Induction of metallothionein synthesis in human peripheral blood leukocytes

ENVIRONMENTAL RESEARCH 42, 377-385 (1987)

Induction of Metallothionein Synthesis in Human Peripheral Blood Leukocytes DUANE L. PEAVY* AND EDWARD J. FAIRCHILD II *School o f Allied Health Sciences and the School o f Public Health, University o f Texas Health Science Center at Houston, Houston, Texas 77225 Received March 6, 1986 Metallothionein (MT), a low molecular weight, metal-binding protein, has recently been shown to protect murine mononuclear phagocytic cells from the cytotoxic effects of bacterial lipopolysaccharides (LPS), the endotoxic component of Enterobacteriaceae. MT appears to function intracellularly as an antioxidant since autolysis results from lipid peroxidation initiated by free radicals of O2. Since this activity is distinct from MT's capacity to specifically sequestrate heavy metals, we examined whether MT synthesis can be induced by direct membrane activation or through interaction with soluble leukocyte mediators. Normal human monocytes, polymorphonuclear neutrophils (PMN), and lymphocytes, isolated from heparinized whole blood, were incubated with and without LPS from Escherichia coli and Salmonella typhosa. MT in cell lysates was quantitated using a Z°3Hg-binding assay employing Sephadex G-10 "minicolumns." When incubated with monocytes, PMN, or lymphocytes, neither preparation of LPS (10-100 txg/ml) was capable of enhancing Z°3Hg-binding activity after 24 or 72 hr incubation. CdCI/(2 txg/ml), however, increased binding activity in monocyte and lymphocyte cultures 4- and 15-fold, respectively. When monocytes and lymphocytes were cocultured with LPS, 2°3Hg-binding activity was not enhanced. Addition of human interleukin 1 (endogenous pyrogen) to these cultures had no significant effect. Leukocyte endogenous mediator (LEM), a product of LPS-activated PMN that possesses hypozincemic activity in vivo, did not induce MT synthesis. Collectively, these results demonstrate that leukocyte MT does not arise from direct LPS activation or from interaction with products secreted by LPS-activated cells. De novo synthesis of MT observed during endotoxemia and gram negative sepsis appears, therefore, to be induced by endogenously released corticosteroids. © 1987Academic Press, Inc.

INTRODUCTION Metallothionein (MT), a cysteine-rich, metal-binding, cytoplasmic protein, functions in the detoxification of H g 2 + , C d 2 + , N i 2 + , and Z n 2 + by sequestering up to seven divalent cations per molecule (Cherian and Goyer, 1978; Kojema and Kagi, 1978; Webb and Cain, 1982). To date, only those metals that bind MT stimulate its synthesis when administered to experimental animals or when added to cell culture (Hager and Palmiter, 1981; Karin et al., 1981b; Klaassen, 1981). Corticosteroids are the only other known inducers, but do not specifically bind MT (Karin et al., 1981a). Recent experimental evidence suggests that MT posseses additional protective functions that are unrelated to its metal-binding activity. Hypozincemia and de n o v o MT synthesis are characteristic features of gram negative sepsis and endotoxemia, the presence of lipopolysaccharides (LPS), or "endotoxin" derived

377 0013-9351/87 $3.00 Copyright© 1987by AcademicPress, Inc. All rightsof reproductionin any formreserved.

