Simultaneous Measurement of Stimulus-induced Changes in Cytoplasmic Ca2+ and in Membrane Potential of Human Neutrophils*

The activation of human neutrophils by chemotactic peptides evokes a rapid chnge in membrane potential and an increase in cytoplasmic Ca” levels. These events are followed up to a minute later by detectable levels of microbicidal agents formed by the oxidative burst. Except for the latter, the sequence of events has remained unclear. We report here that a new fluorescent Ca2+ indicator developed by R. Tsien, Indo-1, has allowed us to resolve the temporal relationship be- tween the rapid and transient cytoplasmic Ca” rise and the membrane potential change and to do so on very small samples by using a fluorescence-activated cell sorter. We have adapted a FACS 440 for simulta- neous single cell membrane depolarization and cytoplasmic [Ca2+] detection in human neutrophils upon stimulation with formyl-methionyl-leucyl-phenylala-nine (fMLP). A membrane potential probe, dipentylox-acarbocyanine, allows us to determine that the mem- brane potential change is fMLP dose-dependent and apparently biphasic. The depolarization is maximal 40 s after stimulation. In contrast, cytosolic [Ca2+], while fMLP-dose dependent, is maximal at 10 s and already decreasing rapidly when the cell has reached its lowest potential. It can be measured with Indo-1 which has a fluorescence emission (Lx = 357 nm) maximum at 485 nm when Ca2+-free and 405 nm when Ca2+-liganded. The ratio of these fluorescences may then be calibrated

Simultaneous Measurement of Stimulus-induced Changes in Cytoplasmic Ca2+ and in Membrane Potential of Human Neutrophils* (Received for publication, February 5, 1986) Kristina G . Lazzari The activation of human neutrophils by chemotactic peptides evokes a rapid chnge in membrane potential and an increase in cytoplasmic Ca" levels. These events are followed up to a minute later by detectable levels of microbicidal agents formed by the oxidative burst. Except for the latter, the sequence of events has remained unclear. We report here that a new fluorescent Ca2+ indicator developed by R. Tsien, Indo-1, has allowed us to resolve the temporal relationship between the rapid and transient cytoplasmic Ca" rise and the membrane potential change and to do so on very small samples by using a fluorescence-activated cell sorter. We have adapted a FACS 440 for simultaneous single cell membrane depolarization and cytoplasmic [Ca2+] detection in human neutrophils upon stimulation with formyl-methionyl-leucyl-phenylalanine (fMLP). A membrane potential probe, dipentyloxacarbocyanine, allows us to determine that the membrane potential change is fMLP dose-dependent and apparently biphasic. The depolarization is maximal 40 s after stimulation. In contrast, cytosolic [Ca2+], while fMLP-dose dependent, is maximal at 10 s and already decreasing rapidly when the cell has reached its lowest potential. It can be measured with Indo-1 which has a fluorescence emission (Lx = 357 nm) maximum at 485 nm when Ca2+-free and 405 nm when Ca2+-liganded. The ratio of these fluorescences may then be calibrated in terms of cytoplasmic Ca2+ levels. Thus, Ca2+ release into the cytoplasm becomes the earliest evidence of neutrophil stimulation by fMLP and occurs in close association with an apparent membrane hyperpolarization.
The stimulus response of human neutrophils has been studied extensively in vitro in cell suspensions and has been shown to involve changes in transmembrane cation gradients within seconds of exposure to soluble or particulate stimuli (1)(2)(3)(4)(5). Simultaneous measurement of these changes, however, has been difficult. The fluorescence-activated cell sorter (FACS') has recently been useful in investigating changes in membrane potential alone (6-8) or simultaneously with oxidative product formation (9). In these experiments, the data * This work was supported in part by National Institutes of Health Grants HL07501, AM31056, HL33565, HL19717B, and 1-S10-RR01949. 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.
for individual cells were collected and analyzed after a minimum time interval of 1.5 min, significantly prolonged when compared to the speed of the neutrophil response. We report here that a FACS may be used to monitor rapidly (at 10-s intervals) and simultaneously the chemotactic peptide formylmethionyl-leucyl-phenylalanine-induced changes in neutrophil cytoplasmic Ca2+ concentrations and membrane potential as a function of stimulus concentration and elapsed time.
