SK channels mediate NADPH oxidase-independent reactive oxygen species production and apoptosis in granulocytes

Abstract

Neutrophils are immune cells that bind to, engulf, and destroy bacterial and fungal pathogens in infected tissue, and their clearance by apoptosis is essential for the resolution of inflammation. Killing involves both oxidative and nonoxidative processes, the oxidative pathway requiring electrogenic production of superoxide by the membrane-bound NADPH oxidase complex. A variety of stimuli, from bacterial chemotactic peptides to complement- or IgG-opsonized microbes, can induce the production of reactive oxygen species (ROS) by neutrophils, presumably by means of NADPH oxidase. We report here that 1-ethyl-2-benzimidazolinone (1-EBIO), an activator of Ca²⁺-activated potassium channels of small conductance (SK) and intermediate conductance (IK), causes production of superoxide and hydrogen peroxide by neutrophils and granulocyte-differentiated PLB-985 cells. This response can be partially inhibited by the SK blocker apamin, which inhibits a Ca²⁺-activated K⁺ current in these cells. Analysis of RNA transcripts indicates that channels encoded by the SK3 gene carry this current. The effects of 1-EBIO and apamin are independent of the NADPH oxidase pathway, as demonstrated by using a PLB-985 cell line lacking the gp91phox subunit. Rather, 1-EBIO and apamin modulate mitochondrial ROS production. Consistent with the enhanced ROS production and K⁺ efflux mediated by 1-EBIO, we found that this SK opener increased apoptosis of PLB-985 cells. Together, these findings suggest a previously uncharacterized mechanism for the regulation of neutrophil ROS production and programmed cell death.

Neutrophils, as key players in the innate immune response, must be able to identify, phagocytose, and neutralize a broad array of pathogenic microbes (1). Killing usually involves isolation of bacteria or fungi within the phagosome, although neutrophils can release toxic contents into their surroundings to kill extracellular microbes (1). These toxic contents include reactive oxygen species (ROS), such as hydrogen peroxide and hypochlorous acid, as well as proteases and antimicrobial peptides. Because these toxins also can damage host cells, limiting their release and clearing apoptotic neutrophils ultimately are necessary for resolving an infection (2). Understanding how neutrophils regulate the production of oxidants is, therefore, important for both resisting pathogens and controlling inflammation.

Although ion channels are best known for their roles in neuronal and cardiac action potentials, their importance in some immune cells, particularly neutrophils, has begun to emerge over the past few years (1, 3–5). The NADPH oxidase of neutrophils produces large amounts of superoxide at the plasma and phagosomal membranes, resulting in an electrogenic efflux of electrons from the cytoplasm (6). An extensive literature indicates that a proton channel, whose molecular identity has recently been identified (7, 8), carries a compensating outward current that prevents depolarization of the cell membrane to potentials that inhibit the NADPH oxidase (9). Still, relatively few ion channels have been identified in neutrophils, and the roles of observed ionic currents in neutrophil functions remain unclear.

We show here that a small conductance, Ca²⁺-activated potassium (SK) channel in neutrophils and neutrophil-like PLB-985 cells modulates ROS production. Activation of SK/IK channels with 1-ethyl-2-benzimidazolinone (1-EBIO) potentiates superoxide production stimulated by the bacterial chemotactic peptide formyl-methionine-leucine-phenylalanine (fMLF) or by IgG-opsonized Staphylococcus aureus. The specific SK blocker apamin, but not the IK blocker clotrimazole, reduces both of these effects of 1-EBIO and also diminishes peroxide production after fMLF or IgG-opsonized S. aureus activation of the respiratory burst. Using whole-cell patch clamp electrophysiology, we identified an apamin-sensitive current in PLB-985 cells, consistent with an SK channel. We also found SK3 mRNA to be present in both neutrophils and PLB-985 cells by RT-PCR. The enhancement of ROS production by 1-EBIO is independent of the NADPH oxidase inhibitor diphenylene iodonium (DPI) and is present even in a cell line lacking the gp91phox subunit of NADPH oxidase. Instead, these ROS are likely of mitochondrial origin. Consistent with the increased ROS production in the presence of 1-EBIO, the SK/IK opener also increased the rate of staurosporine-induced apoptosis in PLB-985 cells. Together, these results point toward a previously uncharacterized pathway for intracellular ROS regulation parallel to NADPH oxidase activity in neutrophils and PLB-985 cells, and involving SK channels. This pathway may be important for neutrophil apoptosis, which is critical to resolving inflammation (2).

