Int. J. lmmunopharmac., Vol. 8, No. 2, pp. 2 2 1 - 226, 1986. Printed in Great Britain.
0 1 9 2 - 0 5 6 1 / 8 6 $3.00+ .00 © 1986 International Society for lmmunopharmacology.
REQUIREMENT FOR PROSTAGLANDIN El (PGE0 FOR THE SECRETION OF SUPPRESSOR CELL INDUCER FACTORS BY SPLEEN CELLS OF T U M O R - B E A R I N G MICE DEVRAJ J. PILLAY and BARBARA L. POPE* Department of Pharmacology, Dalhousie University, Sir Charles Tupper Medical Building, Halifax, Nova Scotia, Canada B3H 4H7
(Received 19 April 1985 and in final form 16 August 1985)
Abstract - - We have demonstrated previously that spleen cells from mice bearing M-1 fibrosarcomas release low molecular weight factors capable of activating suppressor cells from unprimed normal spleen ceils. Using Marbrook vessels to separate inducer and precursor populations, we found in the previous paper that cyclooxygenase inhibitors blocked the activation of suppressor cells and that this activation could be restored by exogenous PGEI. In this paper we have examined the site of action of the prostaglandins in the activation of suppressor cells. To do so, we tested cell-free supernatants from cultured spleen cells of tumor-bearing mice (inducer cells) for the ability to activate suppressor cells from unprimed normal spleen cells (precursor cells). Supernatants from acetyl salicyclic acid (ASA) treated inducer cells did not activate suppressor cells and exogenous PGEI could not restore the activity of this supernatant. In contrast, if inducer cells were treated with ASA and then incubated with PGE1, the supernatant was capable of activating suppressor cells. No role for prostaglandins at the level of the precursor cells or the effector suppressor cells was seen. These data suggest that the inducer cells in tumor-bearer spleens require prostaglandins for the release of an inducer factor but that prostaglandins are not required for the action of this factor on the precursor cells or for the effector function of the activated suppressor cells.
INTRODUCTION
Prostaglandins o f the E series (PGE) have been implicated in suppressor cell circuits at several different levels. First, both PGE1 and PGE2 have been shown to activate suppressor cells (Fischer, Durandy & Griscelli, 1981; Fulton & Levy, 1980; Webb & Jamieson, 1976). Pharmacological concentrations of PGE~ have been demonstrated to act on glass adherent T lymphocytes which subsequently secrete low molecular peptides called Prostaglandin Induced T Suppressor Cell Factors (PITS). These factors suppress the proliferation and function of both T and B cells (Rogers, Nowowiejski & Webb, 1980). Secondly, smaller physiological quantities of PGE2 appear to be necessary for the
activation of suppressor cells by substances such as Coucanavalin A (Con A). In the latter case, it has been postulated that PGE2 is necessary to sensitize suppressor T cells to the activating signal which is released by Con A activated T cells (Fischer et aL, 1981). Thirdly, P G E may be required for the effector function of suppressor factors. This has been found in the case of antigen specific suppressor cells which suppress the effector arm of the delayed hypersensitivity response. It has been shown that PGE2 can reconstitute inactive suppressor factors collected from indomethacin treated suppressor cells (Kato & Askenase, 1984). Lastly, the activity of many suppressor cells activated in vivo have been reversible by cyclooxygenase inhibitors (Goodwin & Webb, 1980). These suppressor cells may act by
*Author to whom correspondence should be addressed. Abbreviations used in this paper: ASA, acetyl salicylic acid; PG, prostaglandin; PGE, prostaglandins of the E series; SD, standard deviation; FBS, fetal bovine serum; PFC, placlue forming cells; SRBC, sheep red blood cells; AMP, adenosine monophosphate; PBS, phosphate buffered saline; Con A, concanavalin A. 221
222
D. J. PILLAYand B. L. POPE