378

PEAVY AND FAIRCHILD

from gram negative cell walls (Pekarek and Evans, 1975; Sobocinski et al., 1978). In vitro, LPS is selectively toxic for monocytes and activated macrophages (Peavy et al., 1978, 1979). Autolysis results from production of H202 and free radicals of 02 generated in response to direct stimulation by LPS (Nathan, 1982). Interestingly, pretreatment of macrophages with Hg 2+, Cd 2+, Zn2+, and corticosteroids, but not Ca2+ or Pb 2÷, inhibits LPS toxicity, and MT synthesis has been demonstrated within protected cells (Patierno et al., 1983a). We have hypothesized that MT may function as a water-soluble antioxidant, protecting membranes and organelles from free radical attack. If true, MT might be induced directly by LPS or indirectly by soluble leukocyte mediators known to stimulate free radicals. These possibilities have been examined in vitro using leukocyte populations isolated from human peripheral blood. The results indicate that neither LPS nor soluble mediators, interleukin 1 or leukocyte endogenous mediator, induce MT in vitro. It appears, therefore, that endogenously released corticosteroids are responsible for initiating MT synthesis that is observed during endotoxemia and gram negative sepsis. MATERIALS AND METHODS Endotoxin. Lipopolysaccharides (LPS or endotoxin), extracted from Escherieha coli 0111 : B4 and 0127 : B8 and Salmonella typhosa using the phenol/water technique, were purchased from Difco Laboratories, Detroit, Michigan. LPS was reconstituted in sterile distilled/deionized H20 immediately before use. Cadmium. Ultrapure cadmium chloride (Alfa Products, Thiokol/Ventron, Danvers, ME) was dissolved immediately before use in distilled/deionized H20 and sterilized by 0.45-p~m filtration. Interleukin 1. Human interleukin 1 (IL-1) was generously provided by Dr. L. B. Lachmann, Department of Cell Biology, M.D. Anderson Hospital and Research Center, Houston, Texas. Cells from patients with monocytic leukemia were obtained by leukapheresis and were incubated overnight at 37°C with formalin-fixed Staphylococcus aureus. Culture supernatants were pooled, clarified by centrifugation, and quantitated for IL-1 activity using PHA-stimulated thymocytes as described previously (Lachmann and Metzger, 1980). Isolation and culture of leukocyte populations. Twenty-five to 50 ml heparinized (10 U/ml) blood, obtained by venipuncture from normal adults, was mixed with an equal volume of 6% dextran (Pharmacia, Uppsula) and incubated at 37°C without disturbance for 1 hr. Leukocytes were transferred to a 50-ml plastic centrifuge tube containing 10 ml RPMI 1640 (GIBCO, Grand Island, NY). Mononuclear and polymorphonuclear (PMN) leukocytes were isolated by fractionation on Ficoll gradients as described previously (Martin et al., 1979). Recovered cells were washed twice in RPMI 1640 and resuspended in Zn2+-deficient, Ham's F-12 medium (formulated by M.A. Bioproducts, Walkersville, MD) supplemented with 5% (v/v) heat-inactivated, autologous serum, 30 mM L-cysteine (GIBCO), 50 ~g/ml gentamycin sulfate (Schering, Kenilworth, NJ), and bicarbonate buffer (M.A. Bioproducts) at 2 × 10 6 cells/ml. When necessary, erythrocytes in PMN

INDUCTION OF LEUKOCYTE METALLOTHIONEIN

379

fractions were lysed by resuspending the cell pellet in 10 ml prewarmed TrisNH4C1, pH 7.3, for 15 min at 37°C. Mononuclear leukocytes were separated into adherent (monocyte) and nonadherent (lymphocyte) populations by adherence to plastic surfaces (Kohl et al., 1977). Two and one-half milliliters of a 10 × 106 cell/ml suspension was added to 60-mm culture dishes (Falcon, Oxnard, CA) and incubated at 37°C on a stationary platform in a 5% C Q in air atmosphere for 2 hr. Nonadherent cells were collected by gently rocking each dish and decanting the supernatant with a sterile pipette. Dishes were rinsed three times with 5 ml complete medium, and adherent cells were removed by scraping the bottom surface with a rubber policeman. All cells were collected by centrifugation at 4°C, washed, counted, and resuspended in complete medium at 2 × 106 cells/ml. Unfractionated leukocytes, monocytes, lymphocytes, and PMN were cultured using identical conditions. One milliliter of each cell suspension was added to a 12 × 75-mm plastic tube; LPS, Cd 2+, or IL-I was added immediately thereafter in 0.1 ml. Cultures were incubated upright and loosely capped at 37°C for 24-72 hr in a humidified chamber containing a 5% CO2 in air atmosphere. Quantitation o f metallothionein. Metallothionein was measured using a 2°3Hgbinding assay recently developed in this laboratory (Patierno et al., 1983b). At culture termination, cells were resuspended by vortexing and pelleted by centrifugation at 400g for 10 min. Cells were washed once in 2 ml Puck's saline A, centrifuged, and resuspended in 1 ml Ca 2÷- and Mg2+-free phosphate-buffered saline (PBS). After removing a 0.1-ml aliquot for counting (Coulter Electronics, Hialeah, FL), the remaining cells were lysed by pipetting the suspension into provials bathed in liquid N2. After thawing at room temperature, 100 ~1 lysate was mixed with 50 pJ 2°3HGC12 (New England Nuclear, Boston, MA, specific activity 1.81 mCi/mg, diluted to approximately 2.5 × 106 dpm/50 pJ) in 1-ml microcentrifuge tubes for 15 min. Thirty microliters of 12% (w/v) trichloroacetic acid was then added and the incubation continued for 15 rain at 4°C. Tubes were centrifuged at high speed in a Dade immunofuge at 4°C for 15 min. One hundred microliters of the clarified supernatant was added to Sephadex G-10 "minicolumns" inserted into 12 × 75-ram culture tubes. Following centrifugation at 400g for 5 sec, 2°3Hg-bound metallothionein was eluted by addition of 100 txl PBS and a 5-sec centrifugation. Tubes were capped with parafilm and counted in a Beckman 4000 gamma counter. Results are expressed as picomoles 2°3Hg bound per 105 cells. A twofold increase in 2°3Hg-binding activity has been shown in previous studies to be statistically significant (Patierno et al., 1983b). Cell viability. Viability was determined by trypan blue exclusion as described previously (Powers et al., 1982). RESULTS