Membrane potential changes, oxidative burst product formation, and degranulation are among the earliest stimulus responses of the neutrophil and are parameters often used as measures of cell activation. Cytoplasmic Ca2+ transients have recently been reported to occur within seconds of ligand binding and may be involved in the ensuing activation of the neutrophil (10-14). It is not yet known whether any of these changes are essential for cell activation, nor has their temporal interrelationship been elucidated. However, the importance of Ca2+ in the process of signal transduction from membrane receptors to the demonstration of a cellular response remains undefined. Evidence obtained from studies of unopsonized Candida albicans hyphae stimulation of human neutrophils suggests that a Ca'+ signal can occur when depolarization and degranulation remain absent.' These results imply that the Ca2+ signal cannot be sufficient to stimulate complete activation of the cellular functions. It has only recently become possible to study the nanomolar levels of Ca'+ within a viable cell without significant perturbation of the system through the development of cell-permeant, fluorescent Ca'+ indicators such as Quin 2, Fura-2, and Indo-1 (16)(17)(18). Whereas useful information has been obtained from Quin 2, the probe suffers from two severe disadvantages: a low extinction coefficient and quantum yield and a significant Ca2+ buffering capability which obscures small intracellular Ca2+ transients. Other interferences such as instrument variation, intracellular dye concentration, and cell thickness also influence measurements of absolute Ca'+ concentration with Quin 2 (17-20). Indo-1 and Fura-2 exhibit a spectral shift upon binding to Ca2+ and therefore do not have many of the problems associated with Quin 2. The ratio of fluorescent intensities at two suitable wavelengths may then be used to calculate [Ca2+]i independently of variations in dye concentration or path length (16,20).
To date, the analyses of intracellular Ca2+ concentration changes have involved spectrofluorometric measurements in cell suspensions which require averaging over lo6 cells/ml.
Since separate experiments must be performed for each parameter being measured, it has been difficult to resolve the sequence of events occurring within the first 30 s after exposure of neutrophils to a specific stimulus. We report here success in resolving this issue, utilizing a flow cytometer equipped with dual lasers and three detection photomultipliers to determine membrane potential and intracellular Ca2+ concentration changes simultaneously within 10 s of stimulation.
We have analyzed the kinetics of Ca2+ fluxes and membrane potential changes upon stimulation of neutrophils using a FACS and have resolved the temporal sequence of events, beginning within 10 s of stimulation. The cytoplasmic Ca2+ concentration peaks within 10 s, whereas the membrane potential change reaches a maximum after 60 s of exposure to formyl-methionyl-leucyl-phenylalanine. Both effects are dose-dependent, with [Ca2+Ii increasing with fMLP concentration to saturation at M, whereas the membrane potential changes within 10 s from more negative (hyperpolarized) at low doses of fMLP (<IO-' M) to less negative (depolarized) at higher doses (>lo-' M).

MATERIALS AND METHODS
Leukocyte Isolation-Peripheral blood polymorphonuclear leukocytes from citrated venous blood of normal human subjects were prepared by dextran sedimentation followed by Ficoll-Hypaque gradient centrifugation and hypotonic lysis to remove contaminating erythrocytes (27). The resulting cells (95% polymorphonuclear leukocytes) were suspended in phosphate-buffered saline (125 mM NaCl, 8 mM Na2HP04, 2 mM NaH2P04.H20, 5 mM KCl, 5 mM glucose), pH 7.4, and stored on ice in a Ca2+-and Mg2+-free medium to minimize aggregation. Cell viability was determined by trypan blue exclusion.
Zndo-l/Acetoxymethyl Ester Loading-The Ca2+ indicator Indo-l/ acetoxymethyl ester (Molecular Probes, Junction City, OR) was diluted to 1 mM stock in dimethyl sulfoxide, aliquoted, and stored desiccated at -20 'C. Cells were loaded with the acetoxymethyl ester of the dye by an adaptation of the technique used for @in 2 (17). Polymorphonuclear leukocytes at 1 X 107/ml in PBS were incubated .with 5 p~ Indo-l/acetoxymethyl ester at 37 "C for 7 min. The suspension was diluted 5-fold in cold PBS, centrifuged at 150 X g at 4 "C for 10 min to wash out any free dye, and resuspended in PBS to approximately 5 X 107/ml. The amount of dye remaining in the cell suspension after loading was calibrated according to the fluorescence titration curve of the cell-impermeant (hydrolyzed) form of Indo-1 in a solution of digitonin-lysed polymorphonuclear leukocytes. No significant leakage of dye was detectable in cells stored on ice for up to 2 h after loading. Cell viability remained unperturbed.
