca2+ entry [39 ], which seems to be a major Ca2+-uptake pathway in almost every mammalian cell type. In spite of intensive efforts, the exact mechanism of capacitative Ca2+ influx is still unknown, however there is growing evidence that IP3-induced emptying of the internal stores might result a conformational change of the IP3-receptor structure, which then activates, either directly or indirectly, a plasma membrane Ca2+ channel [40 ]. Other models of capacitative Ca2+ influx propose that emptying the internal stores leads to channel activation through a soluble messenger substance released from the empty stores. In human neutrophils, agonists of various chemotactic receptors [e.g., fMLP and platelet-activating factor (PAF)] initiate Ca2+ influx by emptying the internal stores. It is interesting that fMLP but not PAF inhibits capacitative Ca2+ influx by a poorly understood mechanism [41 ]. Thus, it appears that the net calcium entry into neutrophils following the fMLP stimulus results from the concurrent presence of activating and inhibiting mechanisms. The protein kinase C activator PMA was also shown to inhibit the capacitative pathway of neutrophils [42 ]. fMLP and PMA are well-known, effective stimuli of the superoxide-producing, NADPH oxidase; therefore, we asked whether a connection exists between inhibition of the calcium influx and stimulation of superoxide production.
Generation of superoxide by the NADPH oxidase results in depolarization of the cells because of transmembrane electron transfer from the intracellular space to the extracellular milieu. Depolarization was shown to inhibit Ca2+ influx in many cell types; therefore, we studied if a relationship exists between the inhibition of Ca2+ influx and NADPH oxidase-induced depolarization. Pharmacologically induced manipulation of the membrane potential revealed a close relationship between depolarization and inhibition of Ca2+ influx [36 ] (Fig. 2 A ). We have also demonstrated that the agonist-mediated inhibition of capacitative calcium influx is absent in the neutrophils of patients suffering from chronic granulomatous disease (Fig. 2B) . As a consequence of this regulatory defect, CGD neutrophils start to accumulate calcium from the extracellular space at an earlier time-point than healthy neutrophils, and the internal Ca2+ stores of CGD neutrophils become refilled more rapidly following receptor stimulation. The pathophysiological significance of this finding and its contribution to the CGD phenotype are currently unknown, however capacitative Ca2+ influx was shown to have a role in regulation of cytokine production of neutrophils [43 ], and the calcium content of internal stores might control the activity of phospholipase A2, an enzyme that has a key role in the inflammatory response [44 ].
![]() View larger version (15K): [in a new window] |
Figure 2. (A) In healthy cells, the NADPH oxidase-induced depolarization inhibits
Ca2+ influx through the capacitative pathway.
(B) In CGD neutrophils, the absence of oxidase-induced depolarization
results in increased Ca2+ entry into the
cells.
|
|
|
|---|
![]() View larger version (22K): [in a new window] |
Figure 3. NADPH oxidase probably affects the movement of several ions among the
extracellular, intracellular, and intraphagosomal spaces. Depending on
its charge and concentration gradient, the oxidase-induced
depolarization can increase (thick arrows) or decrease (thin arrows)
the net transport of a particular ion.
|
We believe that the ion homeostasis of CGD granulocytes should be analyzed carefully in future experiments so that we might gain a better understanding of the pathogenesis and effective treatment [49 ] of chronic granulomatous disease.
Received June 22, 2000; revised October 18, 2000; accepted October 20, 2000.
|
|
|---|
This article has been cited by other articles:
![]() |
N. J. Jann, M. Schmaler, S. A. Kristian, K. A. Radek, R. L. Gallo, V. Nizet, A. Peschel, and R. Landmann Neutrophil antimicrobial defense against Staphylococcus aureus is mediated by phagolysosomal but not extracellular trap-associated cathelicidin J. Leukoc. Biol., November 1, 2009; 86(5): 1159 - 1169. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. J. Coffey, C. H. Serezani, S. M. Phare, N. Flamand, and M. Peters-Golden NADPH oxidase deficiency results in reduced alveolar macrophage 5-lipoxygenase expression and decreased leukotriene synthesis J. Leukoc. Biol., December 1, 2007; 82(6): 1585 - 1591. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Orient, A. Donko, A. Szabo, T. L. Leto, and M. Geiszt Novel sources of reactive oxygen species in the human body Nephrol. Dial. Transplant., May 1, 2007; 22(5): 1281 - 1288. [Full Text] [PDF] |
||||
![]() |
K. Bedard and K.-H. Krause The NOX Family of ROS-Generating NADPH Oxidases: Physiology and Pathophysiology Physiol Rev, January 1, 2007; 87(1): 245 - 313. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. K. Femling, V. V. Cherny, D. Morgan, B. Rada, A. P. Davis, G. Czirjak, P. Enyedi, S. K. England, J. G. Moreland, E. Ligeti, et al. The Antibacterial Activity of Human Neutrophils and Eosinophils Requires Proton Channels but Not BK Channels J. Gen. Physiol., May 30, 2006; 127(6): 659 - 672. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. K Rada, M. Geiszt, C. Hably, and E. Ligeti Consequences of the electrogenic function of the phagocytic NADPH oxidase Phil Trans R Soc B, December 29, 2005; 360(1464): 2293 - 2300. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. von Knethen, A. Tautenhahn, H. Link, D. Lindemann, and B. Brune Activation-Induced Depletion of Protein Kinase C{alpha} Provokes Desensitization of Monocytes/Macrophages in Sepsis J. Immunol., April 15, 2005; 174(8): 4960 - 4965. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. K. Rada, M. Geiszt, K. Kaldi, C. Timar, and E. Ligeti Dual role of phagocytic NADPH oxidase in bacterial killing Blood, November 1, 2004; 104(9): 2947 - 2953. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. E. DeCoursey During the Respiratory Burst, Do Phagocytes Need Proton Channels or Potassium Channels, or Both? Sci. Signal., May 18, 2004; 2004(233): pe21 - pe21. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. L. Petheo, A. Maturana, A. Spat, and N. Demaurex Interactions between Electron and Proton Currents in Excised Patches from Human Eosinophils J. Gen. Physiol., November 24, 2003; 122(6): 713 - 726. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. D. Vinion-Dubiel and J. B. Goldberg Review: Lipopolysaccharide of Burkholderia cepacia complex Innate Immunity, August 1, 2003; 9(4): 201 - 213. [Abstract] [PDF] |
||||
![]() |
T. E. Decoursey Voltage-Gated Proton Channels and Other Proton Transfer Pathways Physiol Rev, April 1, 2003; 83(2): 475 - 579. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Strange Of Mice and Worms: Novel Insights Into ClC-2 Anion Channel Physiology Physiology, February 1, 2002; 17(1): 11 - 16. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. M. McManus, R. C. Bloodworth, T. J. Prihoda, J. L. Blodgett, and R. N. Pinckard Agonist-dependent failure of neutrophil function in diabetes correlates with extent of hyperglycemia J. Leukoc. Biol., September 1, 2001; 70(3): 395 - 404. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||