- Email: [email protected]
Inhibition of Ehrlich ascites tumor in vivo by PAF-antagonists
Int. J. lmmunopharmac., Vol. 12, No. 1, pp. 5 7 - 6 5 , 1990. Printed in Great Britain.
0192-0561/90 $3.00 + .00 © 1989 International Society for Immunopharmacology.
I N H I B I T I O N OF E H R L I C H ASCITES T U M O R I N VIVO BY PAF-ANTAGONISTS DENISE FECCHIO, MOMTCHILO RUSSO, PIERRE SIROIS,* PIERRE BRAQUET t a n d SONIA JANCAR* Departamento de Imunologia, Instituto de Ci~ncias Biom6dicas, Universidade de S~o Paulo, S~o Paulo, Brazil; *Department of Pharmacology, University of Sherbrooke, Sherbrooke, Quebec, Canada; tlnstitut Henri Beaufour Research Laboratories, France (Received 20 December 1988 and in final form 10 July 1989)
Abstract - - Several lines of evidence support that PAF modulates the inflammatory and immune responses,
and that tumors may inhibit both these processes. In the present study we analysed the effect of PAF antagonists on the growth of Ehrlich Ascites Tumor (EAT) in vivo. Mice were inoculated intraperitoneally with 1 × 10 3 EAT ceils and the tumor growth evaluated by counting the number of peritoneal cells, 1, 6 and 10 days after tumor implantation. BN 52021 was administered intraperitoneally, intravenously or subcutaneously once or twice a day, at 1.0, 2.5, 5.0 and 20.0 mg/kg. Control animals received 0.1 ml of the vehicle in the same schedule. It was found that i.p. and i.v. administration of BN 52021 (5 mg/kg, twice a day) significantly inhibited EAT growth (80.8070 and 56.0070 respectively). Other routes and doses were less effective. Another PAF antagonist, SRI 63441 (5 mg/kg, i.p., twice a day) also inhibited EAT growth (80.4°70). The BN 52021 added to EAT cells in culture, at concentration of 10 3 and 10 -4 M, did not affect the viability and proliferation of tumors cells. In an attempt to understand the mechanism of this inhibition, we analyzed the peritoneal macrophages for spreading ability and H202 release. It was found that 24 h after tumor implantation there was an increase in the spreading ability of peritoneal macrophages (75070) and that, as the tumor grew, the spreading index fell to control levels (<10070). In the groups treated with BN 52021 (5 mg/kg/twice a day) the spreading remained elevated (50-60070) at all the times examined. Release of H202, measured by horseradish peroxidase-phenol red oxidation, was below detectable levels throughout tumor growth. Treatment with BN 52021 as above, significantly stimulated the release of H202, 6 and 10 days after tumor implantation (0.4 to 0.6 nmol/105 cells). The data presented show that PAF antagonists inhibit EAT growth in vivo and that this inhibition is concomitant with activation of peritoneal macrophages. These results suggest that macrophages can critically influence tumor growth and that PAF or PAF-like substances may modulate macrophages-tumor interactions.
their p r o d u c t s o n the evolution o f t u m o r growth. A m o n g the p r o d u c t s released by i n f l a m m a t o r y cells, lipid m e d i a t o r s have been s h o w n to play a n essential role in p h a g o c y t e f u n c t i o n s a n d even to interact with i m m u n o c o m p e t e n t cells. P r o s t a g l a n d i n s , particularly PGE2, are k n o w n to exert a negative feedback o n m a c r o p h a g e s a n d to m a r k e d l y affect lymphocytes (for review see G o o d w i n & W e b b , 1984). Evidence is a c c u m u l a t i n g t h a t leukotriene B 4 a n d platelet activating factor ( P A F ) are also very p o w e r f u l i m m u n o r e g u l a t o r s (for review see Rola-Pleszczynski, 1985; B r a q u e t & Rola-Pleszczynski, 1987). E h r l i c h t u m o r can be t r a n s p l a n t e d into the peritoneal cavity o f mice where it grows in ascitic
O n e o f the factors w h i c h m o d u l a t e s t u m o r g r o w t h is the intensity o f the i n f l a m m a t o r y response evoked by the t u m o r cells. Several reports in the literature p o i n t to a n i m p o r t a n t role for i n f l a m m a t o r y cells in the c o n t r o l o f t u m o r g r o w t h (Evans, 1973; Eclles & A l e x a n d e r , 1974; Russel, Gillespie, H a n s e n & C o c h r a n e , 1976). Conversely, the presence o f the t u m o r results in a n i m p a i r m e n t o f m a c r o p h a g e a n d i n f l a m m a t o r y responses in the host a n d this m a y be m e d i a t e d by product(s) released by the t u m o r (Fauve, H e v i n , J a c o b , G a i z z a r d & J a c o b , 1974; Pike & S n y d e r m a n , 1976). T h u s , as i n f l a m m a t i o n usually initiates the m o r e specific i m m u n e response, we felt it relevant to study the role o f i n f l a m m a t o r y cells a n d
:l:Author to whom correspondence should be addressed at: Departamento de Imunologia, Instituto de Ci~ncias Biom6dicas, USP, Caixa Postal 4365, CEP 01051, S~o Paulo (SP), Brazil. 57
D. FECCHIOet al.
58
form (EAT). We have recently shown that the growth of EAT induces a very low influx of polymorphonuclear leukocytes but does not stimulate local macrophages, as assessed by spreading and release of hydrogen peroxide (H2Oz) (Fecchio, Sirois, Russo & Jancar, in press). We have also shown that significant amounts of PGE2 are released into the peritoneal cavity following EAT implantation. In the present study we analysed the effect of PAF antagonists on EAT growth, macrophage spreading ability and the release of H202 during tumor growth.
EXPERIMENTAL PROCEDURES
Animals Mice of the Swiss strain, male, 4 - 6 weeks old, from our own animal facilities were used throughout the experiments.
Ehrlich ascites tumor The tumor was maintained in Swiss mice in the ascitic form. Tumor cells were collected by aspiration with a Pasteur pipette, centrifuged for 10 min at 200 x g, followed by two washes with phosphate buffered saline (PBS), pH 7.2. In all experimental protocols described, mice were injected i.p. with 1 x 103 tumor cells. Cell viability was assessed by trypan blue exclusion test and only the cell suspensions which presented more than 95% viability were used.
Cell harvesting and counting The cells were harvested by peritoneal lavage with 3 ml of sterile PBS and the number of cells determined by counting in a hemocytometer. Differential counts were performed on fixed and stained cell suspensions (0.05% crystal violet dissolved in 30% acetic acid).
Culture of EA T cells in vitro EAT-cells obtained as described above were suspended in Dulbecco modified Eagle's medium supplemented with L-glutamine (216 ng/ml), 2-mercaptoethanol (5 × 10 -6 M), 10 mM Hepes, pholic acid (6/ag/ml), pyruvic acid (110/ag/ml), arginine (116/ag/ml), aspargine (36 mg/1) and normal mouse serum (0.5%). Penicillin (63/~g/ml) and streptomycin (110/ag/ml) were added. EAT cells (1 x 10~/ml) were plated in duplicate T-25 flasks at 37°C in humidified 95% air, 5% CO2 for 48 h. Under these conditions the number of cells increased five fold in 48 h. In the experiments where BN 52021 was added,
the media containing the compounds were sterilized by Millipore filtration.
