Human health and the environment in eastern and central Europe.

More than 40 environmental health scien? tists, from 15 countries of Eastern, Central, and Western Europe and the United States, gathered from 12-15 April 1993 at Charles University in Prague, the Czech Republic, to participate in the con? ference, "Human Health and the Envir? onment in Eastern and Central Europe." This conference, convened by the Conte Institute for Environmental Health, represented a unique nongovernmental effort by international scientists to develop a regional approach to research on the health effects of environmental contamination in

The spleens of high-dose BCG-treated mice also contained suppressor cells that were capable of inhibiting the in vitro reactivity of normal T cells to PHA. Like the cytostatic effect, this suppressor activity was not detected in low-dose mice, and the cells responsible had the properties of macrophages; the effect was lost after the removal of adherent cells by sequential exposure to plastic and colloidal iron, but was conserved after treatment with anti-Thy 1.2 + C'. T-cell-deprived animals, such as "B" or nude mice, also developed suppressor-cell activity when treated with systemic high-dose BCG. Close parallels became evident between the in vivo anti-tumour activity of BCG, the in vitro cytostatic effect, and the suppressor-cell activity. We here discuss the possible role of suppressor cells in the mechanism of action of this agent.
A MAJOR ADVANCE in the field of immunology during the past few years has been the concept that immune reactions are controlled by negative feed-back mechanisms through suppressor cells (Gershon & Kondo, 1971). In many immunological model systems suppressor cells of different types, particularly suppressor T cells (Ha et al., 1974), but also B cells (Zembala et al., 1976) and macrophages (Nelson, 1976), have been implicated. There have been a number of reports that suppressor cells may also play a role in the immunosuppression found in some experimental tumour models (Greene et al., 1977). In this respect, various bacterial vaccines such as BCG have been demonstrated to enhance nonspecifically a number of immune responses (Miller et al., 1973) and are reported to produce a clinical improvement in some cases of leukaemia (Mathe et al., 1974) and melanoma (Morton et al., 1970). Also in various animal models, like the one described in Correspondence to: Dr Simone Orbach-Arbouys, IrLstitut de Canc6rologie et d1'Immunogn6tiqlue, H6pital Paul-Brousse, 14 av. P.V. Couturier, 94804 Villejuif C6dex, France.
13CG-INDUCED SUPPRESSOR ACTIVITY8 this report, BCGr treatment can exert a marked curative effect (Hanna et al., 1972). Studies in our laboratories have demonstrated that these agents can also induce nonspecific cell-mediated suppression of a variety of responses such as mitogen stimulation (Orbach-Arbouys & Poupon, 1978) or GvH and mixed lymphocyte reactions (Orbach-Arbouys & Castes, 1980).
In an attempt to elucidate the possible role of suppressor cells in the anti-tumour activity of BCG, we performed a series of experiments to correlate this effect with the in vitro cytostatic activity of various spleen-cell populations from BCG-treated mice. We demonstrated that systemic high-dose BCG exerted an anti-tumour effect in vivo, and induced the appearance of a cell population that was both cytostatic and suppressive for mitogen response. These cells weire induced, and expressed their activity, in the absence of T cells, and had the properties of macrophages.

Muice
6-8-wk-eek-old specific-pathogen-free C57-BL/6 mice were obtained from the broeding centre of the Centre National de la Recherche Scientifique, Orleans, France. C3H/He mice wvere provided by the Institut de Recherches Scientifiques sur le Cancer, Villejuif, France. Nude (nu/nu) mice were maintained in deliberately outbred colonies under barrierprotected conditions by Dr J. C. Salomon. IRSC, Villejuif, France. "B" mice were prepared by irradiation with 6 Gy, 1 week after thymectomy, and ith 3 Gy a week later (A. J. S. Davies, personal communication). The mice ere not reconstituted with marrow, as this can contain T cells. Such mice mounted no detectable plaque-forming cell (PFC) responses to injections of sheep erythrocytes, but did so against DNP-flagellin.

Tunmours
McC3-1 and McB6-1 were fibrosarcomas induced in C3H/He and C57BL/6 mice respectively by the i.m. injection of 1 mg 3-methylcholanthrene, and maintained by serial isogenic transplantation and freezing of various passages. Tumour grafts Awere performed ventrally by the s.c. introduction of small pieces (, 1 mm3) of non-necrotic tissue via a trocar needle. The day of grafting wvas taken as Day 0 in all experiments. BCG "Immuno BCG", which retains 950o viability after 3 months storage at 4°C, wvas kindly provided by the Pasteur Institute (Paris, France) and injected as described in the results. High doses (3 mg) i.v. caused marked splenomegaly and hepatomegaly, viz., spleen weights of 180 and 450 mg on Days 7 and 14 respectively, and liver wNeights of 1-6 and 2-6 g on these days (controls: 95 mg and 1P3 g respectively).