The capacity of bacterial lipopolysaccharide to induce MT synthesis as a consequence of cell activation was examined by incubating LPS with purified leukocyte populations. Results illustrated in Fig. 1 demonstrate that 10 or 100 txg LPS from E. coli 0127 : B8 or S. typhosa did not increase 2°3Hg-binding activity when

380

PEAVY AND FAIRCHILD

24 hrs.

72 hrs.

3.2

%

2.4

~ 1.6 -r

-°

[

.8

dH20 lO#g lO0~g long lO0#g 2~g E. coli 0127:B8 S. typhosa CdCl 2 LPS

LPS

ooo

ooo

dH20 lO/~g lO0~g lO#g lO0#g 2~g E. coli 0127:B8 S. typhosa CdCl 2 LPS

LPS

FIG. 1. Failure of h u m a n P M N to synthesize metallothionein in vitro. The designated a m o u n t s of LPS, extracted from Escherichia coli 0127 : B8 and Salmonella typhosa, and CdC1 z were incubated at 37°C with 2 x 106 P M N for 24 (left panel) and 48 (right panel) hr. Results are e x p r e s s e d as picomoles 2°3Hg b o u n d per 105 cells.

incubated with PMN for 24 (left panel) or 48 (right panel) hr. Addition of 2 Ixg CdC12also failed to enhance 2°3Hg-binding activity. These results were not due to poor PMN survival or to Cd or LPS toxicity, since viabilities of control (distilled HzO) and LPS- and CdZ+-treated cultures were 91, 93, and 89%, respectively (data not shown). Marked decreases in PMN viability were observed, however, when the incubation period was extended to 72 hr. To determine if LPS induces MT in monocytes and lymphocytes, mononuclear leukocytes isolated by Ficoll gradient centrifugation were separated into adherent (monocyte-rich) and non-adherent (lymphocyte-rich) cell populations as stated in Materials and Methods. Immediately following recovery, each population was counted, resuspended in complete medium, and incubated for 72 hr. 2°3Hgbinding activity in untreated monocytes and lymphocytes was 1.08 and 0.175 pmole/105 cells, respectively (Fig. 2). Addition of 10 or 100 txg E. coli 0127 : B8 did not alter 2°3Hg-binding activity in either cell population. In contrast, inclusion of 2 Ixg CdCI2into culture medium increased binding in untreated monocytes to 4.67 pmole/105 cells and in lymphocyts to 0.5 pmole/105 cells. Recent experimental observations have suggested that interleukin 1 (IL-I), a soluble product secreted by activated monocytes and macrophages, participates in MT synthesis observed following the in vivo administration of LPS (Durnam et al., 1984). Since LPS stimulates IL-1 production in vitro, this possibility was examined by incubating mononuclear leukocytes with varying amounts of LPS. Experimental results, illustrated in Fig. 3, demonstrate that E. coli 0127 : B8 LPS enhanced 2°3Hg-binding activity when quantitated 24 and 48 hr after culture initiation. These increases, however, were less than twofold greater than those of controls (not significant by Student's t-test) and occurred when 4- and 10-fold increases were measured in Cd-treated cells. Mononuclear leukocytes were also

