Twelfth annual Pezcoller symposium: signaling cross-talks in cancer cells.

Signaling as an overall cellular process is a prerequisite for the life of normal and cancer cells and is regulated by a multiplicity of mechanisms including "cross-talks" among pathways. Cross-talk is often critical in determining the decision of a cell to go toward a specific biological direction, such as proliferation, apoptosis, or differentiation. Both signaling pathways and cross-talk among them involve such biochemical processes as protein-to-protein interactions and well-regulated phosphorylation and dephosphorylation processes. This symposium was focused on mechanisms of the regulation of signaling with special emphasis on regulatory cross-talk among pathways and on changes that occur in cancer cells. The phenomena discussed were also considered with reference to opportunities that may be offered toward the development of potential therapeutic intervention.

Selection for high immunogenicity in drug-resistant sublines of murine lymphomas demonstrated by plaque assay.

The immunogenicity of lymphoma L1210 and three L1210 sublines, resistant to methylglyoxal bis(guanylhydrazone), 4,4'-diacetyldiphenylurea bis(guanylhydrazone), or guanazole (L1210/GZL), respectively, was evaluated. Syngeneic DBA/2J mice were given a single i.p. injection of serially diluted suspension of irradiated cells from L1210 or L1210 sublines. Five days later spleen cells from the immunized mice were tested for the presence of plaque-forming cells using the immunizing lymphoma cell lines as target. Sera collected from the animals were examined for cytolytic antibody activity by lysis in gel using the same target cells. For comparison, the H-2 immunogenicity of L1210 and its sublines was investigated in H-2-incompatible allogeneic mice. The following results were obtained. (a) All the sublines showed increased immunogenicity and susceptibility to lysis as compared to L1210 cells. The number of plaque-forming cells/spleen ranged from 100 for L1210 to 4450 for L1210/GZL, the most immunogenic subline, and the antibody titer ranged from 1/8 for L1210 to 1/128 for L1210/GZL. (b) All the sublines carried common tumor-associated antigens that apparently made primary contributions to the increased immunogenicity. (c) The common tumor-associated antigens were also expressed on L1210 cells, although in a lesser defree, as evidenced by the definite, albeit low, capacity of L1210 cells to absorb DBA/2J anti-L1210/GZL antibodies. (d) Spleen and thymus cells of DBA/2J mice as well as unrelated murine ascites tumor cells did not cause significant absorption of these antibodies. (e) Only a partial inverse relationship could be demonstrated between tumor-associated antigens but the lowest for H-2. The above results would seem compatible with the hypothesis that the increased immunogenicity of drug-resistant L1210 sublines is attributable to the selection of preexisting highly immunogenic cells during immunosuppression by treatments selecting for drug resistance.

Reversal of the in vitro methotrexate suppression of cell-mediated immune response by folinic acid and thymidine plus hypoxanthine.

The development of a primary complement-independent cellular cytotoxic immune response in culture by C57BL/6J spleen cells stimulated with X-irradiated allogeneic P815 tumor cells was inhibited more than 50% in the presence of 1.5 X 10(-8) M methotrexate. This immunosuppression by methotrexate was time and dose dependent. Equimolar folinic acid administered at either -4, 0, +4, or +24 hr relative to 1.5 X 10(-8) M methotrexate reversed immunosuppression by more than 50%. Increased folinic acid concentration (5 to 10-fold) completely restored the immune response only if added 4 hr prior to methotrexate. Thymidine plus hypoxanthine (100 microM each) when present throughout the 4-day culture period gave total reversal of immunosuppression. The reversal was also obtained with hypoxanthine alone and was dose dependent. These results indicate that reversal of the methotrexate-induced impairment of cellular immune function depends on several parameters including the concentration of methotrexate and of the reversing agents as well as the time of exposure of relevant target cells to these agents.

Fast- and slow-growing transplantable tumors derived from spontaneous mammary tumors of the DBA/2 Ha-DD mouse.

Nineteen transplantable tumor lines were established from individual spontaneous mammary tumors of the DBA/2 Ha-DD mouse. Of these lines, 5 were classified as fast growing, 8 as medium growing, and 6 as slow growing, based on the time required for the tumors to reach a size of 10 mm in average diameter and on the average survival time of tumor-bearing syngeneic hosts. The relative differences in rate of growth among 5 of these lines remained stable during 11 to 19 transplant generations. In DBA/2J mice, a slow-growing and a fast-growing tumor line were cross-immunogenic. The differences in growth rate between these 2 tumor lines were not primarily related to differences in immunogenicity since they were not abolished in preirradiated hosts. The growth of cell populations from these 2 tumor lines in culture was comparable; however, cells from the fast-growing line had a plating efficiency about 4 times higher than those from the slow-growing line.