releasing high, directly immunosuppressive quantities of PGE. In this case, in addition to the possibility that the PGE acts indirectly by activating suppressor ceils as described above, suppression may be due to the binding of PGE to lymphoid receptors with a subsequent rise in intracellular cyclic AMP and an inhibition of proliferation (Goodwin, Bromberg & Messner, 1981). Both PGE, and PGE2 have been demonstrated to be directly immunosuppressive. Our interest is in the role of PGE in the activation of antigen non-specific suppressor cells seen during the advanced growth of a murine fibrosarcoma. We have found that the emergence of a suppressor T cell population in tumor-bearing mice can be prevented by long term treatment of the mice with indomethacin (Pope, 1985a). In contrast, the activated suppressor cells are not affected by in vivo or in vitro treatment with cyclooxygenase inhibitors. In the accompanying paper we demonstrated that the in vitro activation of suppressor cells by dialysable factors secreted by spleen cells from tumor-bearing mice was blocked by cyclooxygenase inhibitors and that this blockade was reversed by exogenous PGEI. Since PGE~ alone did not activate suppressor cells in control cultures and, since tumor-bearer spleen cells did not secrete abnormally high levels of PGE, we hypothesized that PGE in some way modulated the activity of an inducer factor released by tumorbearer spleen cells. In this paper we provide evidence that the PGE is required for the release of an inducer factor from the tumor-bearer spleen ceils but not for the effect of the factor on the precursor cells or for the effector fimction of the activated suppressor cells.
EXPERIMENTAL
PROCEDURES
Mice and tumor line Male DBA/2J mice (Jackson Laboratories, Bar Harbour, ME) aged 2 to 4 months were used in all experiments. Tumor-bearing mice were injected subcutaneously with 5 x 104 M-1 fibrosarcoma cells from 21 to 28 days before the spleens were removed as a source of inducer cells. The M-1 tumor cell line has been described in detail in the accompanying paper. Activation o f suppressor cells Spleen cells suspensions were prepared by teasing the spleen, filtering the cells through gauze, centrifuging for 10 rain at 200 x g, and resuspending
in culture medium (RPMI 1 6 4 0 medium supplemented with 100 IU/ml penicillin, 100/ag/ml streptomycin, 10°70 fetal bovine serum (GIBCO Canada), 5 x 10-s M 2-mercaptoethanol, and 2 mM glutamine. The method for activating suppressor cells in Marbrook vessels has been described in detail in the accompanying paper. Briefly, 5 x 106 normal spleen cells were cultured in the inner chamber of Marbrook vessels in 1 ml culture medium. The outer chamber contained 5 x 107 spleen cells from tumorbearing mice in 10 ml medium. The two chambers were separated by regular dialysis tubing, thus allowing the 2 cell populations to communicate only by dialysable factors. The chambers were incubated for 18 h and the contents of the inner chamber were tested in a mixing assay for suppressor cell activity. A second method of suppressor cell activation was used in this paper. Spleen cells from tumor-bearing mice (or normal mice as a control) were cultured in 24 well trays at 5 × 106 cells/ml at 37°C in 5070 COs for 24 h. Cell free supernatants were prepared by centrifuging the ceils at 200 × g for 10 min. The supernatants were tested for inducer factors by incubating 5 x 106 normal spleen cells in 1 mi supernatant at 37°C in 5% CO2 for 6 h, washing the cells three times with phosphate buffered saline (PBS), and testing for suppressor cell activity in the usual mixing assay. Mixing assay f o r suppressor cell activity The mixing assay has been described in detail in the accompanying paper. Briefly, 1 × 106 responder cells from normal mice were cultured with 5 × 105 sheep red blood ccells (SRBC) and 5 × l0 s of the cells to be tested for suppressor cell activity. In all experiments control cultures contained spleen cells from age and sex matched normal mice rather than spleen cells from tumor-bearing mice. The cultures were incubated at 37°C in 5% CO2 for 5 days and the contents of 3 wells were pooled and assayed for the number of cells secreting IgM antibodies to SRBC. Data are expressed as the number of plaque Forming Ceils (PFC) per culture _+ the standard deviation (SD) of triplicate determinations. Materials The ASA and PGE1 were obtained from the Sigma Chemical Co., St. Louis, MO. ASA was dissolved in culture medium at 5 mg/ml. PGE, was dissolved in 9507o ethanol at 5 mg/ml and aliquots were stored at - 20°C. Further dilutions were made in medium just prior to addition of the drugs to the assay. In all experiments involving PGE, a vehicle control
PGE, and Secretion of Suppressor Cell Inducer Factor ASA TREATMENT IC --oc None
0
I
PFCKULTURE 1000 500
I
1500
None
10’M
cells alone
Cell-free supernatants from tumor-bearer cells activate suppressor cells
1 O-dM IO-‘M None
contrast, the treatment of the precursor did not block activation.