INDUCTION OF LEUKOCYTE METALLOTHIONEIN 8

Adherent

381

Non-Adherent

~6

%

o~ 4 o

8

-6

2

dH20 10~g 100pg 2~g E. coil 0127:B8 CdCI LPS

dH20 lOng lO0~g 0.2~g 2~g E. coli 0127:B8 CdCl2 CdCl2 LPS

FIG. 2. Induction of metallothionein in human lymphocytes and mononuclear phagocytic cells. Peripheral blood mononuclear leukocytes were fractionated into adherent (monocyte-rich) and nonadherent (lymphocyte-rich) cell populations as described in Materials and Methods. Two × 106 recovered cells were incubated for 72 hr with LPS or CdCI/. Results are expressed as picomoles per 105 cells. Data were obtained from one of five experiments with similar results.

incubated with a culture supernatant containing 0.3, 1, or 3% (v/v) partially purified IL-1. Data shown in Fig. 4 indicate that no measurable increase was found at any concentration tested. An additional product of LPS activation, termed leukocyte endogenous mediator (LEM), a protein derived from PMN, has been demonstrated to produce hypozincemia when injected into a variety of animal species (Pekarek and Evans, 1976). To determine if LEM can induce de n o v o synthesis of MT, dextran-sedimented leukocytes were incubated in the presence and absence of LPS from E. coli 0127 : B8 and 0111 : B4. After 24 hr, Hg-binding activity in control cultures treated with distilled H 2 0 w a s 0.39 pmole/105 cells (Fig. 5, left panel). Addition of 10 or 100 ~g LPS from either strain failed to enhance 2°3Hg-binding activity. In contrast, 2°3Hg-binding activity increased to 0.87 pmole/105 cells when these leukocytes were incubated with 2 ~g CdC12. Similar results were also obtained after 72 hr (Fig. 5, right panel). DISCUSSION

Results of this investigation have demonstrated that the lipopolysaccharides (LPS or endotoxin) of E n t e r o b a c t e r i a c e a e do not induce de n o v o synthesis of metallothionein in human monocytes, lymphocytes, or PMN by direct membrane activation or via secretion of soluble leukocyte mediators. Acute LPS lethality for experimental animals and the toxicity of LPS for mononuclear phagocytic cells can be effectively inhibited by prior treatment with specific transition metal salts ( Z n 2+ , N i z + , C d 2+ , and Mn 2+ but not Ca 2+ o r Pb 2+) or with corticosteroids pos-

382

PEAVY AND FAIRCHILD 24 hrs.

1.6

%

72 h r s .

1.2

"6

E=

dH20 10#g 100~g .2~g 2~g E. coli 0127:B8 CdCI2CdCI 2 LPS

dH2O 10#g 100#g .2/Jg 2/~g E. coli 0127:B8 CdCI 2 CdCI2 LPS

FIG. 3. Metallothionein synthesis in mononuclear leukocytes occurs in response to Cd 2+, but not LPS. Two x 106 mononuclear leukocytes, isolated by Ficoll gradient centrifugation, were incubated with LPS or CdC12. Representation of data is identical to that described for Fig. 1.

sessing anti-inflammatory activity (Patierno et al., 1983a; Snyder and Walker, 1976; Snyder et al., 1977). In vitro, protection of monocytes and macrophages is dependent on de novo synthesis of MT (Patierno et al., 1983a). More recently, ethane exhalation and the hepatic accumulation of malondialdehyde have been observed in mice poisoned with LPS (Peavy and Fairchild, 1986b). Collectively, these observations indicate that free radical-initiated lipid peroxidation underlies the pathogenesis of LPS, and suggest that MT may function in phagocytic cells as 10'

24 hrs.

72 hrs.

m

=6

4

E=

dH20 .3%

1% IL 1

5%

2#g

CdCI2

dH20 .3% 1% IL 1

5%

2~g CdCl2

FIG. 4. Failure of interleukin 1 to induce metallothionein. Two x 106 mononuclear leukocytes were incubated for 24 (left panel) or 72 (right panel) hr in the presence of the indicated final concentrations of human IL-1. Results are expressed as picomoles 2°3Hg bound per 105 cells.

