CELLULAR IMMUNOLOGY 95,46-53 (1985)

Inhibition of Macrophage Activation by Calcium Channel Blockers and Calmodulin Antagonists

B. WRIGHT,',* I. ZEIDMAN,~ R. GREIG,.$+~ AND G. POSTE~$

Departments of *Pharmacology and iPathology and Laboratory Medicine, University of Pennsylvania, Philadelphia, Pennsylvania 19104 and $Department of Tumor Biology. Smith Kline and French

Laboratories, 1500 Spring Garden Street, Philadelphia, Pennsylvania 19101

Received February 7, 1985; accepted May 7, 1985

The biochemical mechanisms by which macrophages become activated to the tumoricidal state are poorly understood. To investigate the role of calcium in this process, the effect of calcium channel blockers and calmodulin antagonists on the acquisition of tumoricidal properties by macrophages activated by a number of different agents was examined. Activation of thioglycollate- stimulated C57BL/6 mouse peritoneal macrophages by macrophage activation factor (MAF) plus LPS, IFN--r plus LPS or the calcium ionophore, A23187, was inhibited in a dose-dependent fashion by the calcium channel blockers nifedipine and verapamil. These agents blocked the influx of 4’Ca into macrophages activated by MAF plus LPS. Macrophage activation was also inhibited by chlorpromazine, W-7, and calmidaxolium at concentrations known to perturb cal- modulin function. The data suggest that activation of macrophages to the tumoricidal state is a calcium-dependent process involving the participation of calcium-regulated biochemical reactions whose activities can be modulated by pharmacological agents that bustrate transmembmne calcium fluxes and/or inhibit calmodulin function. 8 1985 Academic PIW, IIIC.

INTRODUCTION

Numerous structurally diverse materials have been shown to activate macrophages to the tumoricidal state. However, little is known about the biochemical mechanisms involved in the activation process. Previous studies have established a role for calcium in the activation of other immune cell types, including T lymphocytes (l-3) and NK cells (4), but the role of calcium ions in macrophage activation has received less at- tention. However, recent observations showing that the calcium ionophore A23 187 can enhance acid phosphatase activity in macrophages (5) and enhance secretion of cytolytic factor from macrophages (6) raise the possibility that calcium may be an important biochemical mediator in macrophage activation. In this study we have investigated the effects of calcium channel blockers and calmodulin antagonists on macrophage activation in vitro. Calcium channel blockers, such as nifedipine and verapamil, disrupt calcium fluxes across cell membranes and are used clinically in the treatment of several cardiovascular diseases (7). Recently, these agents have been shown to antagonize T-cell mitogenesis (3), to inhibit mitogen-induced stimulation of inter- feron production by T cells (8), and to frustrate NK cell activity (4). Calmodulin

r Present address: Merck Sharp & Dohme Research Laboratories, West Point, PA 19486. 2 To whom reprint requests should be addressed.

46

0008-8749185 $3.00 Gqyi$ttO1985byAcadmicF’nqlnc. All riehts of repmduaion in any form rcsmed.

CALCIUM AND MACROPHAGE ACTIVATION 47

antagonists bind to and disrupt the calcium-binding functions of this molecule (9, 10) and calmodulin antagonists such as the phenothiazines have been shown to inhibit macrophage migration, T-cell mitogenesis, and NK cell cytotoxicity ( 1 l- 13). In this communication we report that activation of macrophages to the tumoricidal state induced by different classes of activating agents is antagonized both by calcium channel blockers and calmodulin antagonists.