I
I
None
lO”M
223
-* 0 Control
q Activatedsuppressor
cells
Fig. 1. Effect of pretreatment of inducer or precursor cells on the activation of suppressor cells in Marbrook vessels. Results are expressed as PFC -+ S.D. of normal responder spleen cells cultured for 5 days with SRBC and control (clear bars) or activated suppressor cells (hatched bars). Control cells were normal spleen cells (precursor cells) from the inner chamber of Marbrook vessels which had been exposed for 18 h to dialysable factors from normal spleen cells (inducer cell control). Activated suppressor cells were normal spleen cells which had been exposed to dialysable factors from tumor-bearer spleen cells (inducer cells). Precursor or inducer cells were treated with ASA and washed before they were cultured in the Marbrook vessels. *Significant suppression (P < 0.05).
containing ethanol was included. Cells were pretreated with 1 x 10m4M ASA for 1 h, washed 3 times with PBS, and resuspended in culture medium.
Data presentation All experiments have been repeated a minimum of 3 times, but data are from representative experiments. Statistical differences between groups were determined by Friedman’s non-parametric two factor analysis of variance by ranks.
RESULTS
ASA pretreatment of the inducer cells prevents activation of the suppressor cells in Marbrook vessels In previous experiments we found that the addition of ASA to Marbrook vessels at the beginning of the incubation period prevented the activation of suppressor cells. In order to determine whether the ASA was acting at the level of the inducer or precursor cells, we pretreated the populations with ASA before culturing them. As can be seen in Fig. 1, pretreatment of either the inducer cells alone or of both the inducer and precursor cells prevented the activation of suppressor cells. In
spleen
In order to better address the role of prostaglandins in the activation of suppressor cells, we developed a protocol in which the inducer factors could be collected and tested for the ability to activate suppressor cells under different conditions. In these experiments, normal precursor spleen cells were cultured for 6 h in supernatants collected from 24 h cultures of tumor-bearer spleen cells. Control cultures contained supernatants collected from normal spleen cells under identical conditions. As seen in Fig. 2, suppressor cells were activated under these conditions.
ASA pretreatment of the inducer cells prevents the release of an inducing factor The experiments in Fig. 1 suggested that the ASA was acting at the level of the inducer cells since the treatment of the precursor cells alone did not block the activation of suppressor cells. However, it was not possible from these data to determine the possible site of action of the prostaglandins since prostaglandins secreted by the inducer cells could reach the precursor cells. To determine whether ASA could be acting by preventing the release of inducer factors, we treated tumor-bearer spleen cells with ASA and then collected the supernatants. As can be seen in Fig. 3, no activation of suppressor cells was seen under these conditions, despite the fact that the precursor cells were capable of synthesizing prostaglandins. In addition, the supplementation of the factor collected from ASA treated inducer cells with 1 x lo-’ M PGE, did not restore the activation
SPLEEN CELL SUPERNATANT
PFWULTURE 0
Normal
B
Tumor-Bearer
m*
500
1000
Fig. 2. Activation of suppressor cells by supernatants from spleen cells of tumor-bearing mice. PFC ? S.D. of normal responder cells cultured for 5 days with SRBC and control (clear bars) or activated suppressor cells (hatched bars). Control cells were normal precursor spleen cells which had been cultured at 5 x IO”cells/ml in 24 h supernatants from normal spleen cells. Activated suppressor cells were normal spleen cells cultured at 5 x IO6cell/ml in 24 h supernatants from tumor-bearer spleen cells. *Significant suppression (P < 0.05).