Assay for spreading of macrophages The spreading ability of peritoneal macrophages from tumor-bearing or control mice was assessed according to the technique described by Rabinovitch, Manejias, Russo & Abbey (1977). Briefly, aliquots of 0.2 ml of the peritoneal cell suspensions were placed over glass coverslips and incubated for 15 min at 37°C. The non-adherent cells were removed by washing with PBS and the adherent cells were incubated in culture medium 199 containing 10 nM Hepes at 37°C in humidified chamber for 1 h. Following this, medium was removed and the cells were fixed with 2.5% glutaraldehyde and examined under phase contrast microscopy. The spread cells with pseudopod-like structures and the round cells were counted.
Assay for adherence-induced release of 1-1202 by peritoneal cells The production of H202 by peritoneal cells was measured by horseradish peroxidase - - phenol red oxidation method developed by Pick & Keisari (1980) and modified by Russo, Teixeira, Marcondes & Barbuto, in press). The technique employed allows the determination of spontaneous, adherenceinduced release of H2Oz, which is independent of further stimulation. Suspensions of peritoneal cells were adjusted to 1 × 106/ml, centrifuged (200 × g for 10 min at 4°C) and suspended again in 1 ml of phenol red buffer containing 50/ag/ml of horseradish peroxidase (type II, Sigma). Aliquots of 100/al were then transferred to 96-well, flat-bottom tissue culture plates (Corning, N.Y.) and incubated for 1 h at 37°C in humidified chamber (5% CO2 in air). The reaction was stopped by addition of 10 tal of 1N NaOH and the optical density at 620 nm was determined on a micro ELISA reader (Titertek Multiscan, Flow Lab.). All determinations were repeated eight times and the absorbance transformed into nanomoles of H202 by deduction from a standard curve. Values below 0.1 nmol were not considered.
Treatment with P A F -
antagonists
BN 52021 (Institut Henri Beafour, France) and SRI 63441 (Sandoz Res. Inst., U.S.A.), were injected one or two times a day at doses of 1.0, 2.5; 5.0 and 20.0 mg/kg. The antagonists were given intraperitoneally (i.p.), intravenously (i.v.) or subcutaneously (s.c.). In parallel, groups of mice were treated with the vehicle using the same protocols above. BN 52021
Inhibition of Ehrlich Ascites Tumor
59
Table 1. Effect of PAF antagonists on the growth of EAT cells in vivo
5 --
Treatment (doses/day)
4
Route of administration
Concentration (mg/kg)
Inhibition (%)
BN 52021 (2 x )
i.p.
BN 52021
i.p.
5.0 2.5 1.0 5.0
80.8 76.9 33.1 37.3
i.v.
5.0
56.0
i.v.
5.0
14.0
s.c.
20.0
zero
i.p.
5.0
80.4
i.p.
5.0
38.3
i.p. or i.v.
0.1 ml
zero
t
E
x
I' /
× 3
/ i /
(lx)
i /
BN 52021 (2x) BN 52021
i
~2 i O Z t
i
i
i
it i
(ix)
i
BN 52021
(ix) SRI-63441 (2x) SRI-63441
I
6
IO
Time (dGys)
Fig. I. Number of cells present in the peritoneal cavity l, 6 and 10 days after i.p. inoculation of 10 3 EAT cells in mice treated with BN 52021 (5 mg/kg, 2 × day, i.p.) ( 0 ) or with the vehicle (0.1 ml, 2 × day, i.p.) (D). Number of cells present in the peritoneal cavity of normal mice treated with BN 52021 as above (A) and in non-manipulated mice (Ill). Each point represents the mean _+ S.E.M. from 10-24 animals. **P<0.001. was dissolved to 5 m g / m l and further diluted with sterile, pyrogen free saline solution. The c o m p o u n d s were m a d e up fresh immediately before use.
Statistical analyses Analyses of variance followed by multiple comparisons according to Tukey were used to analyse the differences between the means. The probability level o f P-.<0.05 was considered significant (Zar, 1984).
RESULTS
Proliferation of E A T in vivo - - effect of PAF antagonists Ehrlich tumor cells (1 × 103) were implanted into mice peritoneal cavity and after 1, 6 and 10 days the animals were sacrificed and the number of cells present in the cavity was determined. It can be seen in Fig. 1 that the growth of E A T was initially inhibited by the host until the sixth day. During this period, the number of cells remained close to that present in the peritoneal cavity o f normal mice
(lx) VEHICLE
(2x) *Percent inhibition of the number of cells in the peritoneal cavity of treated compared to non-treated mice, 10 days after i.p. inoculation of 1 x 103 EAT cells. Data represent the mean from 6 - 2 5 animals. (0.21 × 107/ml). However, after 10 days, the number of cells markedly increased (4.42 × 107/ml). Groups of mice received the P A F antagonist, BN 52021, 5 m g / k g , twice a day, i.p., throughout the time-course of the experiments. Control animals were similarly treated with the vehicle alone. It was found that BN 52021 significantly reduced (80.8%) the number of cells in the peritoneal cavity on the tenth day after tumor implantation. The results are expressed as the total number of peritoneal cells. It should be pointed out that when considering the cells with E A T cell morphology only, the inhibition was even more pronounced (90%) after 10 days. These results are not presented because morphological criteria, although providing a considerable degree of reliability, are not sufficient for a precise identification of the tumor cells. The efficiency of BN 52021 in inhibiting the growth of E A T was also investigated using different doses and routes of administration of the antagonist. Table 1 shows that i.p. administration of BN 52021, twice a day, for 10 days, was the most effective protocol for inhibiting E A T growth. An inhibition of 80.8, 76.9 and 33.1% was observed with 5.0, 2.5 and 1.0 m g / k g of the antagonist, respectively. A single dose of 5 m g / k g was less effective (37.3%) and s.c. administration was ineffective. Intravenous injection of the antagonist, 5 m g / k g / t w i c e a day,
D. FECCHIOet al.
60
Table 2. Effect of BN 52021 on the proliferation of EAT cells in culture
EAT cells in culture with*
Cells after 48 h in culture ( X 106/ml)
Viability after 48 h in culture (%)
5.73 5.37 5.31 5.22
94.5 94.0 93.5 96.0
-BN 52021 (10 -3 M) BN 52021 (10 -4 M) Vehicle
*1 x 106 EAT cells/ml were cultured in DMEM, supplemented, for 48 h. BN 52021 was added at the beginning and after 24 h. Viability assessed by trypan blue exclusion test. IO(3
BN 52021 observed in viva could be ascribed to a direct effect of the drug on tumor cells. E A T cells (1 x 106/ml) were cultured for 48 h in the presence of 1 0 - 3 M and 1 0 - 4 M of BN 52021, or the equivalent volume of the vehicle. The antagonist was added at the beginning of the culture and 24 h later. It was found that after 48 h, the number of E A T cells increased to 5.7 x 106/ml. BN 52021 did not affect the cell proliferation at any of the concentrations tested. Cell viability, as determined by trypan blue exclusion, ranged f r o m 9 3 . 5 - 9 6 . 0 % and was also not affected by the addition of BN 52021 (Table 2).
i
50
-..
l \ \
'\ \ \
I
I
I
I
6
PO
Time (days)
Fig. 2. Spreading (°7o) of peritoneal macrophages l, 6 and 10 days after i.p. inoculation of 1 x 103 EAT cells in mice treated with BN 52021 (5 mg/kg, 2 x day, i.p.) (A) or with the vehicle (0.1 ml, 2 x day, i.p.) (0). Spreading of peritoneal macrophages from normal mice treated with BN 52021 as above (I) and from non manipulated mice (O). Each point represents the mean _+ S.E.M. from 6 - 1 5 animals. *P<0.001. significantly reduced tumor growth (56°7o). Similar inhibition was observed when the animals were treated i.p. with another P A F antagonist, SRI 63441, twice a day (80.4%) or once a day (38.3%). It was also observed that if the treatment with BN 52021 was interrupted, after 10 days of continuous treatment, the tumor regained its capacity to proliferate. One week after the treatment was interrupted the number of cells, in the group that had been treated with the P A F antagonist, was similar to that found in the non-treated group (10.43 +_ 1.15 x 107 and 8.65 _ 0.92 × 107 cells/ml, respectively).