MATERIALS AND METHODS

Reagents. Nifedipine (Pfizer, New York, N.Y.) was prepared in ethanol and diluted in DMEM3 before use. Because of its photosensitivity, experiments using this agent were performed in the dark. Verapamil (Knoll, Whippany, N.J.) was dissolved in ethanol and diluted in DMEM. Thorazine brand of chlorpromazine HCl (Smith Kline and French Laboratories, Philadelphia, Pa.) was diluted from the commercial phar- maceutical formulation and prepared in DMEM. The calmodulin antagonists, W-5 and W-7 (Rikaken, Japan), were dissolved in PBS and diluted in DMEM. A23187 (Calbiochem, La Jolla, Calif.) and calmidazolium (Janssen Pharmaceutics, Beerse, Belgium) were dissolved in DMSO (Sigma, St. Louis, MO.). Stock solutions of reagents were stored at 0°C and diluted in DMEM immediately before use. MAF was prepared as described ( 14). LPS (Escherichia coli 0 127:B8; Difco, Detroit, Mich.) was dissolved in sterile, endotoxin-free water at 50 &ml and stored at 0°C. Dilutions were in DMEM. Murine Interferon-y produced by recombinant DNA technology was supplied by Genentech (San Francisco, Calif.).

Macrophage cultures. Six- to eight-week-old C57BL/6 female mice (Charles River, Portage, Mich.) were injected ip with 2 ml sterile thioglycollate broth (Difco, Detroit). Five days later, animals were sacrificed and peritoneal exudate cells obtained by lavage of the peritoneal cavity with two lo-ml washes with cold HBSS. The harvested cells were centrifuged at 1000 rpm for 10 min at 4°C washed once in PBS (Gibco), resus- pended at 1.5 X 1 06/ml in DMEM, and 1 00-~1 aliquots plated into individual wells of a 96-well microtiter plate (Costar, Cambridge, Mass.). After incubation for 1.5 hr at 37°C to permit the selective adherence of macrophages, cultures were rinsed twice with DMEM to remove unattached cells. The remaining adherent cell population contained >95% macrophages as assessed by phagocytosis of carbon particles and mononuclear morphology (15).

Macrophage activation. Macrophage cultures were activated to the tumoricidal state by incubation in DMEM containing MAF diluted 1:5 in DMEM plus LPS (50 ng/ ml), IFN-y (5 U/ml) plus LPS, or A23 187 ( 1O-6 to 10e9 M) for 24 hr. Control mac- rophages were exposed to DMEM alone. Pharmacological agents were added to mac- rophage cultures immediately before addition of activating stimulus. The extracellular calcium concentration in the activation medium was 1.8 mM as determined by atomic absorption spectroscopy using an inductively coupled plasma emission spectrometer (Model JY-38P, Instruments SA, Inc.). Following the 24-hr activation period, mac- rophage cultures were washed three times with DMEM to remove test drugs and activating agents and assayed for their tumoricidal activity.

3 Abbreviations used: DMEM, Dulbecco’s modified Eagle’s medium; MAF, macrophage activation factor; LPS, lipopolysaccharide; PBS, calcium-, magnesium-free phosphate-buffered saline; DMSO, dimethyl sulf- oxide; FCS, heat-inactivated fetal calf serum; IFN-y, recombinant murine Interferon-y; HBSS, calcium-, magnesium-free Hank’s balanced salt solution; M$, macrophage.

48 WRIGHT ET AL.

Macrophage cytotoxicity. This was performed as described previously ( 14). Briefly, 5 X IO3 B 16FlO clone 1 melanoma cells syngeneic to the C57BL/6 mouse, labeled with 0.25 &i/ml ‘251UDR (sp act 2000 Ci/mmol; New England Nuclear, Boston, Mass., 0.2 cpm/cell) were added to macrophage cultures at an effector:target cell ratio of 30: 1. Tumor cells were plated alone as an additional control for spontaneous isotope release. After the 72-hr coculture period at 37°C in a humidified 95% air:5% CO2 atmosphere, the assay was terminated by washing the wells twice with PBS and lysing the remaining adherent cells with 100 ~1 1 A4 NaOH. The lysate was absorbed onto individual fiber filters (Flow Labs, McLean, Va.) and the remaining cell-associated radioactivity counted in an LKB gamma counter. The percentage specific cytotoxicity was calculated as follows ( 15):

cpm ‘251UDR targets + control M$ - cpm 12’IUDR targets + test M4 x loo cpm ‘251UDR target cells + control M4

The results represent data obtained from at least three experiments. Analysis of sta- tistical significance was performed using Student’s t distribution test.