Results

SK Channel Activity Affects ROS Production.

To assay neutrophil and PLB-985 ROS production, we measured the amount of conversion of dihydrorhodamine 123 (DHR 123) to rhodamine 123 in the presence of various drugs that affect channel activity. DHR 123 is oxidized to its fluorescent derivative by several ROS but most prominently by hydrogen peroxide in the presence of a peroxidase (such as myeloperoxidase). Although most potassium channel blockers had no effect, the SK blocking toxin apamin was found to partially reduce peroxide levels when PLB-985 cells were stimulated with either IgG-opsonized, heat-killed S. aureus or with the bacterial chemotactic peptide fMLF (Fig. 1A). Another SK blocker, d-tubocurarine (10), also decreased oxidation of DHR 123 (data not shown). On the other hand, the SK activator 1-EBIO potently increased rhodamine 123 formation by both neutrophils and PLB-985 cells (Fig. 1B). These effects also were observed by using the fluorogenic probe Peroxyfluor-1 (PF1), a peroxidase-independent hydrogen peroxide sensor (11) (data not shown), suggesting a potential role for SK channels in peroxide formation.

Fig. 1.

SK blocker and opener affect PLB-985 ROS production in PLB-985 cells. (A) Effects of NADPH oxidase inhibitor and SK and BK/IK blockers on opsonized S. aureus-induced DHR 123 oxidation. (B) SK/IK opener 1-EBIO enhances ROS production. (C) Apamin partially inhibits and ionomycin potentiates stimulation of ROS production by 1-EBIO. Data are given in arbitrary units. ∗∗, P < 0.01; ∗∗∗, P < 0.001.

Because 1-EBIO is believed to potentiate SK current by increasing the affinity of Ca²⁺-calmodulin for the SK channel subunit (12), thus causing greater current at lower calcium concentrations, we hypothesized that the effects of 1-EBIO should be increased by a rise in cytosolic calcium. Indeed, treatment of cells with ionomycin, a calcium ionophore, enhanced the effects of 1-EBIO on rhodamine 123 formation, even though ionomycin itself had no effect on resting levels of rhodamine 123 formation (Fig. 1C).

PLB-985 Cells Express a Ca²⁺-Activated Apamin-Sensitive K⁺ Current.

Given the pharmacological evidence for SK channel involvement in ROS production, we set out to demonstrate the presence of an apamin-sensitive current in the promyelocytic leukemia PLB-985 cell line. Stimulation of neutrophils or differentiated promyelocytic cell lines with inflammatory agonists, such as the PKC activator phorbol 12-myristate 13-acetate (PMA) or the bacterial chemotactic peptide fMLF, is known to alter plasma membrane potential by affecting ion channel activity (13–15). Using whole-cell voltage clamp electrophysiology, we found that 1 μM intracellular Ca²⁺ activated an outward current in PLB-985 cells (in four of nine cells tested) with a reversal potential consistent with a potassium-selective channel. This current could be selectively and reversibly inhibited by apamin (Fig. 2), a specific blocker of SK channels (16). We next set out to determine which of the three cloned SK genes, SK1, SK2, and SK3 (17), is expressed in neutrophils and neutrophil-like cell lines.

Fig. 2.

An apamin-sensitive whole-cell current in PLB-985 cells. (A1) Current traces of a PLB-985 cell voltage-clamped from a holding potential of −60 mV to various voltages (−100 to +60 mV in 10-mV increments) for 100 ms followed by a step to −60 mV; pipette internal solution contains 1 μM free Ca²⁺. (A2) Flowing in 3 μM extracellular apamin greatly reduced the current. (A3) Washout of apamin brought the current back to the original level, indicating reversible blocking effect of apamin. (B) Plots of current vs. voltage from experiments in A, with and without apamin, as indicated.

PLB-985 Cells Express SK3 Transcript.

After purifying RNA from human neutrophils, PLB-985 cells, and HL-60 cells, we used reverse transcription and primers specific for SK1, SK2, or SK3 to amplify any transcript present for these channels. Although no transcript was detected for either SK1 or SK2, SK3 was present in all three cell types tested (Fig. 3). Interestingly, SK3 is transcribed in both differentiated and undifferentiated (data not shown) PLB-985 cells. Thus, it appears that SK3 is the most likely candidate for mediating the apamin and 1-EBIO effects on ROS production described above.