224
D . J . PILLAY and B. L. POPE
ASA TREATMENT PGE~ Inducer Precursor in Cells Cells Supernatant 0 I
ASA 0
PFC/CULTURE 500
1000
1500
I
I
I
None
None
None
k\\\\\\\~ ~
1 O-4M
None
None
kk~kkk\\\\\\\\\\\\\\\k,~- ~ k ~
10-'M
None
10-SM
~\\\\\\\\\\\\\\\\~_.
None
10-~M
None
kkL~kkk~\~~
l O-,M
I O-'M
I O-'M
I
> -' F
o f suppressor cells. Similar results were obtained when A S A treated precursor cells were incubated with inducer factors. The supernatants from untreated inducer cells activated suppressor cells whereas the supernatants from A S A treated inducer cells did not (Fig. 3). The addition o f P G E , to the latter g r o u p had no effect.
PGE, restored the ability o f A S A treated inducer cells to release inducer factors Previous experiments showed that the s u p e r n a t a n t f r o m A S A treated inducer cells did not activate
PFC/CULTURE 500 1000
1500
[\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\~
10-4M
None
10-"M
10-SM k \ \ \ \ \ \ \ \ \ \ N \ x t~ .
None
I
~,,:
I
]-~
IO-'M
Fig. 3. Effect of ASA pretreatment of inducer or precursor cells on the activation of suppressor cells by inducer cell supernatants. PFC __+S.D. of responder cells cultured with SRBC and control (clear bars) or activated suppressor cells (hatched bars). Suppressor cells were normal precursor spleen ceils which had been cultured for 6 h with supernatants from normal (control) or tumor-bearer (activated) inducer cells. The precursor or inducer cells were either untreated or pretreated with 1 x 10-4 M ASA before they were cultured. PGE, was added to the inducer cell supernatants at the time of addition to the precursor cells. *Significant suppression (P < 0.05).
TREATMENT OF INDUCER CELLS ASA PGE, 0
PFC/CULTURE 500 1000
I
Fig. 4. Effect of PGEj on ASA treated inducer cells. PFC _+ S.D. of responder cells cultured with SRBC and control (clear bars) or activated suppressor cells (hatched bars). Suppressor cells were activated by a 6 h culture in supernatants from normal (control) or tumor-bearer (activated) spleen cells. The inducer cells were pretreated with 1 × 10-" M ASA, washed, and cultured for 24 h with or without 1 × 10~ PGE, before the supernatants were collected. *Significant suppression (P < 0.05).
Fig. 5. Effect of ASA on the effector function of the activated suppressor cells. PFC _+ S.D, of responder cells cultured for 5 days with SRBC and control (clear bars) or activated suppressor cells (hatched bars). ASA was added to the cultures at the beginning of the assay at 1 × 10-4 M. *Significant suppression (P < 0.05). suppressor cells suggesting that prostaglandin synthesis was required for the release o f an inducer factor. To test this hypothesis, we treated inducer cells with A S A and incubated them with 1 x 10 -5 M P G E , for 24 h b e f o r e collecting the supernatant. As seen in Fig. 4, this treatment did restore the ability o f the inducer cells to secrete an inducer factor.
A S A did not block the effector function o f the activated suppressor cells In order to test whether prostaglandin synthesis was required for the activity o f the activated suppressor cells, we added 1 × 10-4 M A S A to the mixing assay. A S A at this concentration did not inhibit the P F C response o f the responder cells and also did not block the suppressor cell activity (Fig. 5).