“Caj7ux studies. Macrophages (1.5 X lo6 cells/ml) activated with MAF plus LPS were washed three times with warm (37°C) PBS and suspended in DMEM + 5% FCS (1.8 mA4 Ca2+) containing 2 &i/ml 45Ca (45CaC12, New England Nuclear, sp act 4- 50 Ci/g) with or without 10e6 A4 verapamil (or nifedipine). Aliquots (100 ~1) of sus- pension were removed at different times and washed rapidly three times on a Whatman filter with cold DMEM + 5% FCS (1.8 m&f Ca2+). Filters were air dried and radio- activity was measured in a Beckman liquid scintillation counter. Each time point was determined in triplicate.

Macrophage viability. Viability was assessed by monitoring macrophage phagocytic capacity (18) and RNA and protein synthetic effort following a 24-hr exposure to the different pharmacological agents. Cultures were radiolabelled with [3H]uracil (New England Nuclear; sp act, 40-50 Ci/mmol) or [35S]methionine (New England Nuclear; sp act, 40-50 mCi/mmol) for 4 hr and precipitable radioactivity was measured in a Beckman liquid scintillation counter.

Protein measurements. Triplicate protein determinations were performed according to the method of Lowry et al. using bovine serum albumin as a standard (16).

Assay for endotoxin contamination. All culture media and reagents used were de- tectably free of bacterial lipopolysaccharide (endotoxin) by the Limulus amoebocyte lysate assay which is capable of detecting 0.1 rig/ml of endotoxin (17).

RESULTS

Inhibition of macrophage activation by calcium channel blockers. Acquisition of tumoricidal properties in macrophage populations activated with either MAF plus LPS, IFN-y plus LPS, or the calcium ionophore, A23187, was inhibited in a dose- dependent manner by verapamil (Fig. 1). At the highest concentration of verapamil ( lob6 M), 90% inhibition of activation by MAF plus LPS and IFN-y plus LPS was achieved without reduction in macrophage viability. However, activation by A23 187 ( 10m6 M) was less sensitive to antagonism by verapamil. Identical data were obtained with a similar dose range of nifedipine (data not shown).

Inhibition of 45Ca flux in macrophages by calcium channel blockers. Verapamil ( 10m6 M) antagonized the influx of 45Ca into macrophages treated with MAF plus

CALCIUM AND MACROPHAGE ACTIVATION 49

MAF+LPS m-y + LP3 A23107

100

80

z F%o m

5 40 - 3

20

0 IO-6 10-7 lo‘* 10% lo-’ lo- 10% IO -7 10-a

MRAPAMIL (M)

FIG. I. The effect of the olcium channel blocker, verapamil, on activation of mouse peritoneal macrophages by different activation stimuli. The results represent mean values from three experiments and measurements on three samples within each experiment. Bars are standard errors. Each data point displaying greater than 10% inhibition was statistically different (P < 0.05) from control experiments conducted in the absence of drug (0% inhibition).

LPS (Fig. 2). Similar data were obtained for nifedipine ( 10e6 M; data not shown). Verapamil and nifedipine had no detectable effect on 45Ca efflux from macrophage cultures challenged with MAF plus LPS, IFN-y plus LPS or A23 187 (data not shown).