Fig. 3.

RT-PCR analysis of RNA to identify SK1, SK2, and SK3 expression in neutrophils and neutrophil-like cells. Results are shown for primers specific for SK3 (A), SK1 (B), and SK2 (C). Leftmost lanes show molecular weight standards (in base pairs).

Apamin and 1-EBIO Affect Superoxide Production.

Because superoxide is, in most cases, the precursor to hydrogen peroxide, we also measured the effects of apamin and 1-EBIO on superoxide production, by using the superoxide-specific probe dihydroethidium (DHE). DHE is oxidized to red fluorescent ethidium by superoxide and then preferentially associates with DNA. 1-EBIO caused a clear increase in superoxide production, which could be partially inhibited by apamin (Fig. 4A), consistent with the results obtained by using DHR 123. Oxidation of DHE by 1-EBIO was enhanced by ionomycin (Fig. 4B), in agreement with 1-EBIO's activation of a SK channel.

Fig. 4.

SK blocker and opener effects on superoxide production. (A) SK/IK opener 1-EBIO enhances DHE oxidation induced by fMLF, partially inhibited by apamin. (B) Ionomycin increases stimulation of superoxide production by 1-EBIO. ∗, P < 0.05; ∗∗∗, P < 0.001.

1-EBIO Effect Is Independent of NADPH Oxidase.

Although DPI is a relatively specific inhibitor of NADPH oxidase, the only way to unequivocally test the involvement of the oxidase in apamin-sensitive and 1-EBIO-dependent effects is to knock out one of the subunits of the oxidase. This genotype of patients with chronic granulomatous disease (18) has been mimicked in PLB-985 cells by targeted deletion, and we used this gp91phox knockout cell line (PLB-X-CGD) (19) to determine whether the NADPH oxidase is necessary for the effects of 1-EBIO on neutrophil ROS production. Even though PLB-985 wild-type cells have a fMLF-stimulated respiratory burst and PLB-X-CGD cells do not respond to the chemotactic peptide, 1-EBIO induced DHR 123 oxidation in both cell lines (Fig. 5). In addition, although the NADPH oxidase inhibitor DPI decreased ROS production in wild-type cells, DPI increased CGD cells' formation of superoxide and peroxide (Fig. 5). This enhancement of ROS production is likely attributable to inhibition of complex I (NADH dehydrogenase) (20) of the mitochondrial electron-transport chain by DPI, because another complex I inhibitor, rotenone (21), had the same effect (data not shown). These findings show that NADPH oxidase-independent ROS production in neutrophil-like cells is regulated by SK channels, and they strongly indicate that the mitochondrial oxidative phosphorylation cascade is the source of the observed ROS effects.

Fig. 5.

DHR 123 oxidation by wild-type and gp91phox knockout PLB-985 cells. (A) The NADPH oxidase inhibitor DPI decreases, whereas 1-EBIO increases, fMLF-stimulated ROS production in PLB-985-WT cells. (B) PLB-985-X-CGD cells lack fMLF-stimulated ROS production, but ROS production still is stimulated by 1-EBIO. Data are given in arbitrary units. ∗, P < 0.05; ∗∗, P < 0.01; ∗∗∗, P < 0.001.

1-EBIO and Apamin Alter Mitochondrial ROS Production.

To test more specifically the involvement of mitochondria in the effects of 1-EBIO and apamin, we used MitoSOX (22), a superoxide-specific probe, to explore ROS production within the mitochondria. The time course of superoxide production is sensitive to rotenone, a complex I inhibitor that is known to increase superoxide production in the mitochondrial matrix by forcing the electron-transport chain to run in reverse (data not shown). 1-EBIO also caused an increase in the rate of superoxide production, an effect that can be inhibited by apamin (Fig. 6). The fact that these results are qualitatively similar to those obtained with DHR 123, DHE, and PF1 strongly suggests that all of these probes are reporting on the same pathway of ROS production, and that the mitochondria are the most likely source of these ROS. Neither apamin nor 1-EBIO, however, altered the mitochondrial membrane potential, as measured by the inner-membrane-localized, potential-sensitive probe MitoTracker Red (data not shown).

