CELLULAR

IMMUNOLOGY

145372-379

(1992)

SHORT COMMUNICATION Human Recombinant Migration Inhibitory Factor Activates Human Macrophages to Kill Tumor Cells’ Lu-ANN

M. POZZI AND WEISHUI Y. WEISER’

Department of Medicine, Harvard Medical School; and Department gf Rheumatology and Immunology, Brigham and Women k Hospitul, 221 Longwood Avenue, Boston, Mussuchusetts 02115 Received May 11, 1992; accepted August 10, 1992 A recombinant form of human migration inhibitory factor (rMIF) obtained from COS-1 cells transfected with MIF-specific cDNA is able to activate cultured human peripheral blood monocytes and monocyte-derived macrophages, in a dose-dependent manner to become cytotoxic for tumor cells in vitro. The cytotoxicity exhibited by macrophages treated with rMIF is >30% above that of cells incubated with control supernatants or with media and peaks 72 hr after the addition of tumor targets, rMIF also induces macrophages to produce tumor necrosis factor (TNF-a) and interleukin-I@ (IL-IO). These results demonstrate that rMIF is able to modulate macrophage functions and plays a role in cell-mediated immune response. Q 1992 Academic PXSS, hc.

INTRODUCTION Macrophages, the final effector cells in many cellular immune responses,can be activated by various cytokines and LPS to destroy a variety of intracellular pathogens and extracellular neoplastic cells (l-3). It has been demonstrated that upon activation by interferon (EN-~) or granulocyte/macrophage colony-stimulating factor (GM-CSF), human monocyte-derived macrophages are able to kill the intracellular parasite, Leishmania donovani (4, 5) and extracellular tumor targets (6, 7). MIF, the first lymphokine discovered, was originally shown to be produced by antigen-stimulated lymphocytes and has the property of inhibiting the migration of macrophages in vitro (8, 9). It was subsequently found that culture supernatants containing MIF was able to affect a variety of macrophage physiology and functions, including increasedadherence,ruffled border activity, direct monophosphate oxidation, and antimicrobial and tumoricidal activities ( lo- 12). The cloning of a cDNA encoding a human MIF and the availability of MIF in a recombinant form enable us to delineate whether these altered functions of macrophages were induced by MIF or by other factor(s) present in the supernatants and to learn more about the biology of this rMIF (13). In this study, we examined the ability of rMIF to activate human macrophages to inhibit the growth and kill tumor targets. ’ This work was supported by National Institutes of Health Grant AI 2280 I. * To whom correspondence should be addressed. 372 0008-8749/92 $5.00 Copyright 8 1992 by Academic Press, Inc. All rights of reproductmn m any form reserved.

SHORT

COMMUNICATION

373

MATERIALS AND METHODS Monocyte and monocyte-derived macrophages. Human peripheral blood was obtained from healthy volunteers. Mononuclear cells isolated by Ficoll-Hypaque density centrifugation were suspendedat 2 X 1O6cells/ml in medium 199 supplemented with 15% heat-inactivated FCS and 2 mM L-glutamine as previously described (5). The cells were seededat 2 X lo5 cells per well in 96-well plate (Costar) and incubated at 37°C in a 5% CO2 incubator. The adherent cells obtained after incubation for 1 hr were termed fresh monocytes, whereas the monocyte-derived macrophages were adherent cells which were allowed to adhere for 10 days. It is of note that after 10 days, the adherent monolayers consisted of 95-97% macrophages, as determined by morphology and esterasestaining (5). The monocyte and/or macrophage cultures were treated with rMIF, Stop-MIF, or were kept incubated in medium for the desired number of hours. Labeling of tumor targets and cytotoxicity assay. Cells from A375 line, a human melanoma cell line, were used in all experiments. Cells from K562, a myeloma line, were used in some of the experiments. After incubation, target cells were cultured in Dulbecco’s MEM supplemented with 10% heat-inactivated FCS and 2 mA4 L-glutamine. Before counting, the adherent A375 cells were trypsinized. The number of both A375 and K562 cells were determined. Twice as many cells as needed were labeled with 20 &i/ml of [3H]thymidine (NEN) in medium for 24 hr. The cells were again counted and resuspendedin medium at 1 X 1O5cells/ml. Target cells were distributed at 100 &well (effector-to-target ratio = 20:1) in cultures of fresh monocytes and/or macrophageswhich had been treated with rMIF, Stop-MIF, or incubated in medium. After incubation, plates were centrifuged, supernatants were collected, and the counts were obtained from the scintillation counter (Packard Instrument Co). Cytotoxicity was determined by using the following formula: % cytotoxicity = cpm of samples cpm of media/cpm of total lysates - cpm of spontaneous release. Cytokines and antibodies. rMIF, isolated by COS-1 cell expression screening of cDNA from a human T-cell hybridoma line (T-CEMB) for MIF activity, was generated with the assistanceof Genetics Institute (Cambridge, MA) (14). Stop-MIF, a mutant rMIF which has no MIF activity, was produced by inserting a 14-baseoligodeoxynucleotide containing a termination codon at the PstI site in the coding region (13). Recombinant human IFN-7 was obtained from Amgen Biologicals (Thousand Oaks, CA). Neutralizing murine anti-human TNF-(r monoclonal antibody was purchased from Boehringer-Mannheim (Indianapolis, IN). Neutralizing goat anti-human IL- 1p antibody was obtained from R&D System (Minneapolis, MN). TNF bioassay. TNF determinations were performed using TNF-sensitive L929 cells. To generate supernatants for TNF and IL-l assays,human peripheral blood mononuclear cells were seededinto 24-well plate at 2 X lo6 cells/well (Costar) in medium 199 supplemented with 15%FCS and 2 mM L-glutamine. After 10 days of incubation, the cells were stimulated with various concentrations of rMIF. Culture supernatants were harvested at the desired interval and stored frozen at -70°C. For TNF assay, L929 cells suspended in RPM1 supplemented with 10% FCS were seededat 2 X lo4 cells/well in a 96-well plate. After overnight incubation, the supernatants to be tested were added in the presence of actinomycin D and further cultured overnight. The cells were washedand the plates were allowed to dry. Formamazin which was dissolved in acidified isopropyl alcohol and was added to each well. The OD was read at a wavelength of 560 nm. The TNF units were derived by the following formular: % kill