The twentieth international symposium of the Sapporo Cancer Seminar Foundation: gene environment interaction and cancer prevention.

The main goal of this Symposium was to discuss new information that could be used for the development of effective and novel approaches in cancer prevention. Mounting evidence indicates that genetic predisposition to cancer plays an important role in the etiology of the disease and that multistage carcinogenesis is for the most part based on multiple genetic changes, favoring cell survival. It is also evident that a variety of environmental factors lead to carcinogenic changes and determine cancer-causative or cancer-facilitative genetic changes; these factors may be endogenous in origin, as for example those that are endocrinologic in nature, or may come from the external environment, as for example products in tobacco smoke or in industrial pollution. It is therefore obvious that gene-environment interactions play a pivotal role in carcinogenesis and consequently may offer specific sites of intervention that may be useful for the development of cancer prevention. In general, cancer prevention can be implemented through Public Health measures, as is typically the case for intervention on smoking, industrial pollution, aflatoxin B contamination or, when we know more about it, dietary habits. On the other hand, it should become possible to design more rational chemoprevention and immunoprevention trials as more becomes known about the genetic changes predisposing to cancer and about the gene products that are responsible for the phenotypic changes leading to neoplasia.

Increased primary cell-mediated immunity in culture subsequent to adriamycin or daunorubicin treatment of spleen donor mice.

Spleen cell populations from mice treated with Adriamycin or daunorubicin were found to develop a greater complement-independent cellular cytotoxic immune response during culture with allogeneic tumor cells than spleen cells from untreated or cyclophosphamide-treated animals. A temporal and drug dose dependence of this effect was demonstrated. The changes in spleen cell population occurring in the donor mice consequent to drug treatment were evident in the nylon woll-adherent fraction of the spleen cells. The results are consistent with the possibility that the concentration of specific progenitor or accessory cells in the spleen is increased consequent to drug treatment.

Role of tumor necrosis factor and interleukin 1 in gamma-interferon-promoted activation of mouse tumoricidal macrophages.

The purpose of this study was to determine if recombinant murine interleukin 1 beta (rMu-IL-1 beta) alone or in combination with recombinant murine gamma-interferon (rMu-IFN-gamma) could activate murine macrophages to be tumoricidal against tumor necrosis factor (TNF)-insensitive target cells and to evaluate the possible role of interleukin 1 (IL-1) in murine macrophage activation by recombinant murine tumor necrosis factor (rMu-TNF) plus rMu-IFN-gamma. rMu-IL-1 beta and rMu-TNF alone or in combination could neither directly lyse the TNF-insensitive P815 mastocytoma nor activate resident peritoneal macrophages to be tumoricidal for this target. A synergistic induction of tumoricidal macrophage activity against P815 occurred, however, when either of these monokines was combined with rMu-IFN-gamma. The tumoricidal activity obtained was transitory, and the level of activity was dependent upon the monokine concentration and the length of induction period. Murine macrophages stimulated under the same conditions used to induce tumoricidal activity with rMu-TNF plus rMu-IFN-gamma or with rMu-IL-1 plus rMu-IFN-gamma were shown to produce low concentrations of IL-1 or TNF, respectively. Thus, a bidirectional cross-induction of the production of the two monokines occurred. The monokine production was also quite transitory, and the time of peak production of the monokines (12 h) was found to precede the time of peak tumoricidal activation (24 h). Using neutralizing antisera specific for rMu-IL-1s and rMu-TNF, the cross-induced production of TNF was shown to be required for macrophage tumoricidal activation by rMu-IL-1 beta alone (TNF-sensitive targets) or in combination with rMu-IFN-gamma (TNF-insensitive targets). There was no evidence, however, that the production of IL-1 was required for macrophage activation by rMu-TNF in combination with rMu-IFN-gamma.

Alterations in murine host defense functions by adriamycin or liposome-encapsulated adriamycin.