Fig. 6.

Effects of apamin and 1-EBIO on mitochondrial superoxide production measured by MitoSOX oxidation. Differences between control and 1-EBIO and between 1-EBIO and 1-EBIO + Apamin are statistically significant (∗, P < 0.01).

1-EBIO Promotes PLB-985 Cell Apoptosis.

Given the central role that both mitochondrial ROS generation and plasma membrane K⁺ efflux play in apoptosis, we wondered whether SK channel activity might affect programmed cell death. Using an assay for apoptosis, based on differential binding by the DNA-intercalating dyes Yo-Pro (23) and propidium iodide, we showed that 1-EBIO increased, whereas apamin inhibited, staurosporine-induced apoptosis, indicating that SK activity can mediate its effects through a well characterized apoptotic pathway (Fig. 7). Neither apamin nor 1-EBIO alone affected apoptosis of PLB-985 cells.

Fig. 7.

Effects of SK activity on PLB-985 cell apoptosis. (A and B) FACS dot plots showing live (lower left box), apoptotic (lower right box), and necrotic (upper right box) cells after treatment with staurosporine (STA) (A) or STA + 1-EBIO (B). (C) Quantification of effects of apamin and 1-EBIO on STA-induced apoptosis. ∗, P < 0.05; ∗∗, P < 0.01; ∗∗∗, P < 0.001.

Discussion

This work identifies an SK channel in neutrophils and the granulocytic cell line PLB-985 and shows that SK channel activity affects ROS production by these cells. The SK blocker apamin reduces peroxide formation, as measured by oxidation of DHR 123 or PF1, and the SK/IK activator 1-EBIO increases peroxide production in these assays and in assays to detect superoxide formation. Apamin partially compensates for the effects of 1-EBIO, consistent with SK channel pharmacology. An apamin-sensitive current is present in PLB-985 cells, and RT-PCR shows that SK3 is the likely molecular correlate to this current. Because the effects of SK channels seem to mediate ROS production in the mitochondria, rather than through the NADPH oxidase, it is unlikely that this pathway has a direct effect on microbial killing. However, we find that triggering the opening of SK channels enhances staurosporine-induced apoptosis in PLB-985 cells.

Although SK channels have been extensively characterized in neurons (16, 24, 25), where they mediate the afterhyperpolarization, and in some endocrine cells (26–28), where they are thought to regulate hormone secretion, the functions of ionic fluxes in nonexcitable cells, particularly immune cells (29, 30), are only beginning to emerge. The specialized functions of neutrophils require that they release large quantities of ROS, proteases, and basic antimicrobial peptides. Electrogenic superoxide release by the NADPH oxidase has long been thought to be coupled to a voltage-gated proton channel that provides charge balance and prevents voltage-dependent inhibition of the oxidase (9). Segal and coworkers (31) have proposed that the large conductance, Ca²⁺-activated potassium channel (BK) compensates part of the superoxide flux, allowing for potassium-dependent release of basic proteins from the granule proteoglycan matrix (32). A recent study (33), however, casts doubt on the presence of these channels in neutrophils.

Mitochondria are a second, less prominent source of ROS (34) in neutrophils. Because neutrophils rely predominantly on the hexose monophosphate shunt for NAD(P)H production (35), it has been assumed that mitochondria are superfluous for neutrophil function. However, more recent studies have shown that neutrophil mitochondrial activity, although not essential for phagocytosis or respiratory burst initiation, does affect chemotaxis (36) and apoptosis (37, 38). Neutrophil phagocytosis may even involve incorporation of some mitochondrial components into the phagosome (39). It is clear from our findings that neutrophil mitochondria can produce substantial quantities of ROS, and that plasma membrane SK3 channels modulate this process.

We set out, initially, to characterize the effects of K⁺ channel blockers on respiratory burst production and found that apamin, a specific peptide toxin blocker of SK channels, inhibits oxidation of DHR 123 (Fig. 1A). This finding indicated an effect of SK channels on neutrophil peroxide production. Earlier electrophysiological studies showed the presence of an SK channel in neutrophils (40) but used no pharmacological tools to further identify the channel. Our findings provide the identification of an apamin-sensitive SK current and its functional roles in neutrophil-like PLB-985 cells. Although there have been several papers suggesting that Ca²⁺-activated potassium channels affect granule release and ROS production by eosinophils (41–43) and microglia (44), none of these earlier studies made a link among potassium channels, ROS production, and apoptosis. Observation of the apamin-sensitive current in only four of nine cells tested may reflect inhomogeneity of the differentiation states or subpopulations of cells (45) or cell-to-cell variation in glycosylation, which could affect binding of the blocker (46). Even in transfected cells, the percentage of current blocked by apamin can show wide variability (47).