Measurements of Cytosolic Ca" in Cell Suspensions-Intracellular Indo-1 fluorescence was monitored in cell suspensions of 1 X IO6 cells diluted in 1 ml of PBS containing 1.0 mM Ca2+ and 1.5 mM M F on a Perkin-Elmer 650/10 fluorometer in a thermostated l-cm cuvette, with stirring at 37 "C. The dye, when excited at 355 nm, exhibits two emission peaks, one at 405 nm which increases with higher Ca2+ binding and one at 485 nm which decreases with higher Ca2+ binding. Since the binding of CaZ+ shifts the relative emission intensities of the dye at these two wavelengths, the ratio of these intensities was used to quantitate cytosolic Ca2+ according to the method of Gryn-  (22,26). The fluorescence distributions for Indo-1 and for di-O-C,(3) were obtained using a FACS 440 dual laser system (Becton Dickinson, Mountainview, CA) equipped with a digital PDP-11/23 microcomputer (Digital Corp., Maynard, MA). A primary %watt argon laser tuned to a wavelength of 488 nm was used to excite the di-o-c5(3), whereas a secondary 4-watt argon ion laser tuned to 357 nm was used to excite the Indo-1. A 20-ms delay between the primary laser and secondary laser was employed in order to distinguish the fluorescence emitted from the same cell. Three fluorescence emission peaks 405,485, and 575 nm were monitored using 405/20, 485/22, and 575/26 band-pass filters at the respective photomultipliers. All fluorescence parameters were collected using the linear mode in order to calculate mean channel fluorescence ratios. Data were collected for 2000 resting cells in order to establish a baseline distribution. Stimulus was then added directly to the sample, and data were collected, 2000 cells/sample, at 10-5 intervals for 3 min. To determine the extent of depolarization, the difference between the mean channel fluorescence of stimulated cells and unstimulated resting cells at 575 nm was calculated as a percentage of the unstimulated control cell mean fluorescence at each time point. To measure changes in cytoplasmic Ca2+, the ratio of mean channel fluorescence at 405 and 485 nm in the resting cell was calculated and subtracted from the ratio after stimulation. The per cent change in ratio was calculated and compared to the calibration curves obtained as described above in order to calculate the per cent change of [Ca2+Ii.

RESULTS
Alterations in the membrane potential can be measured by changes in the intracellular concentration of fluorescent, lipophilic cations such as the cyanines (21). For observations of cell suspensions, it is preferable to use probes such as di-S-C3(5) which self-associate and hence are quenched inside the cell so that only the residual, extracellular probe fluorescence is observed as an indicator (by difference) of the probe's distribution between the cells and the suspension medium (2, 22, 23). Correlation with membrane potential has been well established with other types of membrane potential measurements (24, 25). In contrast, for observations of single cells in a flow cytometer, a non-self-associating probe such as di-0-C,(3) or di-I-c3(3) must be used such that the observed fluorescence is directly proportional to the internal probe concentration (7, 9,26).
As Fig. 1 (right) indicates, the cytoplasmic Ca2+ increase reaches a maximum by 10 s. In contrast, whereas the membrane potential changes as rapidly (Fig. 1, left), the change appears to be biphasic. At very low fMLP concentrations, one observes a rapid increase in fluorescence which appears t o represent a hyperpolarization of the cell. At higher fMLP concentrations, this hyperpolarization is followed by a slower, but larger loss in fluorescence corresponding to a membrane depolarization which, at high fMLP concentrations, tends to obscure the hyperpolarization and which is maximal at approximately 60 s. Calculations of the concentration of intracellular, cytoplasmic Ca2+ indicate a resting level of approximately 100 nM free Ca2+ in cells suspended in 1.0 mM external CaClz and a maximal increase to approximately 3 PM Ca2+ 10 s after stimulation with a saturating dose of fMLP (Fig. 4).