Effect o f B N 52021 on the proliferation o f E A T in vitro The following experiments were performed in order to investigate whether the inhibitory effect of
Effect o f B N 52021 on peritoneal macrophages f r o m tumor-bearing mice Mice inoculated with 1 × 103 E A T cells were injected i.p., twice a day, with 5 m g / k g of BN 52021 or an equivalent volume of the vehicle. After 1, 6 and 10 days, the animals were sacrificed and the peritoneal cells collected and assayed for spreading ability and release of H202. Macrophages obtained from the peritoneal cavity of normal mice showed very low capacity of rapid spreading on a glass surface (6.9%). It can be seen in Fig. 2 that in tumor-bearing mice, spreading increased to 73.3% in the first day after tumor implantation and decreased thereafter, falling to control levels at 6 and 10 days. In the tumor-bearing mice treated with BN 52021, the macrophage spreading also increased on the first day (83.3o7o), but remained at high levels throughout the experiments. This effect is illustrated in Fig. 3, showing monolayers of mouse peritoneal macrophages (adherent cells) from tumor-bearing mice which exhibit very few cells, most of them with round m o r p h o l o g y (Fig. 3a). In contrast, in the m o n a layers from BN 52021 treated mice the number of cells is strikingly increased and most of the cells have a spread morphology with pseudopod-like elongated
Inhibition of Ehrlich Ascites Tumor
61
Fig. 3. Spreading of peritoneal cells from tumor bearing mice, 10 days after i.p. inoculation of 103 EAT cells in mice treated with BN 52021 (5 mg/kg, 2 x day, i.p.) (b) and mice treated with the vehicle (0.1 ml, 2 x day, i.p.) (a). Phase contrast microscopy, 300 × .
Inhibition of Ehrlich Ascites Tumor
63 DISCUSSION
7/
* I
"".
" ~ o5
~"
-~ ±
z
,
--O---II
~
~
Time (days)
Fig. 4. Spontaneous release of H2Oz (nmol/105 peritoneal cells), 1, 6 and 10 days after i.p. inoculation of 1 × 103 EAT cells, in mice treated with BN 52021 (5 mg/kg, 2 × a day, i.p.) (A) or treated with 0.1 ml of the vehicle (0) and from normal mice treated with BN 52021, same dose as above, (11). Each point represents the mean _+S.E.M. from 10 to 14 animals. *P<0.05. structures (Fig. 3b). Interestingly, normal mice treated with BN 52021 also exhibited high levels of macrophage spreading (Fig. 2). With lower doses of BN 52021, it was found that 2.5 mg/kg stimulated spreading in normal as well as in EAT-bearing animals (59.3%0 and 88.5 o70, respectively, at 10days), and 1 mg/kgstimulated spreading only in EAT-bearing mice (62.7%). For evaluation of H202 release, 1 × 105 peritoneal cells, from the different experimental groups, were analysed individually, according to the technique described under Experimental Procedures. It was found (Fig. 4) that peritoneal cells from normal mice and from EAT-bearing mice, did not release detectable amounts of H202. However, treatment with BN 52021, significantly increased H202 release from peritoneal cells obtained 6 days after tumor implantation (0.54 nmol/105 cells) as well as from those obtained from normal mice. Care should be taken in interpreting the data obtained with the samples collected 10 days after tumor implantation. At this time, the tumor cells account for 8 0 - 90°7o of the peritoneal cell population, and consequently the number of macrophages per sample is very low. Under the experimental conditions employed, it was not possible to work with isolated macrophages since any kind of manipulation would interfere with the assay of the spontaneous, adherence-induced release of H202.
The results presented here show that two PAF antagonists inhibited the growth of Ehrlich ascites tumor in viva. BN 52021 dose-dependently inhibited the tumor growth when injected at the site of tumor growth. It is difficult to understand why other routes of administration were less effective. The only possible explanation is that the antagonist is needed in high concentration at the site of tumor/host cell interaction. This effect of BN 52021 cannot be ascribed to a toxic or anti-mitotic effect, since addition of the antagonist to EAT cells in culture did not affect cell viability and proliferation. These experiments, however, do not exclude the possibility that the compound may interfere with the metabolism of the tumor cell, affecting the synthesis or release of substances which potentially protect the tumor in vivo. In fact, several reports in the literature suggest that tumor cells are able to produce substances which depress host defenses. Products generated by tumor cells were shown to inhibit macrophage spreading, adherence, migration and oxidative metabolism. These substances are either proteins (Szuro-Sudol & Nathan, 1982) or low molecular weight, lipid-like substances (Cantarrow, Cheung & Sundharadas, 1978). Because macrophages operate at the tumor-host interface they may critically influence tumor growth. In the present work we have shown that inhibition of EAT growth occurred concomitantly with activation of peritoneal macrophages as assessed by the spreading ability and release of H202. We observed that the presence of the tumor cells apparently did not activate the local macrophages. Although EAT induced spreading one day after implantation, this was not observed later on. Treatment of the mice with BN 52021 had a marked effect on macrophages, stimulating spreading and release of H202. After 10 days of tumor implantation it is difficult to be certain of the results of H202 release by the method employed, because the number of tumor cells is very high, at this time. Nevertheless, on the sixth day, when the number of cells in both groups (treated and non-treated with BN 52021) is similar, it was clearly observed that only macrophages from the treated group released H202. Although it seems that tumor inhibition is associated with macrophage activation, care should be taken in the interpretation of these results. When looking at individual data we observed that such a direct association is not always observed. The observation that treatment with the PAF antagonist also promoted activation of macrophages from normal mice allows two hypotheses: (i) that the