Inhibition of macrophage activation by calmodulin binding drugs. Chlorpromazine, W-7, and calmidazolium produced a dose-dependent inhibition of macrophage acti- vation by MAF plus LPS (Fig. 3). Complete inhibition of activation was obtained using chlorpromazine or W-7 ( 1 Om6 M) but calmidazolium at the same concentration caused only a 45% decrease in activation. These agents also inhibited macrophage activation by IFN-7 plus LPS, though in this case W-7 was the least effective of the three inhibitors (Fig. 4). Variation in the sensitivity of macrophage activation to al- teration by different calmodulin antagonists depending on the activation stimulus is further illustrated by the finding that the activation of macrophages challenged with

FIG. 2. The effect of verapamil ( lo4 M) on ?a uptake by mouSe peritoneal macrophages activated with MAF plus LPS. Control (0); verapamil(0). The results represent mean values from three experiments and measurements on three samples within each experiment. Bars are standard errors. The differences in the two curves are statistically significant (P < 0.05) at each time point (except 5 min).

50 WRIGHT ET AL.

CHLORPROMAZINE CALMIDAZDLIUM

CONCENTRATION (M 1

FIG. 3. The effect of calmoduhn antagonists on activation of mouse peritoneal macrophages by MAF plus LPS. The results represent mean values from three experiments and measurements on three samples within each experiment. Bars are standard error. Cytotoxicity of MAF plus LPS activated macrophages was 57%. Each data point displaying greater than 10% inhibition was statistically different (P < 0.05) from control experiments conducted in the absence of drug (0% inhibition).

an optimal dose of A23 187 ( 10m6 M) was decreased by only 25% following exposure to chlorpromazine ( 10d6 M), while W-7 and calmidazolium were highly effective, causing a 90% reduction in activation at the highest drug concentration tested ( low6 M; Fig. 5). W-5, a chlorine-deficient derivative of the parent naphthalene sulfonamide W-7 which interacts only weakly with calmodulin (19), displayed weak inhibition of macrophage activation by any activating stimulus (data not shown).

Macrophage viability. Over the concentration range at which the pharmacological agents used in this study were effective in antagonizing macrophage activation, parallel drug-treated cultures retained full phagocytic capacity and displayed unimpaired RNA

CHLORPROMAZINE w-7 CALMIDAZOLIUM

90

80

70

560 F = 50 r g 40

8 30

20

IO

0 10-S lo-’ lo-’

CONCENTRATION (Ml

FIG. 4. The effect of calmodulin antagonists on activation of mouse peritoneal macrophages by IFN-7 plus LPS. The results represent mean values from three experiments and measurements on three samples within each experiment. Bars are standard error. Cytotoxicity of IFN-y + LPS activated macrophages was 65%. Each data point displaying greater than 10% inhibition was statistically different (P < 0.05) from control experiments conducted in the absence of drug (0% inhibition).

CALCIUM AND MACROPHAGE ACTIVATION 51

CHLORPROMAZINE w-7 CALMIDAZOLIUM

CONCENTRATION (MI

FIG. 5. The effect of calmodulin antagonists on activation of mouse peritoneal macrophage by A23 187. The results represent mean values from three experiments and measurements on three samples within each experiment. Bars are standard error. Cytotoxicity of A23187 activated macrophages was 80%. Each data point displaying greater than 10% inhibition was statistically different (P < 0.05) from control experiments conducted in the absence of drug (0% inhibition).

and protein synthesis. Significant toxicity was observed only at concentrations greater than 1 O-’ M (data not shown).

DISCUSSION

The results obtained in the present study suggest that activation of tumoricidal properties in mouse peritoneal macrophages induced by different activating agents is a calcium-dependent process. This conclusion is supported by data showing that mac- rophage activation induced by a variety of stimuli is inhibited in a dosedependent manner by the calcium channel blockers, verapamil and nifedipine, and by agents that bind to and disrupt calmodulin function.