DISCUSSION We d e m o n s t r a t e d in the previous paper in this series that P G E is required for the activation o f suppressor cells by inducer factors secreted by spleen cells f r o m t u m o r - b e a r i n g mice. The data in this paper best support the hypothesis that P G E is required for the release o f a factor f r o m the inducer cells and that P G E is not required to reconstitute the factor, to activate the precursor cells, or for the effector function o f the activated suppressor cells. We started with the hypothesis that P G E is needed in a m o d u l a t o r y capacity during the process o f activation o f suppressor cells. It was possible the P G E acted: (a) at the level o f the inducer cells and was necessary for the release o f the factor, (b) as an active p o r t i o n o f the inducer factor, (c) at the level o f the precursor cell, or (d) at the effector level o f the activated suppressor cells. Our first a p p r o a c h was to treat the inducer a n d precursor cells with A S A before placing them in M a r b r o o k vessels for the activation
225
PGE~ and Secretion of Suppressor Cell Inducer Factor of suppressor cells. In these experiments it was clear that treatment of inducer, but not precursor cells, prevented the activation of suppressor cells. These data suggested that prostaglandin synthesis was most important at the level of the inducer cells. The problem with this approach was that prostaglandins secreted by the untreated inducer cells could reach the treated precursor cells. In order to more clearly establish the stage at which PGE is required for the activation of suppressor cells we altered our methodology to a three stage procedure. In the first step, we collected supernatants from inducer or control cells. In step two, supernatants were incubated with precursor cells for the activation of suppressor cells. Thirdly, the activated cells were tested for suppressor cell activity in a mixing assay. Using this approach it was possible to inhibit prostaglandin synthesis or add exogenous PGE at different times during the activation process. Our first observation was that the supernatant from ASA treated inducer cells was incapable of activating suppressor cells. This factor could not be reconstituted with exogenous PGE, and had no activity when incubated with either ASA treated or untreated precursor cells. In contrast it was possible to recover an active inducer factor from ASA treated inducer cells which had been cultured with PGE1. These data suggested that PGE was necessary for the release of an active inducer factor from the tumorbearer spleen cells and that PGE could not reconstitute an inactive factor released from inducer cells in the absence of PGE. There was no evidence that prostaglandin synthesis was required at the level of the precursor cells since suppressor cells were activated equally well by active inducer factors when the precursor cells were either ASA treated or untreated. However, it is not possible to totally exclude a role for PGE at this level, since inducer supernatants contained low levels (approx. 1 x 10-9 M) of PGE and it is possible that this concentration was necessary, but sufficient for the factor to interact with and activate the precursor ceils. Similarly, there was no evidence that prostaglandins were required for the effector function of the activated suppressor cells since the addition of ASA to the mixing assay did not block suppressor cell activity. We believe that this is the first report that prostaglandins are required for the release of a suppressor cell inducer factor from spleen cells activated during tumor growth. It is well documented that both PGE, and PGE2 are immunosuppressive (Goodwin & Webb, 1980) and
that suppressor cells found in a variety of circumstances, including malignancy, act via secreted prostaglandins (Goodwin, 1980). Postulated mechanisms for this suppression by PGE include: (a) a direct binding of PGE to receptors on lymphoid cells with a resultant rise in cyclic A MP and consequently an inhibition of proliferation (Goodwin et al., 1981), and (b) binding of PGE to suppressor cells which, upon activation, secrete suppressive peptides (Rogers et al., 1980). We would now like to add a third possible mechanism, that PGE stimulates inducer cells to release factors which activate suppressor cells. One of the issues to be resolved is the nature of the inducer cell population. It is clear that under the experimental conditions employed to activate the suppressor cells, significant suppression was seen only with tumor-bearer spleen cells and not with normal spleen cells. If there is a distinct population of inducer cells which binds PGE and releases inducer factors, then tumor-bearing mice may have increased numbers of these cells, may have inducer cells which express more PGE receptors, or may have inducer ceils which have a greater sensitivity to PGE. Evidence supporting these mechanisms at the effector cell level has been found in other models of suppressor cells. Kaszubowski & Goodwin (1982) have demonstrated that T cells with receptors for PGE express receptors for the Fc portion of IgG and act as suppressor cells following interaction with PGE. Delfraissy, Galanaud, Wallon, Balovoine & Dormont (1982) have shown that peripheral mononuclear lymphocytes from aged individuals have an increased sensitivity to PGE and that this correlates with the presence of radiosensitive suppressor T cells. It is not clear if the suppression is
Suppressor Inducer Factor
~
Thy-1
. ~k~upp~Ue~lSor).Lyt.1 ~_.~" Lyt-2
.SuppressOrFactor
Fig. 6. Theoretical model proposing a role for PGE in the activation of suppressor cells by soluble factors from tumor-bearing spleen cells.