374

SHORT

COMMUNICATION TABLE

rMlF Induced Cytotoxicity

1

Against A375 and K562 Cells by Monocytes and Macrophages Percent cytotoxicity”

Experiment

1

Experiment 2

Experiment 3

Treatment

A375

K562

A315

K562

A375

K562

Monocytes MIF I:5 MIF I:10 MIF I:20 IF-N + LPSb stop I:5 stop 1:lO

37.7 35.1 27.8 48.6 -1.5 -3.1

57.0 55.6 30.1 32.0 4.1 -5.8

48.6 37.8 29.7 40.3 -6.2 -4.8

54.5 45.3 31.4 38.4 2.4 4.4

34.4 28.6 25.1 37.1 -2.2 6.1

50.1 36.3 32.4 43.8 1.6 -3.4

Macrophages MIF 1:5 MIF 1:lO MIF I:20 IFN + LPSb stop 1:5 stop 1:lO

42.3 30.7 27.2 45.1 1.5 2.2

36.2 29.8 33.6 25.6 0.2 2.4

40.8 36.7 26.3 53.8 1.7 -3.1

55.4 41.0 29.7 42.1 -4.9 -4.2

56.7 38.2 31.3 50.7 6.0 -5.5

54.4 34.1 29.3 49.8 1.3 -4.7

’ Each stimulation condition was carried out in triplicate with values representing the means, Cytotoxicity was determined 72 hr after stimulation b 1000 U/ml of IFN-7 and 10 rig/ml of LPS were used.

= [ I-(average OD of sample/average OD of control)] X 100. TNF units = % kill X dilution factor/50. One unit of TNF caused 50% lysis of L929 monolayers. IL-1p determination. Supernatants were generated and collected as described for the TNF assay. The presence of IL- l/3 was examined by using the IL- I P-specific ELISA Quantikine kit (R&D System). The amount of IL-lp produced was calculated by linear regression from the standard curve. H,O, microassay. Hydrogen peroxide production from cultured macrophages was performed in microtiter wells by measuring the release of H202 from PMA-challenged cells with a scopoletin-based fluorometeric assay as described by De la Harpe and Nathan (15). Human monocyte-derived macrophages were incubated with rMIF, rIFNy, Stop-MIF, or medium for 24 or 48 hr. The cells were washed with PBS and further incubated for 60 min at 37°C with Krebs-Ringer-buffered salt solution containing 20 PALM scopoletin (Sigma), 70 rig/ml PMA, and 1.9 purpurogallin U/ml horseradish peroxidase. To control for variation in cell numbers, the data were normalized to micrograms of DNA per culture. DNA was measured using a calorimetric assay (16). Data are expressed in nmol H,O,/pg DNA/hour. RESULTS

AND DISCUSSION

We have previously shown that human rMIF activates human monocyte-derived macrophages to inhibit the growth of and/or kill promastigotes and amastigotes of an intracellular parasite, L. donovani (17). To learn more about the interaction of this