64
D. FECCHIOet al.
antagonist is a partial agonist; or (ii) that the agonist (PAF) is an endogenous modulator of macrophage function. Since the PAF antagonist usually has no agonistic effect on platelets and other tissues, the latter hypothesis seems more plausible. In the experiments where administration of BN 52021 was interrupted after 10 days of treatment, the tumor regained its capacity to grow, suggesting that the effect of the compound is reversible. This result also suggests that the antagonist needs to be present during the tumor growth to exert its inhibitory activity. Inhibition of EAT growth by PAF antagonists also indicates that PAF or PAF-like substances are involved in the host-tumor interaction, possibility favoring tumor growth. Since inhibition of EAT growth was parallel with macrophage activation, the possibility exists that macrophages are the targets for the action of PAF or PAF-like substances. Recently, several experimental data have appeared in the literature which indicate that PAF may have an important role in macrophage function. PAF regulates IL-I production, either stimulating (low doses) or inhibiting (high doses) the release of IL-1 from LPS-stimulated monocytes (Pignol, Henane, Mencia-Huerta, Rola-Pleszczynski & Braquet, 1987). It is also known that PAF can induce release of cyclo-oxygenase metabolites and that prostaglandins are important modulators of macrophage function as well as of the immune response (Goodwin & Webb, 1980). In addition, we have found high levels of PGE2 in the ascitic fluid from Ehrlich tumor (Fecchio et al., in press). Since release of prostaglandins is a consequence of activation of phospholipid metabolism, it is possible that PAF is also generated. PAF and prostaglandins have been shown to inhibit the expression of class II antigens in macrophage cellular membrane (Gebhardt, Bazan & Bazan, 1988; Snyder, Beller & Unaue, 1982), thus affecting macrophage/lymphocyte interaction with
consequences on the specific immune response. Indeed, we cannot exclude the possibility that PAF antagonists inhibit tumor growth by affecting the immune response. PAF has been described to suppress alloantigen recognition in MLC and generation of cytotoxic lymphocytes in vitro, the latter effect being reversed by BN 52021 (Gebhardt et al., 1988). This mediator has also been shown to inhibit lymphocyte proliferation and IL-2 production and to generate suppressor cells, either directly or by inducing the release of cyclo-oxygenase metabolites from monocytes (Rola-Pleszczynski, Pignol, Pouliot & Braquet, 1987a,b). The cell type responsible for the release of PAF or PAF-like substances in the Ehrlich Ascites Tumor remains to be determined. Macrophages are known to produce PAF (Mencia-Huerta & Benveniste, 1979) and unusually high levels of O-alkyl glycero phospholipids have been detected in several mammalian tumors including EAT (Snyder & Wood, 1968; Friedberg, Halpert & Barnwell, 1985). Overall, our data suggest that tumor-derived or effector cell-derived factor(s) inhibits macrophages and that this inhibition favors tumor growth. Since PAF antagonists inhibited tumor growth and stimulated macrophages, the possibility exists that the inhibitory factor is PAF or PAF-like. Nevertheless, as pointed out in the discussion, several points remain obscure and more work is needed before we can venture any hypothesis. Unravelling the cellular and molecular mechanisms that govern the inflammatory component of tumor development should provide a better understanding of the nature of the tumor-host interactions and eventually the possibility of modulating these interactions in favor of the host. Acknowledgements - - The authors wish to acknowledge
Funda~o de Amparo ~ Pesquisa do Estado de S~o Paulo (FAPESP) for financial support and Miss Grazia de Giacobbi for technical assistance.
REFERENCES
BRAQUET,P. & ROLA-PLESZCZYNSKI,M. (1987). Platelet activating factor and cellular immune responses. Immun. Today, 8, 345 - 352. CANTARROW,W. D., CHEUNG,H. T. • SUNDHARADAS,G. (1978). Modulation of spreading, adhesion and migration of peritoneal macrophages by a low molecular weight factor extracted from mouse tumors. J. Reticuloendothel. Soc., 24, 657 - 666. ECCLES,S. A. & ALEXANDER,P. (1974). Macrophages content of tumors in relation to metastatic spread and host immune reaction. Nature, Lond., 250, 667-669.
Inhibition of Ehrlich Ascites Tumor
65
EVANS, R. (1973). Macrophages and tumor bearing host. Br. J. Cancer, 28, 19-25. FAUVE, R. M., HEVIN, B., JACOB, H., GAIZZARD, J. A. & JACOB, F. (1974). Anti-inflammatory effects of murine malignant cells. Proc. natn. Acad. Sci. U.S.A., 10, 4052-4056. FECCHIO, D., SIRO~S,P., RUSSO, M. & JANCAR,S. Studies on the inflammatory response induced by Ehrlich tumor in mice peritoneal cavity. Inflammation. In press. FRIEDBERC, S. J., HALPERT, M. & BARNWELL, G. M. (1985). The rate of formation of surface membrane ether lipids in Ehrlich ascites tumor cells: Kinetic consideration. Arch. Biochern. Biophys., 243, 504-514. GEBHARDT, B. M., BAZAN, H. E. P. & BAZAN,N. G. (1988). Ginkgolide BN 52021 blocks PAF-mediated suppression of cellular immunity. In The Ginkgolides - - Chemistry, Biology, Pharmacology and Clinical Perspectives (ed. Braquet, P.) pp. 703-729. J. R. Prous, Barcelona. GOODWIN, J . S . & WEBB, D . R . (1980). Regulation of the immune response by prostaglandins. Clin. Imrnun. [mrnunopath., 15, 106-122. MENCIA-HUERTA, J. M. & BENVENISTE,J. (1979). Platelet-activating factor and macrophages. I. Evidence for the release from rat and mouse peritoneal macrophages and not from mastocytes. Eur. J. Immun., 9, 409- 415. PICK, E. & KEISARI,Y. (1980). A simple colorimetric method for measurement of hydrogen peroxide produced by cells in culture. J. Imrnunol. Methods, 38, 161- 172. PIrCE, M. C. & SNYDERMAN,R. (1976). Depression of macrophage function by a factor produced by mouse neoplasms: A mechanism of abrogation of immune surveillance. J. lmrnun., 117, 1243- 1249. PIGNOL, B., HENANE, S., MENCIA-HUERTA,J. M., ROLA-PLESZCZYNSKI,M. & BRAQUET, P. (1987). Effect of PAF-acether and its specific antagonist, BN 52021, on interleukin l (IL-1) synthesis and release by rat monocytes. Prostaglandins, 33, 391 - 396. RABINOVITCH, M., MANEJIAS, R. E., RUSSO, M. & ABBEY, E. E. (1977). Increased spreading of macrophages from mice treated with interferon inducers. Cell Irnrnun., 29, 86-95. ROLA-PLESZCZYNSKI,M. (1985). Immunoregulation by leukotrienes and other lipoxygenase metabolites. Irnrnun. Today, 6, 302 - 307. ROLA-PEESZCZYNSKI,M., PIGNOL, B., POULIOT, C. & BRAQUET,P. (1987a). Inhibition of human lymphocyte proliferation and interleukin 2 production by platelet-activating factor (PAF-Acether). Reversal by a specific antagonist, BN 52021. Biochern. Biophys. Res. Cornrnun., 142, 745- 760. ROEA-PLESZCZYNSKI, M., POULIOT, C., PIGNOL, B. & BRAQUET, P. (1987b). Platelet-activating factor induces human suppressor cell activity. Fedn Proc. Fedn Am. Socs exp. Biol., 46, 444-450. Russo, M., TEIXEIRA, H. L., MARCONDES, M. C. G, & BARBUTO, J. A. M. Superoxide-indpendent hydrogen peroxide release by activatd macrophages. Brazilian J. ivied. Biol. Res. In press. RUSSEL, S. W., GILLESPIE, G. Y., HANSEN, C. B. & COCHRANE, C. G. (1976). Inflammatory cells in solid murine neoplasms. II-cell types found as the cause of Moleney sarcoma regression or progression. Int. J. Cancer, 18, 331-338. SNVDER, F. & WOOD, R. (1968). The occurrence and metabolism of alkyl and alk-1-enyl ethers of glycerol in transplantable rat and mouse tumors. Cancer. Res., 28, 972-978. SNVDER, D. S., BELLER,D. I. & UNAUE,E. R. (1982). Prostaglandins modulate macrophage Ia expression. Nature, Lond., 299, 163 - 165. SZt3Ro-SUDOL, A. & NATHAr~, C. F. (1982). Suppression of macrophage oxidative metabolism by products of malignant and non-malignant cells. J. exp. Med., 156, 945- 961. ZAR, J. H. (1984). Biostatistical Analysis. pp. 162-205. Prentice-Hall Inc., New Jersey.