Cell preparations
Spleen, cells. Spleens from groups of 4 mice wvere aseptically removed, placed in Medium 199 (Eurobio, France) and gently squeezed between 2 sterile glass slides. The cell suspensions thus obtained wvere then filtered through gauze and washed wvith cold sterile Medium 199. For subsequent tests the suspensions wvere prepared in RPMI 1640 (Eurobio, France) supplemented wvith 2mM glutamine (Gibco, Grand Island, New-York) and containing 500 fresh human AB serum (decomplemented by heating at 56°C for 30 min), 100 i.u./ml penicillin and 100 ,Lg/ml streptomycin. When irradiated, the cells received 15 Gy delivered by a caesium bomb (Gravaton Industries, Gosport, Hampshire).
Enriched T cells. T cells wvere isolated with nylon w-ool columns using the technique of Julius et al. (1973). Anti-Thy 1.2 serumn (Lespinats & Poupon, 1977) plus guinea-pig complement killed 94%o of the normal Tenriched suspension. This population is referred to throughout this report as nylonpurified T cells, or simply spleen T cells. Such preparations obtained from BCG-treated mice contained only 7500 of cells killed by this treatment.
Adherent spleen cells. 108 spleen cells in 10 ml of RPMI Awith 50 inactivated foetal calf serum (Eurobio. France) wvere incubated horizontally in a 250ml plastic bottle (Falcon Plastics, California, U.S.A.); then after 60 min the supernatant was discarded and the cells adhering to the plastic were rinsed twrice with PBS, incubated 3-10 min with 0 M. CASTES, N. R. LYNCH, G. LESPINATS AND S. ORBACH-AR13O0YS 5 ml of Ca-and Mg free Earle's medium (Gibco) containing 0-02% EDTA. The cells were harvested, washed and adjusted to a concentration of 5 x 106/ml. Depletion of adherent spleen cells usi,ng carbon yl-irotI particles. 2 x 107 spleen cells in 20 ml of nutrient medium supplemented with 10% FCS were incubated at 37°C for 30 rnin w%Aith 50 mg of carbonyl-iron particles (particles E; GAF, Louvres, France) in a 250ml plastic bottle (Falcon Plastics). The cells adhering to the particles were magnetically removed by the procedure of Goldstein & Blomgren (1973). Treatment ith anti-Thy 1.2+C' killed 3%0 of these cells.
Depletion? of Thy 1.2 cells. Anti-Thy 1.2 serum wNas prepared as described by Reif & Allen (1964). This antiserum killed 95%o of normal thymocytes hen incubated for 30 muin at a dilution of 1:27 in the presence of guinea-pig complement. Lyophilized guineapig serum w% as obtained from Institut Pasteur Production, Paris, France. and was absorbed wN-ith agar as described by Cohen & Schlesinger (1970).
In vitro Culture technriques Cytostatic test. Continuously cultivated McC3-1 and McB6-1 tumour cells, growing exponentially in RPMI 1640 supplemented with 500 inactivated foetal calf serum, were harvested by trypsinization and used as targets in cytostatic assays.
The Imethod used was as described by Lespinats & Poupon (1977) with minor modifications. The tumour cells were washed and adjusted to 2x 104 cells/ml in RPMI 1640 plus 2% inactivated foetal calf serum, 2mM glutamine and antibiotics. 100,ul aliquots of the lymphocyte suspensions were added to 2500 tumour cells at ratios of 12-5:1, 25:1 and 50:1, in a total volume of 200 jA in microtitre plates (Falcon,3040 Mitogen responsiveness.-To test the responsiveness of spleen-cell populations to mitogens. 2-5 or 5 x 105 cells were placed in wells of Falcon 3040 microplates in a volume of 250 ,ul. The mitogen dose used was that found to be optimal for stimulation of normal spleen cells: 04 ug per w%ell of PHA HA16 (Wellcome Research Laboratories, Beckenham, England).
To demonstrate the suppressive properties of cells from BCG-treated mice, 5 x 105 cells from normal or BCG-treated animals were cultivated with 2-5 x 105 normal T cells and the same amount of PHA was added as previously. All cultures were performed in triplicate and incubated for 48 h in an atmosphere of 5% CO2 in air. Five h before the end of the culture, 1 ,uCi of [3H]dT w%ias added to each well. The cells were harvested on glassfibre filters, as described above, and the results expressed as mean ct/min of triplicate samples.
The suppressive effect of one population of spleen cells (B) against another (A) was calculated by the formula: ct/min (A + B) stimulatedct/min (A + B) unstimulated et/mmn (A) stimulatedct/min (A) unstimulated Preliminary studies demonstrated that the results of tests performed in this way were inot influenced by factors such as cell crowding or nutritional alterations, and the optimal PHA concentration was not changed, even after BCG treatment.
Statistical methods.-The statistical significance of results obtained in survival experiments, cytostatic assays and mitogen stimulation were analysed by the t test. The exact method of calculating significance by the x2 test, using Yates' correction for small sample sizes, was also used in the analysis of survival data.