SHORT COMMUNICATION

a .Z u

g 0 :: 5

50 40

:: i;;

5

30

0

E :

20

Pk

10

E : ; P

60 50 40 30 20 10 0

0 24

46 Hours after

72

96

24

48

stimulation

96

72

Hoursafter stimulation

FIG. 1. Development of cytotoxicity by rMIF-treated monocytes and macrophages.Monocytes and macrophagescultures obtained from human peripheral blood were incubated with I:5 dilution of rMIF or StopMIF (a mutant of rMIF which contained a termination codon in the &I site of its coding region) for 24 hr. ‘H-labeled tumor targets, A375 (left) or K562 (right) were added to rMIF-treated monocfles (open squares),rMIF-treated macrophages(solid squares),Stop-MIF-treated monocytes (open circles), and StopMIF-treated macrophages (shaded circles) at an effector-to-target ratio of 20: 1. The effector cells and the ‘H-labeled tumor targets were incubated for additional 24, 48, 72, and 96 hr. Cytotoxicity from each time point was determined. The data are expressedas means of three experiments f SEM.

rMIF with macrophages,we further examined its effect on macrophages for killing of extracellular targets. Human peripheral blood monocytes and monocyte-derived macrophageswere treated with rMIF or Stop-MIF, a mutant MIF containing a termination codon in its coding region. After treatment, 3H-labeled A375 cells or K562 cells were added to the cultures. Cytotoxic activity was found in cultures of fresh monocytes and monocyte-derived macrophages treated with rMIF but not with Stop-MIF (Table 1). The cytotoxic activity induced by rMIF at 15 dilution was comparable to that of rIFN-y at 1000 U/ml. Optimal cytotoxicity toward the tumor cells was observed < 72 hr after the addition of tumor targets to rMIF-treated macrophages (Fig. 1).

60 -

. rz. 50-. ; 2 p

50 -

Macrophages

40-

* .= e ;

30-

0

0"

hlonocytes

4030-

5

zo-

E : ; 0

lo-

20 E g ; 0.

loo-lol,.#.,.,.,.,r -12

0

_

.lO,,.,.,,,.,.,# 6

12

24

Pretreatment time

46

72

-12

6

12

24

46

72

Pretreatmenttime

FIG. 2. Time courSeof rMIF-induced cytotoxicity. Monocyte/macrophages cultures were incubated with 1:5 dilution of rMIF (solid squares)or I:5 dilution of Stop-MIF (open squares)for 6, 12, 24, 48, and 72 hr followed by the addition of ‘H-labeled A375 cells. rMIF and Stop-MIF were also administered to monocyte/ macrophagescultures I2 hr after the addition of A375 cells. Cytotoxicity to A375 cells by monocytes (right) and macrophages(left) was determined ~72 hr after the addition of target cells. The data represent means + SEM of six quantifications of one experiment.

376

SHORT

COMMUNICATION

60

e"

50

10 ::

40

z 5

30

E

20

: G P

10 0 l/l00

1150

1120

1110

l/5

Dilution

FIG. 3. Dose response of rMIF-induced macrophage cytotoxicity. Macrophages were treated with dilutions of rMlF for 24 hr as indicated on the abscissa. A375 cells (open squares) or K562 cells (solid squares) were incubated with treated macrophages for ~72 hr and the percentage cytotoxicity was determined as described under Materials and Methods.

Next we investigated the time course of rMIF-induced cytotoxic activity by macrophages.We found that in order to exhibit cytotoxicity in the MIF system, monocytes and macrophages need to be pretreated with rMIF, and the rMIF-induced cytotoxic activity is maximal when macrophages have been pretreated with rMIF 24 hr before the addition of tumor targets (Fig. 2). No cytotoxic effect was found when the effector cells were treated after the addition of tumor targets. Also, rMIF-induced cytotoxicity against the tumor cells was dose dependent (Fig. 3). It is of note that the maximal antileishmanial effect of macrophages induced by rMIF treatment required longer pretreatment time with rMIF; usually 48 to 72 hr of treatment before infection with L. donovani is needed for the macrophages to expressthis function (17). We have previously found that macrophages treated with rMIF showed enhanced levels of TNF-a and IL- l/3 mRNA expression ( 13, 18). To gain some insights into the mechanism responsible for the induced tumor killing by rMIF, we examined whether

70 60 -

-10

'

I 1

I 3

I 6

12

I 24

Hours after simulation

I 48

I 72

-10'

I 1

3

I 6

I 12

I 24

I 48

I 72

Hours after stimulation

FIG. 4. Production of TNF-a and IL-10 by rMIF-treated macrophages. Macrophages were incubated with rMIF (solid squares) or Stop-MIF (open squares). Supernatants were harvested at intervals as indicated in the abscissa. The amount of TNF released (left) was obtained through bioassay whereas the presence of IL18 (right) was determined by ELISA as described under Materials and Methods.