0192-0561/90 $3.00 + .00 © 1989 International Society for Immunopharmacology.
I N H I B I T I O N OF E H R L I C H ASCITES T U M O R I N VIVO BY PAF-ANTAGONISTS DENISE FECCHIO, MOMTCHILO RUSSO, PIERRE SIROIS,* PIERRE BRAQUET t a n d SONIA JANCAR* Departamento de Imunologia, Instituto de Ci~ncias Biom6dicas, Universidade de S~o Paulo, S~o Paulo, Brazil; *Department of Pharmacology, University of Sherbrooke, Sherbrooke, Quebec, Canada; tlnstitut Henri Beaufour Research Laboratories, France (Received 20 December 1988 and in final form 10 July 1989)
Abstract - - Several lines of evidence support that PAF modulates the inflammatory and immune responses,
and that tumors may inhibit both these processes. In the present study we analysed the effect of PAF antagonists on the growth of Ehrlich Ascites Tumor (EAT) in vivo. Mice were inoculated intraperitoneally with 1 × 10 3 EAT ceils and the tumor growth evaluated by counting the number of peritoneal cells, 1, 6 and 10 days after tumor implantation. BN 52021 was administered intraperitoneally, intravenously or subcutaneously once or twice a day, at 1.0, 2.5, 5.0 and 20.0 mg/kg. Control animals received 0.1 ml of the vehicle in the same schedule. It was found that i.p. and i.v. administration of BN 52021 (5 mg/kg, twice a day) significantly inhibited EAT growth (80.8070 and 56.0070 respectively). Other routes and doses were less effective. Another PAF antagonist, SRI 63441 (5 mg/kg, i.p., twice a day) also inhibited EAT growth (80.4°70). The BN 52021 added to EAT cells in culture, at concentration of 10 3 and 10 -4 M, did not affect the viability and proliferation of tumors cells. In an attempt to understand the mechanism of this inhibition, we analyzed the peritoneal macrophages for spreading ability and H202 release. It was found that 24 h after tumor implantation there was an increase in the spreading ability of peritoneal macrophages (75070) and that, as the tumor grew, the spreading index fell to control levels (<10070). In the groups treated with BN 52021 (5 mg/kg/twice a day) the spreading remained elevated (50-60070) at all the times examined. Release of H202, measured by horseradish peroxidase-phenol red oxidation, was below detectable levels throughout tumor growth. Treatment with BN 52021 as above, significantly stimulated the release of H202, 6 and 10 days after tumor implantation (0.4 to 0.6 nmol/105 cells). The data presented show that PAF antagonists inhibit EAT growth in vivo and that this inhibition is concomitant with activation of peritoneal macrophages. These results suggest that macrophages can critically influence tumor growth and that PAF or PAF-like substances may modulate macrophages-tumor interactions.
their p r o d u c t s o n the evolution o f t u m o r growth. A m o n g the p r o d u c t s released by i n f l a m m a t o r y cells, lipid m e d i a t o r s have been s h o w n to play a n essential role in p h a g o c y t e f u n c t i o n s a n d even to interact with i m m u n o c o m p e t e n t cells. P r o s t a g l a n d i n s , particularly PGE2, are k n o w n to exert a negative feedback o n m a c r o p h a g e s a n d to m a r k e d l y affect lymphocytes (for review see G o o d w i n & W e b b , 1984). Evidence is a c c u m u l a t i n g t h a t leukotriene B 4 a n d platelet activating factor ( P A F ) are also very p o w e r f u l i m m u n o r e g u l a t o r s (for review see Rola-Pleszczynski, 1985; B r a q u e t & Rola-Pleszczynski, 1987). E h r l i c h t u m o r can be t r a n s p l a n t e d into the peritoneal cavity o f mice where it grows in ascitic
O n e o f the factors w h i c h m o d u l a t e s t u m o r g r o w t h is the intensity o f the i n f l a m m a t o r y response evoked by the t u m o r cells. Several reports in the literature p o i n t to a n i m p o r t a n t role for i n f l a m m a t o r y cells in the c o n t r o l o f t u m o r g r o w t h (Evans, 1973; Eclles & A l e x a n d e r , 1974; Russel, Gillespie, H a n s e n & C o c h r a n e , 1976). Conversely, the presence o f the t u m o r results in a n i m p a i r m e n t o f m a c r o p h a g e a n d i n f l a m m a t o r y responses in the host a n d this m a y be m e d i a t e d by product(s) released by the t u m o r (Fauve, H e v i n , J a c o b , G a i z z a r d & J a c o b , 1974; Pike & S n y d e r m a n , 1976). T h u s , as i n f l a m m a t i o n usually initiates the m o r e specific i m m u n e response, we felt it relevant to study the role o f i n f l a m m a t o r y cells a n d
:l:Author to whom correspondence should be addressed at: Departamento de Imunologia, Instituto de Ci~ncias Biom6dicas, USP, Caixa Postal 4365, CEP 01051, S~o Paulo (SP), Brazil. 57
D. FECCHIOet al.
58
form (EAT). We have recently shown that the growth of EAT induces a very low influx of polymorphonuclear leukocytes but does not stimulate local macrophages, as assessed by spreading and release of hydrogen peroxide (H2Oz) (Fecchio, Sirois, Russo & Jancar, in press). We have also shown that significant amounts of PGE2 are released into the peritoneal cavity following EAT implantation. In the present study we analysed the effect of PAF antagonists on EAT growth, macrophage spreading ability and the release of H202 during tumor growth.
EXPERIMENTAL PROCEDURES
Animals Mice of the Swiss strain, male, 4 - 6 weeks old, from our own animal facilities were used throughout the experiments.
Ehrlich ascites tumor The tumor was maintained in Swiss mice in the ascitic form. Tumor cells were collected by aspiration with a Pasteur pipette, centrifuged for 10 min at 200 x g, followed by two washes with phosphate buffered saline (PBS), pH 7.2. In all experimental protocols described, mice were injected i.p. with 1 x 103 tumor cells. Cell viability was assessed by trypan blue exclusion test and only the cell suspensions which presented more than 95% viability were used.
Cell harvesting and counting The cells were harvested by peritoneal lavage with 3 ml of sterile PBS and the number of cells determined by counting in a hemocytometer. Differential counts were performed on fixed and stained cell suspensions (0.05% crystal violet dissolved in 30% acetic acid).
Culture of EA T cells in vitro EAT-cells obtained as described above were suspended in Dulbecco modified Eagle's medium supplemented with L-glutamine (216 ng/ml), 2-mercaptoethanol (5 × 10 -6 M), 10 mM Hepes, pholic acid (6/ag/ml), pyruvic acid (110/ag/ml), arginine (116/ag/ml), aspargine (36 mg/1) and normal mouse serum (0.5%). Penicillin (63/~g/ml) and streptomycin (110/ag/ml) were added. EAT cells (1 x 10~/ml) were plated in duplicate T-25 flasks at 37°C in humidified 95% air, 5% CO2 for 48 h. Under these conditions the number of cells increased five fold in 48 h. In the experiments where BN 52021 was added,
the media containing the compounds were sterilized by Millipore filtration.