Verapamil and nifedipine are structurally unrelated calcium channel blocking drugs that are employed clinically for their antianginal and antiarrhythmic activities. Their in vitro and in vivo pharmacological effects have been attributed to their ability to recognize and bind specifically to calcium channels located in the plasma membrane and thereby frustrate transmembrane calcium fluxes [for review see (7)]. Other phar- macological actions, particularly a-adrenergic affects, have also been associated with these agents (20) and thus we cannot unequivocally attribute the antagonism of mac- rophage activation by these agents to their calcium channel blocking activity. However, the present investigation has shown that at nontoxic concentrations which inhibit macrophage activation, verapamil and nifedipine block calcium uptake into activated macrophages. Under identical conditions these agents failed to affect 45Ca efflux. The observation that verapamil ( 1 Oe6 M) antagonizes macrophage activation by 90% while causing only a 30% inhibition of calcium uptake suggests a complex stoichoimetric relationship between intracellular calcium levels and expression of macrophage tu- moricidal capacity. Alternatively, expression of tumoricidal activity might depend not only on absolute intracellular calcium levels but also on the compartmentalization of calcium within different subcellular organelles, a process that may also be affected by calcium channel blockers (7). In addition, since macrophages are functionally heter- ogenous (2 1) it is possible that the macrophage subpopulation capable of being rendered tumoricidal is considerably more sensitive to the calcium-blocking drugs than the macrophage population at large. Experiments at the single cell level will be required to resolve this issue.

52 WRIGHT ET AL.

The present results extend previous reports demonstrating alterations in macrophage function in vitro after treatment with calcium channel blockers. For example, con- canavalin A-stimulated release of superoxide anion from rat alveolar macrophages is inhibited by methoxyverapamil(22). Verapamil at concentrations much higher than those employed here (100 N) blocks production of a cytolytic factor by macrophages treated with A23 187 (23). Collectively, these observations and the present data suggest that macrophages possess functional calcium channels whose activity is intimately linked with the biochemical reactions responsible for the development and expression of macrophage tumoricidal properties.

Calmodulin is an intracellular protein that regulates a number of calcium-dependent processes (9, 10). Pharmacological interference with calmodulin activity in vitro has been shown to inhibit NK cell activity ( 13,24) and T-cell mitogenesis ( 12). The present results provide additional evidence to suggest that calmodulin may be important in cell-mediated immune reactions by showing that macrophage activation is blocked by a number of structurally distinct calmodulin antagonists at concentrations that are nontoxic to macrophages and that correspond well to their order of potency in inhibiting calmodulin activity in cell-free preparations (19, 35, 26). Calmidazolium, which has a reported I& of 5 X 10m9 M for calmodulindependent activation of brain phos- phodiesterase (26), and which blocks Epstein-Barr virus infection of B lymphocytes at 1.5 X low6 M (27), proved to be the most reproducible and effective inhibitor of macrophage activation in the present study. At the lowest concentration tested ( low8 M) this agent produced a greater than 50% inhibition of activation by MAF plus LPS. W-7 was an effective antagonist of macrophage activation elicited by MAF plus LPS but was considerably less potent in blocking activation induced by IFN-y plus LPS. While these data may suggest that MAF and IFN-7 provoke alterations in macrophage function through distinct calcium-dependent mechanisms (W-7-sensitive and retractory mechanisms, respectively) it is also possible that W-7, at its effective concentration (lo+ M), is acting nonspecifically.

The ability to use pharmacologic probes to modulate specific aspects of macrophage activation offers a new experimental approach for identification of the biochemical pathways involved at different stages of the activation process. The present results using agents that perturb cellular calcium homeostasis via effects on the transmembrane transport of calcium ions and calmodulin function suggest that calcium may be an important mediator in the activation process. Further evidence to support this hy- pothesis has been obtained recently in preliminary experiments demonstrating that the calcium agonist, Bay K8644 (28) at lo-* M was able to induce signilicant mac- rophage tumoricidal activity (Wright, Greig and Poste; unpublished observations). Additional studies to examine the effects of immunoregulatory agents on the level and distribution of intracellular calcium in macrophages in distinct stages of activation and the possible role of changes in phosphoinositol lipid metabolism in mobilizing calcium ions (29, 30) following exposure to activation agents are in progress.

ACKNOWLEDGMENTS

We thank Pfizer Pharmaceuticals, New York, N.Y., for supplying nifedipine, Knoll Pharmaceuticals, Whippany, N.J., for verapamil, Genentech for murine Interferon-y, and Nancy E. Signora for excellent secretarial assistance.