377

SHORT COMMUNICATION

the elevation in TNF-a! and IL-lp messagesfollowed with increased production of respective proteins. Thus, supernatants from macrophages incubated with rMIF or Stop-MIF were examined for the presenceof TNF-a and IL- 1,CIproteins. L929 bioassay was usedto detect the presenceof soluble releasedform of TNF-cu,whereasthe presence of IL- 1p was determined by ELISA. As shown in Fig. 4,4 both TNF-(r and IL- l/3 were found in culture supernatants of macrophages treated with rMIF but not with StopMIF. The time course of the releaseof both cytokines by rMIF-treated macrophages paralleled those with the expression of cytotoxicity. TNF-a in tumor killing has been well documented ( 19-23) and IL- 1p has also been shown to cause cytostasis of various tumor cell lines including A375 cells (24, 25). However, in five independent experiments, administration of neutralizing anti-TNF01antibody (20 to 100 units) or neutralizing anti-IL- lp antibody ( 10 ng to 2.5 pg/ml) was not able to abolish the cytotoxicity induced by rMIF, although cytotoxicity induced by treatment of macrophageswith rIFN-y was abrogated by anti-TNF-cY antibody and was reduced by anti-IL-10 antibody. A representative experiment is shown in Table 2. It is known that cells of the monocyte/macrophage lineage can effect extracellular killing of nucleated mammalian cells via different mechanisms, including cytokines, complement components, neutral proteases,arginase, and reactive oxygen intermediates. The susceptibility of the target cells utilized and the choice of activating and

TABLE 2 Effect of Antibody to TNF-(U or IL- Ifl on Cytotoxicity Mediated by rMIF or rIFN-y Treatment’

% Cytotoxicity

MIF I:5 +aTNF 100 U +aTNF 80 U +aTNF 60 U +aTNF 40 U +aTNF 20 U +aILl 2.5 pg taIL1 1.0 pg faILI 0.1 pg +aILl 10 ng Stop-MIF rIFN-y +aTNF 100 U +aTNF 80 U +aTNF 60 U +aTNF 40 U +aTNF 20 U +aILl 2.5 pg taIL1 1.0 pg +aILl 0.1 pg +aILI 10 ng

41 ir6 29 k 5 33 * 4 28 + 5 21*2 39 f 7 31 +4 40 2 3 37 f 5 36 + 7 5fl 47 * 4 2_+1 6&l 9k2 13 f 2 26 f 3 19 f 2 28 f 4 37 +4 49 + 6

’ Macrophage cultures were treated with I:5 dilution of rMIF, 1000 U/ml of rIFN-y in the absence or presenceof various concentrations of anti-TNF-ol or anti-IL-ID antibody, or 1:5 dilution of Stop-MIF for 24 hr. After the incubation, ‘H-labeled A375 cells were added. Cytotoxicity was determined ~72 hr after the addition of target cells. Data represent the means f SEM of four quantifications of one experiment.