RESULTS
Anti-tuamour activity of BCG as measured in vitro BCG was injected at high (3 mg) or lowr (0-1 or 0 5 mg) i.v. doses into mice bearing the transplantable MeC3-1 fibrosarcoma. The BCG was injected either on the day of the graft (Day 0), 14 days 830) "B" mice BCG "B" mice 4 Normal before (Day -14) or 7 days after (Day 7). In Table I, presenting the results of one of our experiments, it can be seen that all the untreated animals died about 35 days after grafting. When 3 mg BCG was injected on Day -14, only 2/7 of the tumours were rejected. Although this proportion was not statistically significant, the survival of the animals that died was significantly lengthened. When the high dose of BCG was injected on Day 0, 5/7 animals survived, this proportion being statistically significant, and the mean time of death of the remainder was considerably delayed. The injection of 3 mg BCG on Day 7 caused both a significant tumour rejection (5/10) and an increase in survival time of the mice that died. Low doses of BCG on Day 0 (0 1 mg) or Day 7 (0.5 mg) had no curative effect, though they did prolong life. It is of interest to note that when 1-5 mg of BCG was injected twice, with an interval of 7 days, into small growing tumours about 10 days after grafting (diameter 5 mm), there was no rejection in 9 mice, though the mean survival time was increased. This cytotoxicity was also tested for its sensitivity to irradiation (15 Gy) or heating (56°C, 30 min). Table II presents the results of 4 representative experiments, in which it can be seen (Exp. 1) that whole spleen-cell preparations from high-dose BCG animals inhibited the DNA synthesis of tumour cells with which they were co-cultivated in vitro.
This cytostatic effect was also exerted by the adherent cells from such populations, and treatment with anti-Thy 1.2+ C' did not diminish their activity (Exp. 2).
It can also be seen (Exp. 3) that the spleen cells from high-dose BCG-treated "B" mice were cytostatic to the growth of McB6-1 tumour cells.
In contrast to these results, cells from low-dose BCG animals had no significant effect in these tests (Exp. 4).
The figure demonstrates that the cytostasis was not abolished when the whole (or adherent) spleen cells were irradiated (15 Gy) before culture. In contrast, heating at 56°C for 30 min eliminated this effect. preparations was detected, and heatinactivated normal spleen cells had no inhibitory activity. In fact, increased DNA synthesis by the cultured tumour cells was found when heated cells were added, possibly because of nutritional factors. This effect was also noted when irradiated cells from normal animals were added to the cultures (Fig. 1).

Responsiveness to mitogen
It had been previously demonstrated in our laboratories that the i.v. injection of high doses of BCG into mice led to a series of modifications of the in vitro reactivity of lymphoid cells (Orbach-Arbouys & Poupon, 1978). We therefore extended our earlier studies of the development of cells capable of suppressing responses to PHA, and attempted to confirm their identity.
Suppression by whole and adherent spleencell populations from BCG-treated mice of T-cell response to PHA To determine whether the depression of mitogen responsiveness of spleen cells from high-dose BCG mice was an active phenomenon, such cells from BCG-treated animals were co-cultivated with normal T cells, and the response of the mixed population to PHA was measured. It can be seen in Table III that the addition of unfractionated spleen cells from high-dose BCG mice diminished the response of the normal T cells to this mitogen.
In a similar manner, when adherent spleen cells from high-dose BCG mice were added to normal T cells, there was a severe depression of the response, whereas adherent cells from normal mice had virtually no effect. This suppression persisted when the   Poupon, 1978), possibly owing to the presence of some residual adherent cells. It can also be seen in Table III that mice treated with low doses of BCG (0.5 mg) did not develop significant suppressive activity, whether the BCG was injected 14 or 21 days before the experiment. When the spleen cells of such mice were tested 6 weeks after injection, a low but significant (38%, P < 0.01) suppression was detected (results not presented).
It is of interest that the spontaneous [3H]dT incorporation by normal spleen cells was significantly lower than that of the BCG population in all our experiments.