The most robust of the pharmacological effects observed is the potentiation of hydrogen peroxide formation by 1-EBIO (Fig. 1), as measured by oxidation of DHR 123. Even without activating the respiratory burst with a physiological stimulus such as fMLF or opsonized bacteria, the SK/IK opener dramatically increased rhodamine 123 fluorescence. This effect was Ca²⁺-dependent, because the calcium ionophore ionomycin potentiated the response to 1-EBIO (Fig. 1C). In addition, apamin, but not the IK blockers clotrimazole (data not shown) or charybdotoxin (Fig. 1A), was able partially to reduce the 1-EBIO effect, indicating that 1-EBIO is acting primarily or exclusively through SK channels to increase ROS production. This inability of apamin to entirely inhibit the effect of 1-EBIO is consistent with electrophysiological studies of SK (47) and may be attributable to heterogeneity of glycosylation of surface SK channels, some of which may respond to the cytosolic 1-EBIO but not to the cell-impermeant peptide apamin.

Although DHR 123 is a widely used probe for studying the respiratory burst in neutrophils, its susceptibility to oxidation by several other ROS in addition to hydrogen peroxide makes it important to verify our findings with another probe. A recently characterized probe, PF1 (11), is converted from a nonfluorescent form to fluorescein in the presence of peroxide (independent of peroxidase), responds less to hydroxyl radical, and is insensitive to hypochlorous acid and peroxynitrite. The effects of 1-EBIO and apamin on PF1 oxidation by PLB-985 cells were qualitatively similar to those obtained with DHR 123. The superoxide-specific probe DHE also showed an increased conversion to its fluorescent, oxidized form, ethidium, in the presence of 1-EBIO (Fig. 4), and this effect could be partially inhibited by apamin. Differences in the magnitudes of effects in these experiments are most likely because of the varying sensitivities of each probe to particular ROS. Thus, experiments performed with several distinct probes indicate that apamin and 1-EBIO alter the level of ROS.

To distinguish NADPH oxidase-dependent and -independent pathways of granulocyte ROS production, it is necessary to remove NADPH oxidase activity. Although the NADPH oxidase inhibitor DPI is a potent pharmacological tool, it also inhibits the NADH dehydrogenase of complex I of the mitochondrial respiratory chain (48). It is therefore important to use available genetic techniques to further characterize the role of NADPH oxidase activity. Patients with X-linked chronic granulomatous disease (18) are susceptible to infections by pathogenic bacteria and fungi, as their neutrophils are incapable of producing a respiratory burst in response to stimuli such as fMLF or opsonized microbes. The molecular basis for this disease is a nonfunctional NADPH oxidase (49), most commonly resulting from mutations in the gp91phox subunit (50). A stable line of PLB-985 cells with targeted knockout of the gp91phox subunit (19) can be differentiated to neutrophil-like cells. As expected, we found that the CGD line did not respond to opsonized bacteria. In addition, DPI increased these cells' ROS production (Fig. 5B), presumably in mitochondria. Potentiation of ROS production by 1-EBIO, in contrast, was the same in wild-type and CGD cell lines, demonstrating that the actions of this drug are independent of NADPH oxidase (Fig. 5).

NADPH oxidase is the source of ROS involved in oxidative microbial killing and possibly in the release of serine proteases (51), but mitochondria provide another, less well characterized source of ROS in neutrophils (34, 52). Mitochondria produce superoxide when electrons escape the electron-transport chain and react with oxygen (51). This leakage is thought to occur primarily at complex I (20) and complex III (53). Inhibition of either complex I or complex III actually increases superoxide formation, presumably by causing greater flavoprotein autooxidation and electron leakage (54). In most cells, mitochondria are the predominant source of ROS, and disruption of the electron-transport chain by complex I inhibitors like rotenone can trigger apoptosis in many cell types (55, 56). Even in neutrophils, which are capable of much greater ROS production than other cells, mitochondria are involved in apoptosis (37), and neutrophil apoptotic pathways can be triggered by ROS (37).