CALCIUM AND MACROPHAGE ACTIVATION 53

REFERENCES

1. Luckasen, J. R., White, J. M., and Kersey, J. H., Proc. Natl. Acad. Sci. USA 71, 5088, 1974. 2. Plaut, M., Bubbers, E., and Henney, C. S., J. Immunol. 116, 150, 1976. 3. Birx, D., Berger, M., and Fleisher, T. A., J. Immunol. 133,2904, 1984. 4. Hiserodt, J. C., Britvan, L. J., and Targan, S. C., J. Zmmunol. 129,2266, 1982. 5. Hand, W. L., King, N. L., Johnson, J. D., and Lowe, D. A. Nature (London) 265, 543, 1977. 6. Drysdale, B. E., Zacharchuk, C. M., and Shin, H. S., J. Zmmunol. 131,2362, 1983. 7. Janis, R. A., and Triggle, D. J., J. Med. Chem. 26,775, 1983. 8. Dianzani, F., Capobianchi, M. R., and Facchini, J., J. Virol. 50,964, 1984. 9. Cheung, W. Y., Science 207, 19, 1980.

10. Means, A. R., and Dedman, J. R., Nature 285, 73, 1980. 11. Onozaki, K., Takenawa, T., Homma, Y., and Hashimato, T., Cell Immunol. 75,242, 1983. 12. Cheung, R. K., Grinstein, S., and Gelfand, E. W., J. Immunol. 131,2291, 1983. 13. Rochette-Egly, C., and Touey, M. G., Biochem. Biophys. Rex Commun. 121,478, 1984. 14. Fidler, 1. J., and Raz, A., “Lymphokine Reports” (E. Pick, ed.), Vol. 3, p. 345. Academic Press, New

York, 1981. 15. Poste, G., Kirsh, R., Fogler, W. E., and Fidler, I. J., Cancer Res. 39, 881, 1979. 16. Lowry, 0. H., Rosenbrough, N. J., Farr, A. L., and Randall, R. J., J. Biol. Chem. 193,265, 1951. 17. Pabst, M. J., Hedegaard, H. B., and Johnston, R. B., Jr., J. Immunol. 128, 123, 1982. 18. Tucker, S. B., Pierre, R. V., and Jordan, R. E., J. Immunol. Methods 14,267, 1977. 19. Hidaka, H., Asano, M., and Tanaka, T., Mol. Pharmacol. 20,571, 1981. 20. Fairhurst, A. S., Whittaker, M. L., and Ehlert, F. J., B&hem. Pharmacol. 29, 155, 1980. 21. Koestler, T. P., Kirsh, R., Kline, T., Rieman, D., Greig, R., and Poste, G., Cell. Immunol. 97, 1, 1984. 22. Forman, H. J., and Nelson, J., J. Appl. Physiol. 54, 1249, 1983. 23. Drysdale, B. E., and Shin, H. S., Fed. Proc. 43, 1744, 1984. 24. Moon, T. D., Morley, J. E., Vessella, R. L., and Lange, P. H., &and. J. Immunol. 18,255, 1983. 25. Hidaka, H., Sasaki, Y., Tanaka, T., Endo, J., Ohno, S., Fugii, Y., and Nagata, T., Proc. Natl. Acad. Sci.

USA 78,4354, 1981. 26. Van Belie, H., Cell Calcium 2,483, 198 I. 27. Nemerow, G. R., and Cooper, N. R., Proc. Natl. Acad. Sci. USA 81,4955, 1984. 28. Schramm, M., Thomas, G., Towart, R., and Franchowiak, G., Nature (London) 303,535-537, 1983. 29. Berridge, M. J., Bio/Technology 2, 541, 1984. 30. Wightman, P. D., and Raetz, C. R. H., J. Biol. Chem. 259, 10048, 1984.