Suppression of T-cell PHA responses by spleen cells from BCG-treated B and nude mice
In the light of the previous results, it was thought important to determine whether in fact T cells were necessary for the generation or expression of the suppressive effect on PHA responsiveness.
In in their spleen-cell populations. It is interesting that normal nude mice had very little spleen-cell response indeed to PHA, and the treated mice even less.

DISCUSSION
An anti-tumour effect of BCG in experimental animal models has often been reported (Hanna et al., 1972), but in many systems this effect is prophylactic rather than therapeutic (Finklestein et al., 1972). When, however, therapeutic effects were found against solid tumours, it was usually necessary also to inject the BCG directly into the lesion (Hanna et al., 1972). In our experiments with a solid transplantable :3-methylcholanthrene-induced fibrosarcoma of C3H mice, a significant antitumour effect was found when high doses of BCG were injected systemically at the same time as, or 7 days after, the graft. This protection was expressed as a significant proportion of the animals controlling and subsequently rejecting their tumours. Even the animals that died survived longer. In this model, high-dose systemic BCG exerted much less antitumour activity when applied prophylactically. Also, the intra-tumoural injection of BCG exerted very little protection, detectable only as a lengthened survival time after grafting. The fact that BCG administered systemically exerted an effect in this model, but not in many others, may be explained by the requirement for a high dose (3 mg); lower doses (0.1 or 0 5 mg) that are commonly used had no curative effects, though they did increase survival times. It is of interest to note here that animals that had rejected their tumour grafts after BCG treatment were resistant to further challenge grafts, whereas only about half of the animals that had been surgically cured of their tumour rejected a second graft, (results not presented). This observation can be interpreted to mean that high-dose systemic BCG activates the immune mechanism, or at least allows longer and more effective contact of the animal with the tumour antigens.
We decided, however, to examine in more detail nonspecific mechanisms that might contribute to BCG-stimulated tumour control. Considering the well documented effect of BCG on the reticuloendothelial system (Hibbs, 1975) we evaluated more precisely the role of the macrophage.
We chose a cytostatic assay to examine any nonspecific in vitro anti-tumour effects of high-dose systemic BCG, and also studied the suppressive effect of spleen cells from such animals on T-cell PHA response. We found that spleen cells from animals treated systemically with high (but not low) doses of BCG greatly inhibited the proliferation of cultured fibrosarcoma cells and PHA-stimulated lymphoid cells. The cells responsible were found to be strongly adherent to plastic, and the activity was removed by the sequential exposure of the spleen cell population to plastic and colloidal iron. In earlier studies (Orbach-Arbouys & Poupon, 1978) it was reported that, in addition to the adherent fraction, suppressive and cytostatic activities were found in T-cell populations that had been prepared by passage through nylon-wool columns. In the present study, however, we demonstrated that only 75%o of such preparations expressed the surface markers of T cells, in contrast to 9400 in cell populations obtained from normal animals. The apparent contamination of the former may represent T cells that were not expressing Thy-1.2 anitigen due to their stage of proliferation or differentiation, or non-T cells that were weakly adherent for similar reasons.
We found that the cytostatic and suppressive activities were not affected by treatment with anti-Thy 1.2+ C', were not sensitive to irradiation, but were lost upon heating. Confirmation of the non-Tcell character of the cells responsible, and their independence of a functional thymus, was obtained by our studies of BCGtreated "B" or nude mice.
The cytostatic and suppressive cells have, therefore, the properties of macro-phages. Similar results have been reported in other systems (Klimpel & Henney, 1978) though the i.v. route of BCG injection has been suggested as ineffective (Germain et al., 1975). This latter difference from our results can probably be explained by the dependence of the effect on the use of high doses of BCG. In fact in other systems, such as infection with Trichinella 8piralis (Jones et al., 1976) or Corynebacterium parvurn (Scott, 1974), while the injection of high doses is largely suppressive, small doses are predominantly stimulatory for a variety of responses.
The T-cell independence of these activities is most important, as Mackaness (1 970) and Evans & Alexander (1972) propose as a mechanism the following scheme: T cells specific for the antigens of BCG, upon reaction with these, secrete mediators (lymphokines) that provoke the activation of monocytes and macrophages, which are then cytostatic. Our results suggest, however, a more direct influence of BCGon the macrophage. It should also be noted here that in various acute infections suppressive activity appears to be due to either T cells or macrophages, but not both together (Jones et al., 1976;Bullock et al., 1978;Scott, 1974).