Assay for spreading of macrophages The spreading ability of peritoneal macrophages from tumor-bearing or control mice was assessed according to the technique described by Rabinovitch, Manejias, Russo & Abbey (1977). Briefly, aliquots of 0.2 ml of the peritoneal cell suspensions were placed over glass coverslips and incubated for 15 min at 37°C. The non-adherent cells were removed by washing with PBS and the adherent cells were incubated in culture medium 199 containing 10 nM Hepes at 37°C in humidified chamber for 1 h. Following this, medium was removed and the cells were fixed with 2.5% glutaraldehyde and examined under phase contrast microscopy. The spread cells with pseudopod-like structures and the round cells were counted.
Assay for adherence-induced release of 1-1202 by peritoneal cells The production of H202 by peritoneal cells was measured by horseradish peroxidase - - phenol red oxidation method developed by Pick & Keisari (1980) and modified by Russo, Teixeira, Marcondes & Barbuto, in press). The technique employed allows the determination of spontaneous, adherenceinduced release of H2Oz, which is independent of further stimulation. Suspensions of peritoneal cells were adjusted to 1 × 106/ml, centrifuged (200 × g for 10 min at 4°C) and suspended again in 1 ml of phenol red buffer containing 50/ag/ml of horseradish peroxidase (type II, Sigma). Aliquots of 100/al were then transferred to 96-well, flat-bottom tissue culture plates (Corning, N.Y.) and incubated for 1 h at 37°C in humidified chamber (5% CO2 in air). The reaction was stopped by addition of 10 tal of 1N NaOH and the optical density at 620 nm was determined on a micro ELISA reader (Titertek Multiscan, Flow Lab.). All determinations were repeated eight times and the absorbance transformed into nanomoles of H202 by deduction from a standard curve. Values below 0.1 nmol were not considered.
Treatment with P A F -
antagonists
BN 52021 (Institut Henri Beafour, France) and SRI 63441 (Sandoz Res. Inst., U.S.A.), were injected one or two times a day at doses of 1.0, 2.5; 5.0 and 20.0 mg/kg. The antagonists were given intraperitoneally (i.p.), intravenously (i.v.) or subcutaneously (s.c.). In parallel, groups of mice were treated with the vehicle using the same protocols above. BN 52021
Inhibition of Ehrlich Ascites Tumor
59
Table 1. Effect of PAF antagonists on the growth of EAT cells in vivo
5 --
Treatment (doses/day)
4
Route of administration
Concentration (mg/kg)
Inhibition (%)
BN 52021 (2 x )
i.p.
BN 52021
i.p.
5.0 2.5 1.0 5.0
80.8 76.9 33.1 37.3
i.v.
5.0
56.0
i.v.
5.0
14.0
s.c.
20.0
zero
i.p.
5.0
80.4
i.p.
5.0
38.3
i.p. or i.v.
0.1 ml
zero
t
E
x
I' /
× 3
/ i /
(lx)
i /
BN 52021 (2x) BN 52021
i
~2 i O Z t
i
i
i
it i
(ix)
i
BN 52021
(ix) SRI-63441 (2x) SRI-63441
I
6
IO
Time (dGys)
Fig. I. Number of cells present in the peritoneal cavity l, 6 and 10 days after i.p. inoculation of 10 3 EAT cells in mice treated with BN 52021 (5 mg/kg, 2 × day, i.p.) ( 0 ) or with the vehicle (0.1 ml, 2 × day, i.p.) (D). Number of cells present in the peritoneal cavity of normal mice treated with BN 52021 as above (A) and in non-manipulated mice (Ill). Each point represents the mean _+ S.E.M. from 10-24 animals. **P<0.001. was dissolved to 5 m g / m l and further diluted with sterile, pyrogen free saline solution. The c o m p o u n d s were m a d e up fresh immediately before use.
Statistical analyses Analyses of variance followed by multiple comparisons according to Tukey were used to analyse the differences between the means. The probability level o f P-.<0.05 was considered significant (Zar, 1984).
RESULTS
Proliferation of E A T in vivo - - effect of PAF antagonists Ehrlich tumor cells (1 × 103) were implanted into mice peritoneal cavity and after 1, 6 and 10 days the animals were sacrificed and the number of cells present in the cavity was determined. It can be seen in Fig. 1 that the growth of E A T was initially inhibited by the host until the sixth day. During this period, the number of cells remained close to that present in the peritoneal cavity o f normal mice
(lx) VEHICLE
(2x) *Percent inhibition of the number of cells in the peritoneal cavity of treated compared to non-treated mice, 10 days after i.p. inoculation of 1 x 103 EAT cells. Data represent the mean from 6 - 2 5 animals. (0.21 × 107/ml). However, after 10 days, the number of cells markedly increased (4.42 × 107/ml). Groups of mice received the P A F antagonist, BN 52021, 5 m g / k g , twice a day, i.p., throughout the time-course of the experiments. Control animals were similarly treated with the vehicle alone. It was found that BN 52021 significantly reduced (80.8%) the number of cells in the peritoneal cavity on the tenth day after tumor implantation. The results are expressed as the total number of peritoneal cells. It should be pointed out that when considering the cells with E A T cell morphology only, the inhibition was even more pronounced (90%) after 10 days. These results are not presented because morphological criteria, although providing a considerable degree of reliability, are not sufficient for a precise identification of the tumor cells. The efficiency of BN 52021 in inhibiting the growth of E A T was also investigated using different doses and routes of administration of the antagonist. Table 1 shows that i.p. administration of BN 52021, twice a day, for 10 days, was the most effective protocol for inhibiting E A T growth. An inhibition of 80.8, 76.9 and 33.1% was observed with 5.0, 2.5 and 1.0 m g / k g of the antagonist, respectively. A single dose of 5 m g / k g was less effective (37.3%) and s.c. administration was ineffective. Intravenous injection of the antagonist, 5 m g / k g / t w i c e a day,
D. FECCHIOet al.
60
Table 2. Effect of BN 52021 on the proliferation of EAT cells in culture
EAT cells in culture with*
Cells after 48 h in culture ( X 106/ml)
Viability after 48 h in culture (%)
5.73 5.37 5.31 5.22
94.5 94.0 93.5 96.0
-BN 52021 (10 -3 M) BN 52021 (10 -4 M) Vehicle
*1 x 106 EAT cells/ml were cultured in DMEM, supplemented, for 48 h. BN 52021 was added at the beginning and after 24 h. Viability assessed by trypan blue exclusion test. IO(3
BN 52021 observed in viva could be ascribed to a direct effect of the drug on tumor cells. E A T cells (1 x 106/ml) were cultured for 48 h in the presence of 1 0 - 3 M and 1 0 - 4 M of BN 52021, or the equivalent volume of the vehicle. The antagonist was added at the beginning of the culture and 24 h later. It was found that after 48 h, the number of E A T cells increased to 5.7 x 106/ml. BN 52021 did not affect the cell proliferation at any of the concentrations tested. Cell viability, as determined by trypan blue exclusion, ranged f r o m 9 3 . 5 - 9 6 . 0 % and was also not affected by the addition of BN 52021 (Table 2).
i
50
-..
l \ \
'\ \ \
I
I
I
I
6
PO
Time (days)
Fig. 2. Spreading (°7o) of peritoneal macrophages l, 6 and 10 days after i.p. inoculation of 1 x 103 EAT cells in mice treated with BN 52021 (5 mg/kg, 2 x day, i.p.) (A) or with the vehicle (0.1 ml, 2 x day, i.p.) (0). Spreading of peritoneal macrophages from normal mice treated with BN 52021 as above (I) and from non manipulated mice (O). Each point represents the mean _+ S.E.M. from 6 - 1 5 animals. *P<0.001. significantly reduced tumor growth (56°7o). Similar inhibition was observed when the animals were treated i.p. with another P A F antagonist, SRI 63441, twice a day (80.4%) or once a day (38.3%). It was also observed that if the treatment with BN 52021 was interrupted, after 10 days of continuous treatment, the tumor regained its capacity to proliferate. One week after the treatment was interrupted the number of cells, in the group that had been treated with the P A F antagonist, was similar to that found in the non-treated group (10.43 +_ 1.15 x 107 and 8.65 _ 0.92 × 107 cells/ml, respectively).