378

SHORT

COMMUNICATION

triggering agents are also of crucial determinants of the killing mechanisms at play in different in vitro systems. The production of hydrogen peroxide by macrophages is an important mechanism of cytotoxicity. We examined rMIF-treated macrophages for H202 production by measuring its releaseafter challenge with PMA. Macrophages were treated with rMIF ( 1:5 dilution), Stop-MIF, or rIFN-y ( 1000 U/ml) for 24 or 48 hr. In two preliminary experiments, no H202 was found when cells were treated with rMIF for 24 hr, whereas 2 and 3 nmol/pg DNA/hour was produced from rIFN-y-treated cells of the same donors. However, after 48 hr, both rMIF-treated and rIFN-y-treated macrophages were found to produce H202 (2 to 3 and 3 to 5 nmol/pg DNA/hour, respectively). It is of note that the amount of H202 produced by rIFN-y-treated macrophages is in agreement with the findings of Lehn et al. (26). H202 production was not found in cells treated with Stop-MIF or incubated in medium. Since we found no TNF-(u, ILl& nor H202 production from monocytes/macrophages treated with Stop-MIF, and Stop-MIF was not able to induce cytotoxicity from these cells, it is therefore possible that cytotoxicity induced by treatment of macrophageswith rMIF is through the ability of rMIF to induce TNF-(u, IL- 10, and H202 together with other factors which remain to be determined. The mechanisms of MIF-induced cytotoxicity for intracellular organisms and for extracellular targets are currently under investigation. MIF activity has been detected in patients with rheumatoid polyarthritis and in a variety of chronic inflammatory loci (27). The presenceof MIF at sitesof inflammation suggestsa role for MIF in regulating the function of macrophages. Recently, rMIF has been shown to induce strong T-cell proliferative response and increase antibody production to soluble protein antigens, bovine serum albumin, and HIV-gp120 (28). The data presented here together with previous findings demonstrate that rMIF is an activator of macrophages and is likely to be involved in cell-mediated immune host defenses. REFERENCES 1. Nacy, C. A., Oster, C. N., James, S. L., and Meltzer, M. S., Contemp. Top. Immunobiol. 13, 146, 1984. 2. Murray, H. W., Contemp. Top. Immunobiol. 13, 97, 1984. 3. Adams, D. 0.. and Hamilton, T. A., dnnu. Rev. Immunol. 2, 283, 1984. 4. Murray, H. W., Rubin, B. Y., and Rothermel, C. D., J. Clin. Invesl. 72, 1506, 1983. 5. Weiser, W. Y., Van Niel, A., Clark, S. C., David, J. R., and Remold, H. G.. J. Exp. Med. 166, 1436, 1987. 6. Kleinerman, E. S., Schroit, A. J., Foggier, W. E., and Fidler, I. J., J. Clin. Invest. 72, 304, 1983. 7. Grabstein, K. H., Urdal, D. L., Tushiuski, R. J., Mochizuki, D. Y., Price, V. L., Cantrell, M. A., Gillis, S., and Conlon, P., Science 232, 506, 1986. 8. Bloom, B. R., and Bennett, B., Science 153, 80, 1966. 9. David, J. R., Proc. Natl. Acad. Sci. USA 65, 12, 1966. 10. Nathan, C. F.. Karnovsky, M. L., and David, J. R., J. Exp. Med. 133, 1356, 1971. 11. Nathan, C. F., Remold, H. G.. and David, J. R., J. Exp. Med. 137, 275, 1973. 12. Churchill, W. H.. Piessens, W. F., Sulid, C. A., and David, J. R., J. Immunol. 115, 78 1, 1975. 13. Weiser, W. Y., Temple, P. A., Witek-Giannotti, J. S., Remold, H. G., Clark, S. C., and David, J. R., Proc. Nad. Acad. Sci. USA 86, 7522, 1989. 14. Weiser, W. Y., Remold, H. G., and David, J. R., Cell. Immunol. 90, 167, 1985. 15. De la Harpe, J., and Nathan, C. F., J. Immunol. Methods 78, 323, 1985. 16. Labarca, C., and Paigen, K., Anal. Biochem. 102, 344, 1980. 17. Weiser, W. Y.. Pozzi, L. M., and David, J. R., J. Immunol. 147, 2006, 1991. 18. Weiser, W. Y., Fourlong, S., and David, J. R., FASEB J. 5, A1487, 1991.

SHORT COMMUNICATION

379

19. Carswell, E. A., Old, L. J., Kassel, R. L., Green, S., Fiore, N., and Williamson, B., Proc. Natl. Acad. Sci. USA 12, 3666, 1975. 20. Old, L., Science 230, 630, 1985. 21. Mantovani, A., In “Human Monocytes” (M. Zembala and G. L. G. Asherson, Eds.), pp. 303-33 1. Academic Press,London, 1989. 22. Feinman, R., Henriksen-DeStefano, D., Tsujimoto, M., and Vilcek, J., J. Immunol. 138, 635, 1987. 23. Baglioni, C., In “Tumor Necrosis Factors” (B. Beutler, Ed.), pp. 425-438. Raven Press, New York, 1992. 24. Keller, R., and Keist, R., Cell. Immunol. 10, 659, 1987. 25. Lachman, L., Dinarello, C., Liansa, N., and Fidler, I. J., J. Immunol. 136, 3098, 1986. 26. Lehn, M., Weiser, W. Y., Engelhorn, S., Gillis, S., and Remold, H. G., J. Immunol. 143, 3020, 1989. 27. Odink, K., Cerletti, N., Bruggen, J., Clerc, R. G., Tarcsay, L., Zwadlo, G., Gerhards, G., Schlegel, R., and Sorg, C., Nature 330, 80, 1987. 28. Weiser, W. Y., Pozzi, L. M., Titus, R. G., and David, J. R., Proc. Nat/. Acad. Sci. USA. 1992, in press.