INDUCTION OF LEUKOCYTE METALLOTHIONEIN 1.0"

24 h r s .

383

72 h r s .

.8 % .6 ¸ o

"5

E

dH20 10~g 100~g 10~g 100/~g 2~g E. coli 0111:B8 ~ CdCI2 LPS E. coli 0127:B8 LPS

dH20 long loo~g long loo#g 2~g E.coli0111:B4 ~ CdCI2 LPS E. coli 0127:B8 LPS

FIG. 5. Induction of metallothionein in cultures of dextran-sedimented leukocytes. Two × 106 nucleated leukocytes, obtained by dextran sedimentation, were incubated for 24 (left panel) or 72 (right panel) hr. Experimental details and representation of results are as described for Fig. 1.

a water-soluble, antioxidant. It was anticipated, therefore, that MT would be induced in response to LPS activation, a process known to generate free radicals of O2. Endogenous corticosteroids, released in response to LPS, apparently are of primary importance in initiating MT synthesis, at least in these cell types. Metallothionein is known to function chiefly in detoxification of transition heavy metals by sequestrating these ions immediately after internalization into the cells. Results reported here and elsewhere have demonstrated that inducibility of MT in human and murine mononuclear phagocytic cells and its capacity to inhibit free radical-initiated lipid peroxidation reactions (Patierno et al., 1983a). It appears, therefore, that in monocytes and macrophages, MT acts to protect against free radicals of O2 generated as products of the phagocyte respiratory burst. Interestingly, Cagen and Klassen have demonstrated that Zn 2+ administration protects rats against CC14 hepatotoxicity and have detected complexes of MT associated with radiolabeled CC14in hepatic homogenates. These investigators have suggested that complex formation arises from interaction with metabolites of CC14(Cagen and Klassen, 1979). The failure to observe MT synthesis following incubation of monocytes and macrophages with LPS or with soluble mediators may have resulted from an inability of specific membrane receptors to provide appropriate intracellular signals. Cytoplasmic receptors specific for heavy metals and corticosteroids have been demonstrated in a variety of cell types (Hager and Palmiter, 1981; Karin et al., 1981b). It is currently unknown whether or not additional membrane receptors for IgG and activated complement components, particularly germane to the phagocytic process, participate in inducing MT. Investigations are currently under way in this laboratory to determine these possibilities. Nevertheless, corti-

384

PEAVY AND FAIRCHILD

costeroids must still play an important role, since functional receptors are known

to exist in human and murine monocytes, macrophages, and macrophage cell lines (Werb et al., 19878a, b). ACKNOWLEDGMENTS The authors thank Lisa V. de Sousa and Jyoti J. Patel for excellent technical assistance, and Debra Hines for skilled preparation of the manuscript. This investigation was supported by grants from the American Cancer Society, the American Society for Medical Technology Education and Research Fund, Inc., and the Baker Instruments Company.