It is of considerable importance to discuss a problem related to the technique of the cytostatic and suppressive assays used in our studies. Waldman & Gottlieb (1973) have presented evidence that macrophages are able to secrete a factor that apparently inhibits mitosis, and Badger et al. (1973) have described a factor with similar effects secreted by mitogen-stimulated lymphocytes. Calderon et al. (1974) and Stadecker et al. (1977) suggested that many of these inhibitors may in fact have no such effect in reality, but only modify the incorporation of [3H]dT into DNA. Large quantities of endogenously produced dT competing with [3H]dT could, for example, exert such an effect. These authors report that assays using certain tumour cell lines (such as the EL4) are particularly sensitive to this type of inhibition. In order to clarify the possibility that this phenomenon occurred in our assay system, we evaluated the influence of washing the cultures before labelling, to remove possible soluble factors liberated by the spleen cells. This had no significant effect (results not presented). Such washing, however, can only eliminate soluble factors from the culture medium, and would have no effect on materials already incorporated into the cells. It is important to note in this respect that previous publicationis from our laboratory (Orbach- Arbouys & Poupon, 1978) and our present PHA results demonstrate that the spontaneous incorporation of [3H]dT by BCGtreated spleen cells was considerably more than that of normal populations. This was the case for whole spleen cell preparations, as well as for T or adherent cells obtained from them. It can therefore be asserted that if autoinhibition of [3H]dT incorporation by the spleen cell did not occur this is even less likely for bystander cells.
The fact that significant in vivo and in vitro anti-tumour activities were always paralleled by the generation of nonspecific suppressor-cell activity in our system poses the question of the involvement of the latter cells in these mechanisms. In effect, considerable uncertainty surrounds the role of immune responses, specific or not, in the control of tumour growth. This probably arises largely from the great complexity of the host-tumour relationship, and the participation of numerous effector and suppressor systems. Considerable information exists in the literature on the effect of suppressor cells on antitumour responses. Anergy, or immunodeficiency, have been associated with the development of suppressor cells in neoplasias such as myelomas (Broder et al., 1975) or Hodgkin's disease (Twomey et al., 1975). It appears, however, that when the suppression is specific, suppressor T cells can be demonstrated. As the injection of BCG causes the development of nonspecific suppressor cells, probably macrophages, it is reasonable to postulate that the effects on the host-tumour relationship might be 836 N. CASTES, N. R. LYNCH, G. LESPINATS AND S. ORBACH-ARBOUYS different. Some evidence does, in fact, implicate suppressor cells in tumour control. For example, the existence of nonspecific concomitant immunity to the growth of a tumour different from the primary ( Van der G(aarg & McCullagh, 1978) may be due to tumour-induced suppressor cells. It has also been reported that adherent macrophage-like cells that can inhibit lymphocyte response to mitogens can be correlated with a better prognosis in human cancer (Mikulski & Muggia, 1978), and that cytotoxic T cells appear to be most effective when in a non-dividing state (Cantor & Jandinski, 1974). It can therefore be suggested that a suppressive activity, that is most conveniently measured by inhibition of mitogen response, may be of great importance on account of its capacity to interfere with the proliferation of a number of cell types, lymphoid or tumoral. In such a case, chemotherapy protocols that are designed to diminish suppressor-cell activity (Orbach-Arbouys & Castes, 1979) and studies using mitogen response as an indicator of the "immunocompetence" of cancer patients (Mekori et al., 1974) should be evaluated with care. A notable lack of correspondence between the anti-tumour effects of indomethacin and activation of depressed mitogen response for example, has been reported, and severely depressed mitogen response does not necessarily mean the existence of a general state of immune depression (Lynch et al., 1978). It is obvious that the generation of suppressor cells could be disadvantageous to the host if they limited specific immune response, though considering the possible "blocking" activity of antibody, a preferential expression of nonspecific mechanisms may aid in tumour control. It can be suggested that excessive production of suppressor cells, such as might occur when tumours are particularly large (Poupon et al., 1976) or when very high doses of BCG are injected (Olsson et al., 1978), might facilitate tumouir growth.
In conclusion, we can coinsider 3 possible mnechanisms by which BCG-induced sup-pressor cells could be involved in tumour control. One is that the suppressor cells are generated in parallel with the actual effector cells, and might regulate, positively or negatively, the activity of such cells. A second possibility is that the suppressor cells are precursors of the effector cells, and a third is that the suppressor cells are actually a population of anti-tumour effector cells. Considering the parallels that we have found between the suppressor cell and anti-tumour activities, we favour the third possibility. Thus the activity of these cells, though detected as an in vitro suppression of various responses, may be an important mechanism of tumour control.