226
D. J. PILLAY and B. L. POPE
due to increased n u m b e r s o f these cells or due to increased sensitivity o f the cells to PGE2. Fischer et al. (1981) have s h o w n that PGE2 is necessary for the activation o f suppressor cells by C o n A a n d have postulated that the C o n A sensitizes the suppressor cells to P G E . In all o f these examples, it is assumed t h a t the P G E acts at the level o f the suppressor effector cell r a t h e r t h a n at the level of the suppressor inducer cell. Kato a n d A s k e n a s e (1984) have s h o w n that PGE2 is necessary for the reconstitution o f a suppressor factor p r o d u c e d by antigen specific suppressor ceils treated with i n d o m e t h a c i n . They have postulated t h a t the PGE2 is a n active c o m p o n e n t o f the factor
a n d delivers the ultimate suppressive signal to the cell binding the suppressor factor. We have excluded this as the only role for P G E in our inducer factor since P G E , did not reconstitute the factor f r o m A S A treated inducer cells. The most p r o b a b l e e x p l a n a t i o n for o u r results at this time is t h a t P G E is required for the release o f a n inducer factor f r o m cells which accumulate or b e c o m e m o r e sensitive to P G E d u r i n g t u m o r growth. This model is presented in d i a g r a m m a t i c form in Fig. 6. Acknowledgements--Supported by grants from the National Cancer Institute of Canada and the Medical Research Council of Canada.
REFERENCES
DELFRAISSY,J. F., GALANAUD, P., WALLON,C., BALAVOINE,J. F. & DORMONT,J. (1982). Abolished & vitro antibody response in elderly: exclusive involvement of prostaglandin-induced T suppressor cells. Clin. lmmunol. lmmunopathol., 24, 3 7 7 - 385. FISCHER,A., DURANDY,A. & GRISCELLI,C. (1981). Role of prostaglandin E~ in the induction of non-specific T lymphocyte suppressor activity. J. lmmunol., 126, 1452- 1455. FULTON,A. M. & LEVY, J. G. (1981). The induction of nonspecific T suppressor lymphocytes by prostaglandin E,. Cell. lmmunol., 59, 5 4 - 6 0 . GOODWIN, J. S. (1980). Prostaglandin synthetase inhibitors as immunoadjuvants in the treatment of cancer. J. Immunopharmacol., 2, 397- 424. GOODWIN,J. S., BROMBERG,S. & MESSNER,R. P. (1981). Studies on the cyclic AMP response to prostaglandins in human lymphocytes. Cell. Immunol., 60, 298-307. GOODWIN, J. S. & WEBB, D. R. (1980). Regulation of the immune response by prostaglandin. Clin. lmmunol. Immunopathol., 15, 106-122. KASZUBOWSKI,P. A. & GOODW1N,J. S. (1982). Monocyte-produced prostaglandin induces Fcy receptor expression on human T cells. Cell. Immunol., 68, 3 4 3 - 348. KATO,K. & ASKENASE,P. W. (1984). Reconstitution of an inactive antigen-specific T cell suppressor factor by incubation of the factor with prostaglandins. J. Irnmunol., 133, 2025- 2031. POPE, B. L. (1985a). The effect of indomethacin on the activation and effector function of suppressor cells from tumorbearing mice. Cancer Immunol. Immunother., 19, 101 - 108. POPE, B. L. (1985b). Activation of suppressor T cells by low molecular weight factors secreted by spleen cells from tumorbearing mice. Cell. Immunol., 93, 364-374. ROGERS, T. J., NOWOWlEJSKI, I. & WEBB, D. R. (1980). Partial characterization of a prostaglandin-induced suppressor factor. Cell. Immunol., 50, 8 2 - 9 3 . WEBB, D. R. & JAMIESON, A. T. (1976). Control of mitogen-induced transformation: characterization of a splenic suppressor cell and its mode of action. Cell. Immunol., 24, 4 5 - 57.