Effect o f B N 52021 on the proliferation o f E A T in vitro The following experiments were performed in order to investigate whether the inhibitory effect of
Effect o f B N 52021 on peritoneal macrophages f r o m tumor-bearing mice Mice inoculated with 1 × 103 E A T cells were injected i.p., twice a day, with 5 m g / k g of BN 52021 or an equivalent volume of the vehicle. After 1, 6 and 10 days, the animals were sacrificed and the peritoneal cells collected and assayed for spreading ability and release of H202. Macrophages obtained from the peritoneal cavity of normal mice showed very low capacity of rapid spreading on a glass surface (6.9%). It can be seen in Fig. 2 that in tumor-bearing mice, spreading increased to 73.3% in the first day after tumor implantation and decreased thereafter, falling to control levels at 6 and 10 days. In the tumor-bearing mice treated with BN 52021, the macrophage spreading also increased on the first day (83.3o7o), but remained at high levels throughout the experiments. This effect is illustrated in Fig. 3, showing monolayers of mouse peritoneal macrophages (adherent cells) from tumor-bearing mice which exhibit very few cells, most of them with round m o r p h o l o g y (Fig. 3a). In contrast, in the m o n a layers from BN 52021 treated mice the number of cells is strikingly increased and most of the cells have a spread morphology with pseudopod-like elongated
Inhibition of Ehrlich Ascites Tumor
61
Fig. 3. Spreading of peritoneal cells from tumor bearing mice, 10 days after i.p. inoculation of 103 EAT cells in mice treated with BN 52021 (5 mg/kg, 2 x day, i.p.) (b) and mice treated with the vehicle (0.1 ml, 2 x day, i.p.) (a). Phase contrast microscopy, 300 × .
Inhibition of Ehrlich Ascites Tumor
63 DISCUSSION
7/
* I
"".
" ~ o5
~"
-~ ±
z
,
--O---II
~
~
Time (days)
Fig. 4. Spontaneous release of H2Oz (nmol/105 peritoneal cells), 1, 6 and 10 days after i.p. inoculation of 1 × 103 EAT cells, in mice treated with BN 52021 (5 mg/kg, 2 × a day, i.p.) (A) or treated with 0.1 ml of the vehicle (0) and from normal mice treated with BN 52021, same dose as above, (11). Each point represents the mean _+S.E.M. from 10 to 14 animals. *P<0.05. structures (Fig. 3b). Interestingly, normal mice treated with BN 52021 also exhibited high levels of macrophage spreading (Fig. 2). With lower doses of BN 52021, it was found that 2.5 mg/kg stimulated spreading in normal as well as in EAT-bearing animals (59.3%0 and 88.5 o70, respectively, at 10days), and 1 mg/kgstimulated spreading only in EAT-bearing mice (62.7%). For evaluation of H202 release, 1 × 105 peritoneal cells, from the different experimental groups, were analysed individually, according to the technique described under Experimental Procedures. It was found (Fig. 4) that peritoneal cells from normal mice and from EAT-bearing mice, did not release detectable amounts of H202. However, treatment with BN 52021, significantly increased H202 release from peritoneal cells obtained 6 days after tumor implantation (0.54 nmol/105 cells) as well as from those obtained from normal mice. Care should be taken in interpreting the data obtained with the samples collected 10 days after tumor implantation. At this time, the tumor cells account for 8 0 - 90°7o of the peritoneal cell population, and consequently the number of macrophages per sample is very low. Under the experimental conditions employed, it was not possible to work with isolated macrophages since any kind of manipulation would interfere with the assay of the spontaneous, adherence-induced release of H202.
The results presented here show that two PAF antagonists inhibited the growth of Ehrlich ascites tumor in viva. BN 52021 dose-dependently inhibited the tumor growth when injected at the site of tumor growth. It is difficult to understand why other routes of administration were less effective. The only possible explanation is that the antagonist is needed in high concentration at the site of tumor/host cell interaction. This effect of BN 52021 cannot be ascribed to a toxic or anti-mitotic effect, since addition of the antagonist to EAT cells in culture did not affect cell viability and proliferation. These experiments, however, do not exclude the possibility that the compound may interfere with the metabolism of the tumor cell, affecting the synthesis or release of substances which potentially protect the tumor in vivo. In fact, several reports in the literature suggest that tumor cells are able to produce substances which depress host defenses. Products generated by tumor cells were shown to inhibit macrophage spreading, adherence, migration and oxidative metabolism. These substances are either proteins (Szuro-Sudol & Nathan, 1982) or low molecular weight, lipid-like substances (Cantarrow, Cheung & Sundharadas, 1978). Because macrophages operate at the tumor-host interface they may critically influence tumor growth. In the present work we have shown that inhibition of EAT growth occurred concomitantly with activation of peritoneal macrophages as assessed by the spreading ability and release of H202. We observed that the presence of the tumor cells apparently did not activate the local macrophages. Although EAT induced spreading one day after implantation, this was not observed later on. Treatment of the mice with BN 52021 had a marked effect on macrophages, stimulating spreading and release of H202. After 10 days of tumor implantation it is difficult to be certain of the results of H202 release by the method employed, because the number of tumor cells is very high, at this time. Nevertheless, on the sixth day, when the number of cells in both groups (treated and non-treated with BN 52021) is similar, it was clearly observed that only macrophages from the treated group released H202. Although it seems that tumor inhibition is associated with macrophage activation, care should be taken in the interpretation of these results. When looking at individual data we observed that such a direct association is not always observed. The observation that treatment with the PAF antagonist also promoted activation of macrophages from normal mice allows two hypotheses: (i) that the