A common feature of apoptosis is shrinking of cells, where the loss of water is caused by an efflux of K⁺ and Cl⁻, the most abundant cytoplasmic ions (57). This volume loss, and in particular the drop in cytoplasmic K⁺ concentration, has been shown to activate both proteases and nucleases involved in programmed cell death, which are suppressed at normal K⁺ concentrations (58). Up-regulation of efflux through K⁺ channels has been associated with apoptosis in neurons (59), tumor cells (60), vascular smooth muscle cells (61), erythrocytes (62), T lymphocytes (63, 64), and many other cell types. ROS are common triggers for apoptosis, and in neutrophils they are known to activate formation of a death receptor complex that leads to a drop in cellular glutathione levels and programmed cell death (38, 65). Loss of the mitochondrial membrane potential, release of cytochrome c from the mitochondrial matrix, and the formation of mitochondrial ROS often accompany apoptosis, even in neutrophils, which do not depend on mitochondria for most metabolic activity (36, 52). Our findings that the SK opener 1-EBIO enhances staurosporine-mediated apoptosis (Fig. 7) clearly are consistent with potassium loss and ROS formation causing cell death. The SK channel pathway appears to be NADPH oxidase-independent, because 1-EBIO can trigger ROS even in gp91phox-deficient PLB-985-CGD cells (Fig. 5B), but its stimulation of ROS production could potentially be upstream of Cl⁻ channel activation (66).

One of the unexpected findings from this study is that a plasma membrane channel can affect intracellular ROS generation, most likely by mitochondria. Apamin, because it is membrane-impermeant, does not have access to organellar membranes except through endocytosis, so it most likely inhibits only plasma membrane channels. Because neither apamin nor 1-EBIO altered cytoplasmic calcium levels, either in resting cells or in fMLF-stimulated cells, it is unlikely that SK channels are affecting calcium release. A possible explanation is that the K⁺ efflux stimulated by 1-EBIO affects cellular energy homeostasis, which leads to a reduction in the NADPH/NADP⁺ ratio, and a more oxidizing intracellular environment. This environment could enhance ROS production and the oxidation of fluorescent ROS probes in compartments such as the mitochondria and endoplasmic reticulum, which maintain oxidizing environments. Activation of K⁺ efflux could, for example, alter the metabolic oscillations present in neutrophils (67) or increase glutathione efflux (68), as it has been found to do in airway epithelia (69). A proposed pathway summarizes these possibilities (Fig. 8, which is published as supporting information on the PNAS web site), but further experiments will be necessary to fully characterize the mechanism of SK channel regulation of granulocyte ROS production and apoptosis.

In conclusion, we have found that neutrophils and neutrophil-like PLB-985 cells express SK channels, and that these channels' activity regulates ROS production in granulocytes. Rather than modulating the classical NADPH oxidase-dependent pathway for ROS production, however, SK channels appear to affect ROS generation by mitochondria. Given the importance of ROS in triggering apoptosis and the critical role of K⁺ efflux in programmed cell death, one might expect that SK channels could be part of the apoptotic pathway. Indeed, we showed that activating SK channels enhanced staurosporine-induced apoptosis. This work demonstrates a role for a K⁺ channel in neutrophil programmed cell death and may suggest a target for antiinflammatory therapies to potentiate granulocyte clearance during the resolution of inflammation. Given the critical role of mitochondrial ROS in cell signaling and redox homeostasis, SK channels also may be important for regulating neutrophil turnover or differentiation.

Materials and Methods

Culture of PLB-985 Cells.

PLB-985 cells (University of California, San Francisco, Cell Culture Facility) and PLB-X-CGD cells (19) were cultured as previously described (19).

Purification of Human Neutrophils.

Fresh human blood collected under informed consent (University of California, San Francisco, Committee on Human Research approval no. H677-22987-03) was anticoagulated with heparin or citrate-dextrose solution (ACD; Sigma, St. Louis, MO), and neutrophils were isolated with Polymorphprep (Axis-Shield, Dundee, U.K.), according to the manufacturer's instructions. The neutrophil layer was removed and washed three times with Hanks' balanced salt solution (HBSS) −Ca²⁺/−Mg²⁺ with 20 mM Hepes (pH 7.3), and red blood cells were removed by hypotonic lysis.