DISCUSSION
Changes in transmembrane potential are among the earliest detectable events following ligand-receptor binding and have been interpreted by some as being necessary but not sufficient for the formation of the oxidative metabolic response (2-4, 22, 28). Calcium has been strongly implicated as a second messenger in this process of cell activation (10,35)' and is also thought to play a role in the regulation of cytoskeletal protein polymerization (6, 30). Increases in cytosolic [Ca"] have been demonstrated to be involved in, but insufficient for, the triggering of depolarization, degranulation, or superoxide production (12).' Evidence that cell stimulation occurs in the presence of a chelator of extracellular Ca2+ has indicated that an influx of divalent cations is not essential for activation, and an intracellular calcium reservoir appears to be responsible for the cytosolic Ca2+ increase observed after stimulation by fMLP (12,14,31,35).
Studies of chronic granulomatous disease patients whose neutrophils, in most cases, are unable t o produce a normal oxidative response or a normal depolarization have indicated that these two responses are interrelated (2,22,32), whereas in all cases, the influx in intracellular Ca'+ upon stimulation appeared to be normal (32). From other studies involving the artificial reduction of transmembrane potential using high K+/low Na+ buffers (3,33) and from the measurement of simultaneous depolarizations and oxidative product formation by flow cytometry (9), it appears that receptor-mediated depolarization, whereas occurring before the oxidative burst, is not by itself sufficient to trigger the oxidative metabolic response.
The temporal relationship between increases in [Ca2+Ii and depolarization has been less clear. Chemotactic factor-induced increases in cytoplasmic Ca2+ have been detected previously in human neutrophils and are confirmed here (11). However, these experiments were performed separately (not simultaneously) on cell suspensions, where it is technically difficult to observe small changes within the cells. We demonstrate here, by means of a FACS, an increase in [Ca2+Ii which occurs almost instantly, peaking within 10 s of activation, simultaneously with an increase in transmembrane potential, and long before the later depolarization(and still later detection of oxidative products). These responses may, therefore, represent separate events in the neutrophil activation process. Simultaneous measurements have not been previously achieved.
The biphasic membrane potential response to activation has been noted by other investigators, both in neutrophils (9, 23, 26) and macrophages (34). In other studies, the apparent hyperpolarization has been attributed to cell population heterogeneity or probe-dependent artifacts (7). Whereas it is possible that the apparent initial rise in di-0-C5(3) shown here represents an artifact, the following evidence suggests that there is indeed a rapid hyperpolarization of the cell. The observed hyperpolarization was not dependent on di-O-CJ3) concentration and therefore is unlikely to be related to intracellular probe distribution. The hyperpolarization has also been observed in cell suspensions on a spectrofluorometer which records a response averaged over 2 million cells (data not shown) and therefore does not appear to be a consequence of population heterogeneity. Furthermore, we have reported elsewhere3 that freshly isolated human granulocyte precursor Sullivan, R., Griffin, J. D., Melnick, D. A., Meshulam, T., Simons, E. R., Lazzari, K. G., a n d P r o t o , P. (1986) J. Biol. Chern., submitted for publication. cells exhibit a similar increase in di-0-C5(3) fluorescence (measured on a FACS) in response to saturating doses (10" M) of fMLP. As these cells differentiate in culture over a period of 8 days, their response to fMLP stimulation gradually becomes more characteristic of a mature neutrophil; the detectable initial hyperpolarization decreases as the later, larger depolarization increases. In view of these findings, we believe that there may indeed be a hyperpolarization occurring within the first 10 s of stimulation which is observable only at very low doses of fMLP and is indicative of a biphasic response.
The studies reported here demonstrate that, by using a FACS and simultaneous detection of Ca2+ concentration and membrane potential changes, we can document a fMLP dosedependent rise in cytoplasmic Ca2+ and a change in membrane potential which are uniform in the entire population of human neutrophils. At low doses, the cells appear to hyperpolarize as the cytoplasmic Ca2+ increases until both responses reach a maximum at 10 s. With increasing fMLP, the maximum [Ca2+Ii change increases, but the apparent hyperpolarization decreases as the slower depolarization becomes larger. These events taken together indicate that the [Ca2+Ii increase is independent of and precedes the fMLP-induced depolarization of human neutrophils, but may be concomitant with a faster hyperpolarization.