REFERENCES Cagen, S. Z., and Klaassen, C. D. (1979). Protection of carbon-tetrachloride-induced hepatotoxicity by zinc: Role of metallothionein. Toxicol. Appl. Pharmacol. 51, 107-116. Cherian, M. G., and Goyer, R. A. (1978). Metallothioneins and their rote in the metabolism and toxicity of metals. Life Sci. 23, 1-10. Durnam, D. M., Hoffman, J. S., Quaife, C. J., Benditt, E. E, Chert, H. Y., Brinster, R. L., and Palmiter, R. D. (1984). Induction of mouse metallothionein-I mRNA by bacterial endotoxin is independent of metals and glucocorticoid hormones. Proc. Natl. Acad. Sci. U.S.A. 81, 1053-1056. Hager, L. J., and Palmiter, R. D. (1981). Transcriptional regulation of mouse liver metallothionein-I gene by glucocorticoids. Nature (London) 291, 340-342. Karin, M., Anderson, R. D., and Herschman, H. R. (1981a). Induction of metallothionein mRNA in HeLa ceils by dexamethasone and by heavy metals. Eur. J. Biochem. 118, 527-531. Karin, M., Slater, E. E, and Herschman, H. R. (1981b). Regulation of metallothionein synthesis in HeLa cells by heavy metals and glucocorticoids. J. Cell. Physiol. 106, 63-74. Kiaassen, C. D. (1981). Induction of metallothionein by adrenocortical steroids. Toxicology 20, 275-279. Kohl, S., Start, S. E., Oleski, J. M., Shore, S. L., Ashman, R. B., and Namias, A. J. (1977). Human monocyte macrophage-mediated antibody-dependent cytotoxicity to herpes simplex-virus-infected cells. 118, 729-735. Kojema, Y., and Kagi, J. H. R. (1978). Metallothionein. Trends in Biochem. Sci. 3, 90-92. Lachmann, L. B., and Metzger, R. S. (1980). Characterization of high and low molecular weight lymphocyte activating factor (interleukin 1) from P388D 1 and J774.1 mouse macrophage cell lines. J. Reticulo. Soc. 27, 621-629. Martin, D. K., Nelms, C. D., Mackler, B. E, and Peary, D. L. (1979). Lymphoproliferative responses induced by streptococcal antigens in recurrent aphthous stomatitis and Behcet's syndrome. Clin. Immunol. Immunopathol. 13, 146-155. Nathan, C. E (1982). Secretion of oxygen intermediates: Role in effector functions of activated macrophages. Fed. Proc. 41, 2206-2210. Patierno, S. R., Costa, M., Lewis, V. M., and Peary, D. L. (1983a). Inhibition of LPS toxicity by metallothionein-inducing agents. J. Immunol. 130, 1924-1929. Patierno, S. R., Pellis, N. R., Evans, R. M., and Costa, M. (1983b). Application of a modified Z°3Hg binding assay for metallothionein. Life Sci. 32, 1629-1636. Peary, D. L., Baughn, R. E., and Musher, D. M. (1978). Strain-dependent cytotoxic effects of endotoxin for mouse peritoneal macrophages. Infect. lmmun. 21, 310-319. Peary, D. L., Baughn, R. E., and Musher, D. M. (1979). Effects of BCG-infection on the susceptibility of mouse macrophages to endotoxin. Infect. lmmun. 24, 59-64. Peary, D. L., and Fairchild, E. J., II (1986). Evidence for lipid peroxidation in endotoxin-poisoned mice. Infect. Immun. 52, 613-616. Peary, D. L., and Fairchild, E. J., II (1987). Toxicologic interactions between ozone and bacterial endotoxin. Environ. Res. 42, 63-71. Pekarek, R. S., and Evans, G. W. (1975). Effect of acute infection and endotoxemia on zinc absorption in the rat. Proc. Soc. Exp. Biol. Med. 150, 755-757.

INDUCTION OF LEUKOCYTE METALLOTHIONEIN

385

Pekarek, R. S., and Evans, G. W. (1976). Effect of leukocyte endogenous mediator (LEM) on zinc absorption in the rat. Proc. Soc. Exp. Biol. Med. 152, 573-575. Powers, C. N., Peavy, D. L., and Knight, V. (1982). Selective inhibition of functional lymphocyte subpopulations by rivavirin. Antimicrob. Agents and Chemo. 22, 108-114. Snyder, S. L., and Walker, R. I. (1976). Inhibition of lethality in endotoxin-challenged mice with zinc chloride. Infect. Immun. 13,998-1000. Snyder, S. L., Walker, R. I., and Moniot, J. V. (1977). Protection against endotoxin-induced mortality in mice treated with transition metal salts. Infect. Immun. 15, 337-339. Sobocinski, P. Z., Canterbury, W. J., Mapes, C. A.,and Dinterman, R. E. (1978). Involvement of hepatic metallothioneins in hypozincemia associated with bacterial infection. Amer. J. Physiol. 234, E399-406. Webb, M., and Cain, K. (1982). Functions of metallothionein. Biochem. Pharmacol. 31, 137-142. Werb, Z., Foley, R., and Munck, A. (1978a). Interaction of glucocorticoids with macrophages. Identification of glucocorticoid receptors in monocytes and macrophages. J. Exp. Med. 147, 1684-1694. Werb, Z., Foley, R., and Munck, A. (1978b). Glucocorticoid receptors and glucocorticoid-sensitive secretion of neutral proteases in a macrophage line. J. Immunol. 121, 115-121.