Detection of SK Transcripts by RT-PCR.

Total RNA was isolated from PLB-985 cells, HL-60 cells, or neutrophils by using an RNEasy Mini Kit (Qiagen, Valencia, CA), according to the manufacturer's instructions. After purification, RNA was treated with DNase I (Ambion, Austin, TX) to digest genomic DNA. cDNA was made by synthesizing the first strand from RNA with oligo(dT) primers and Stratascript Reverse Transcriptase (Stratagene, La Jolla, CA), according to the manufacturer's protocol. The first strand then was used as a template for PCR with primers specific to human SK1, SK2, or SK3. Primer pairs (Integrated DNA Technologies, Coralville, IA) are reported in Supporting Text, which is published as supporting information on the PNAS web site.

Measurement of ROS Production by Neutrophils and PLB-985 Cells.

PLB-985 cells or neutrophils were resuspended in HBSS plus 0.2% BSA with either 500 nM DHR 123, 5 μM DHE, or 7.7 μM PF1 (11) in the presence or absence of drugs. Cell suspensions were rotated at 37°C for 10 min to allow uptake of fluorescent probes. To stimulate cells, 100 nM fMLF (Sigma-Aldrich, St. Louis, MO) was added and cells were further incubated for 10 min at 37°C before quenching on ice. For stimulation with opsonized bacteria, ≈4 mg of heat-killed S. aureus (Invitrogen, Carlsbad, CA) was incubated for 2–3 h in PBS plus 10 mg/ml human IgG (Sigma) at 37°C. Opsonized bacteria then were washed with PBS and added to neutrophils or PLB-985 cells at a ratio of ≈10–20 bacteria per granulocyte. The suspension was centrifuged briefly at 500 × g to promote binding and phagocytosis and then incubated again at 37°C for 10–15 min before quenching on ice. Fluorescence spectra were measured by using a Fluoromax-3 (Jobin Yvon Horiba, Edison, NJ). DHR 123 was excited at 501 nm, and emission was measured from 510–600 nm (531-nm maximum); DHE was excited at 492 nm with emission spectra collected from 525–650 nm (600-nm maximum); and PF1 was excited at 488 nm with emission spectra collected from 500–600 nm (518-nm maximum). Statistical significance was determined by ANOVA.

Measuring Apoptosis of PLB-985 Cells.

PLB-985 cells differentiated for 6 days with 0.5% DMF (10⁶ cells) were treated for 6 h at 37°C and 5% CO₂ with 500 nM staurosporine (EMD Biosciences, San Diego, CA) with or without 1-EBIO or apamin. Cells were washed and resuspended in cold PBS/10 nM Yo-pro-1/1 μg/ml propidium iodide (Invitrogen) and incubated on ice for 30 min. Fluorescence of 10,000 cells was analyzed by using a FACSort (BD Biosciences, San Jose, CA) flow cytometer. Statistical significance was determined by ANOVA.

Whole-Cell Patch Clamp Electrophysiology.

PLB-985 cells in a recording chamber superfused with HBSS external solution were used for whole-cell recordings at room temperature with patch electrodes (5–7 MΩ) filled with pipette solution containing 120 mM K-gluconate, 20 mM KCl, 2 mM MgCl₂, 5 mM Na₂ATP, 0.3 mM Na₃GTP, 1.1 mM EGTA, 5 mM Hepes, and CaCl₂ to bring free Ca²⁺ to 1 μM (pH 7.4) with KOH. Recordings were amplified with Axopatch 200B, filtered at 5 kHz, sampled at 20 kHz, and stored directly to the disk by using pClamp9 (Axon Instruments, Sunnyvale, CA). Cell capacitance was typically between 3 and 7 pF, and series resistance was <20 MΩ. Drugs were applied in bath solution by means of an N₂-pressurized perfusion system (ALA Scientific Instruments, Westbury, NY). Analysis of the digitized records was performed by using Clampfit9 software to obtain G/V curves.

Abbreviations

ROS

reactive oxygen species

1-EBIO

1-ethyl-2-benzimidazolinone

fMLF

formyl-methionine-leucine-phenylalanine

DPI

diphenylene iodonium

DHR 123

dihydrorhodamine 123

PF1

Peroxyfluor-1

DHE

dihydroethidium.