64
D. FECCHIOet al.
antagonist is a partial agonist; or (ii) that the agonist (PAF) is an endogenous modulator of macrophage function. Since the PAF antagonist usually has no agonistic effect on platelets and other tissues, the latter hypothesis seems more plausible. In the experiments where administration of BN 52021 was interrupted after 10 days of treatment, the tumor regained its capacity to grow, suggesting that the effect of the compound is reversible. This result also suggests that the antagonist needs to be present during the tumor growth to exert its inhibitory activity. Inhibition of EAT growth by PAF antagonists also indicates that PAF or PAF-like substances are involved in the host-tumor interaction, possibility favoring tumor growth. Since inhibition of EAT growth was parallel with macrophage activation, the possibility exists that macrophages are the targets for the action of PAF or PAF-like substances. Recently, several experimental data have appeared in the literature which indicate that PAF may have an important role in macrophage function. PAF regulates IL-I production, either stimulating (low doses) or inhibiting (high doses) the release of IL-1 from LPS-stimulated monocytes (Pignol, Henane, Mencia-Huerta, Rola-Pleszczynski & Braquet, 1987). It is also known that PAF can induce release of cyclo-oxygenase metabolites and that prostaglandins are important modulators of macrophage function as well as of the immune response (Goodwin & Webb, 1980). In addition, we have found high levels of PGE2 in the ascitic fluid from Ehrlich tumor (Fecchio et al., in press). Since release of prostaglandins is a consequence of activation of phospholipid metabolism, it is possible that PAF is also generated. PAF and prostaglandins have been shown to inhibit the expression of class II antigens in macrophage cellular membrane (Gebhardt, Bazan & Bazan, 1988; Snyder, Beller & Unaue, 1982), thus affecting macrophage/lymphocyte interaction with
consequences on the specific immune response. Indeed, we cannot exclude the possibility that PAF antagonists inhibit tumor growth by affecting the immune response. PAF has been described to suppress alloantigen recognition in MLC and generation of cytotoxic lymphocytes in vitro, the latter effect being reversed by BN 52021 (Gebhardt et al., 1988). This mediator has also been shown to inhibit lymphocyte proliferation and IL-2 production and to generate suppressor cells, either directly or by inducing the release of cyclo-oxygenase metabolites from monocytes (Rola-Pleszczynski, Pignol, Pouliot & Braquet, 1987a,b). The cell type responsible for the release of PAF or PAF-like substances in the Ehrlich Ascites Tumor remains to be determined. Macrophages are known to produce PAF (Mencia-Huerta & Benveniste, 1979) and unusually high levels of O-alkyl glycero phospholipids have been detected in several mammalian tumors including EAT (Snyder & Wood, 1968; Friedberg, Halpert & Barnwell, 1985). Overall, our data suggest that tumor-derived or effector cell-derived factor(s) inhibits macrophages and that this inhibition favors tumor growth. Since PAF antagonists inhibited tumor growth and stimulated macrophages, the possibility exists that the inhibitory factor is PAF or PAF-like. Nevertheless, as pointed out in the discussion, several points remain obscure and more work is needed before we can venture any hypothesis. Unravelling the cellular and molecular mechanisms that govern the inflammatory component of tumor development should provide a better understanding of the nature of the tumor-host interactions and eventually the possibility of modulating these interactions in favor of the host. Acknowledgements - - The authors wish to acknowledge
Funda~o de Amparo ~ Pesquisa do Estado de S~o Paulo (FAPESP) for financial support and Miss Grazia de Giacobbi for technical assistance.
REFERENCES
BRAQUET,P. & ROLA-PLESZCZYNSKI,M. (1987). Platelet activating factor and cellular immune responses. Immun. Today, 8, 345 - 352. CANTARROW,W. D., CHEUNG,H. T. • SUNDHARADAS,G. (1978). Modulation of spreading, adhesion and migration of peritoneal macrophages by a low molecular weight factor extracted from mouse tumors. J. Reticuloendothel. Soc., 24, 657 - 666. ECCLES,S. A. & ALEXANDER,P. (1974). Macrophages content of tumors in relation to metastatic spread and host immune reaction. Nature, Lond., 250, 667-669.
Inhibition of Ehrlich Ascites Tumor
65
EVANS, R. (1973). Macrophages and tumor bearing host. Br. J. Cancer, 28, 19-25. FAUVE, R. M., HEVIN, B., JACOB, H., GAIZZARD, J. A. & JACOB, F. (1974). Anti-inflammatory effects of murine malignant cells. Proc. natn. Acad. Sci. U.S.A., 10, 4052-4056. FECCHIO, D., SIRO~S,P., RUSSO, M. & JANCAR,S. Studies on the inflammatory response induced by Ehrlich tumor in mice peritoneal cavity. Inflammation. In press. FRIEDBERC, S. J., HALPERT, M. & BARNWELL, G. M. (1985). The rate of formation of surface membrane ether lipids in Ehrlich ascites tumor cells: Kinetic consideration. Arch. Biochern. Biophys., 243, 504-514. GEBHARDT, B. M., BAZAN, H. E. P. & BAZAN,N. G. (1988). Ginkgolide BN 52021 blocks PAF-mediated suppression of cellular immunity. In The Ginkgolides - - Chemistry, Biology, Pharmacology and Clinical Perspectives (ed. Braquet, P.) pp. 703-729. J. R. Prous, Barcelona. GOODWIN, J . S . & WEBB, D . R . (1980). Regulation of the immune response by prostaglandins. Clin. Imrnun. [mrnunopath., 15, 106-122. MENCIA-HUERTA, J. M. & BENVENISTE,J. (1979). Platelet-activating factor and macrophages. I. Evidence for the release from rat and mouse peritoneal macrophages and not from mastocytes. Eur. J. Immun., 9, 409- 415. PICK, E. & KEISARI,Y. (1980). A simple colorimetric method for measurement of hydrogen peroxide produced by cells in culture. J. Imrnunol. Methods, 38, 161- 172. PIrCE, M. C. & SNYDERMAN,R. (1976). Depression of macrophage function by a factor produced by mouse neoplasms: A mechanism of abrogation of immune surveillance. J. lmrnun., 117, 1243- 1249. PIGNOL, B., HENANE, S., MENCIA-HUERTA,J. M., ROLA-PLESZCZYNSKI,M. & BRAQUET, P. (1987). Effect of PAF-acether and its specific antagonist, BN 52021, on interleukin l (IL-1) synthesis and release by rat monocytes. Prostaglandins, 33, 391 - 396. RABINOVITCH, M., MANEJIAS, R. E., RUSSO, M. & ABBEY, E. E. (1977). Increased spreading of macrophages from mice treated with interferon inducers. Cell Irnrnun., 29, 86-95. ROLA-PLESZCZYNSKI,M. (1985). Immunoregulation by leukotrienes and other lipoxygenase metabolites. Irnrnun. Today, 6, 302 - 307. ROLA-PEESZCZYNSKI,M., PIGNOL, B., POULIOT, C. & BRAQUET,P. (1987a). Inhibition of human lymphocyte proliferation and interleukin 2 production by platelet-activating factor (PAF-Acether). Reversal by a specific antagonist, BN 52021. Biochern. Biophys. Res. Cornrnun., 142, 745- 760. ROEA-PLESZCZYNSKI, M., POULIOT, C., PIGNOL, B. & BRAQUET, P. (1987b). Platelet-activating factor induces human suppressor cell activity. Fedn Proc. Fedn Am. Socs exp. Biol., 46, 444-450. Russo, M., TEIXEIRA, H. L., MARCONDES, M. C. G, & BARBUTO, J. A. M. Superoxide-indpendent hydrogen peroxide release by activatd macrophages. Brazilian J. ivied. Biol. Res. In press. RUSSEL, S. W., GILLESPIE, G. Y., HANSEN, C. B. & COCHRANE, C. G. (1976). Inflammatory cells in solid murine neoplasms. II-cell types found as the cause of Moleney sarcoma regression or progression. Int. J. Cancer, 18, 331-338. SNVDER, F. & WOOD, R. (1968). The occurrence and metabolism of alkyl and alk-1-enyl ethers of glycerol in transplantable rat and mouse tumors. Cancer. Res., 28, 972-978. SNVDER, D. S., BELLER,D. I. & UNAUE,E. R. (1982). Prostaglandins modulate macrophage Ia expression. Nature, Lond., 299, 163 - 165. SZt3Ro-SUDOL, A. & NATHAr~, C. F. (1982). Suppression of macrophage oxidative metabolism by products of malignant and non-malignant cells. J. exp. Med., 156, 945- 961. ZAR, J. H. (1984). Biostatistical Analysis. pp. 162-205. Prentice-Hall Inc., New Jersey.