Peritoneal exudate cells (PEC) from C57BL/6 mice were collected on different days following an i.p. injection of Adriamycin (10 mg/kg) as free drug (ADM) or encapsulated in multilamellar liposomes (ADM/Lip). Macrophages harvested from mice at various times (Days 4-14) after either drug treatment were responsive to in vitro lipopolysaccharide induction of tumoricidal activity, maximum response being seen on Day 7. In addition, 18 days after treatment, significant macrophage tumoricidal activity was observed only in the ADM/Lip-treated group. When supernatants from cultures of PEC obtained 7 days after treatment were assayed for interleukin 1 following lipopolysaccharide stimulation, activity was found with both ADM- and ADM/Lip-treated cells. Without lipopolysaccharide stimulation, only PEC from ADM-treated mice elaborated factor(s) with interleukin 1-like activity. Both ADM and ADM/Lip induced significant PEC-natural killer (PEC-NK) activity by Day 4, while the ADM/Lip treatment sustained PEC-NK activity more effectively than free drug at later time points (7 or 11 days posttreatment). Drug-induced PEC-NK activity (Day 7) was (a) ablated by treatment in vitro with anti-asialo GM1 antibody and complement, and (b) associated with a population of PEC nonadherent to plastic. A transient suppression of splenic NK activity was seen 4 days following either ADM or ADM/Lip administration with recovery to control level by Day 7. These data demonstrate that following ADM or ADM/Lip administration some of the changes necessary for macrophage tumoricidal activation must have occurred in vivo. Liposome encapsulation of ADM extended the duration of ADM-induced augmentation of certain host defenses.

Tumoricidal activation of murine resident peritoneal macrophages by interleukin 2 and tumor necrosis factor alpha.

The capacity of recombinant human interleukin 2 (rH-IL2), alone or in combination with recombinant tumor necrosis factor (r-TNF alpha), to activate murine resident peritoneal macrophages to a tumoricidal state was examined. Resident peritoneal exudate cells from C57BL/6 mice were cultured for 18 h with activating agents and washed and the adherent cells (macrophages) were assessed for cytolytic activity against radiolabeled target tumor cells (EL4, P815). Under these conditions, rH-IL2 alone activated macrophages to a tumoricidal state in a concentration dependent fashion. Neither murine nor human r-TNF alpha alone had any activating effect but, when combined with rH-IL2, further stimulated rH-IL2-inducible responses. Using polymyxin B, it was shown that macrophage activation was not due to an inadvertent lipopolysaccharide contamination of the r-TNF alpha or rH-IL2 preparations. It was also unlikely that target cell lysis was a direct result of increased TNF alpha production by rH-IL2 stimulated macrophages since P815 is totally resistant to lysis by r-TNF alpha. Although the lytic effector function was mediated by adherent cells, nonadherent peritoneal exudate cells were required for activation to occur. Furthermore, antisera against murine gamma-interferon, when added to activation cultures, reduced the level of cytolytic activity which developed. These data suggest that rH-IL2-induced peritoneal macrophage activation requires stimulation of nonadherent cells and is dependent upon gamma-interferon mediated mechanisms.

Augmentation of the development of immune responses of mice against allogeneic tumor cells after adriamycin treatment.

In C57BL/6J mice, depending on the dose of P815 cells used for immunization, Adriamycin exerted different effects on the cell-mediated lytic response and complement-dependent cytotoxicity. At the dose of 3 X 10(7) P815 cells, Adriamycin treatment had no apparent effect on cell-mediated lytic response regardless of timing of drug treatment. At lower doses of antigen (10(7) or 5 X 10(6) cells), the response was augmented in Adriamycin-pretreated mice. Similarly, under conditions which led to a suboptimal complement-dependent humoral response of untreated control, Adriamycin pretreatment resulted in an augmented response; under conditions of maximal response, Adriamycin was suppressive. Suppression was maximal if the drug was injected at either the same time or shortly before or after antigen. The cell-mediated lytic response was proportional to the dose of antigen used, while the complement-dependent humoral lytic response was inversely proportional to dose of antigen in the range used in these experiments. Secondary cell-mediated lytic response in culture was also augmented if mice had been pretreated with Adriamycin 5 days before the primary immunization. The cell-mediated lytic response of spleen and peritoneal exudate cells from mice immunized with relatively low doses of P815 cells 5 days after treatment with Adriamycin was increased 12 to 15 days after immunization. The cytotoxic effects were present in both plastic adherent and nonadherent fractions of either spleen or peritoneal cell populations. All these effector cells were found to be anti-Thy 1.2 sensitive. The phagocytic activity of spleen cells was increased after immunization, but no drug effect was observed; following 24 hr of culture, however, cells from drug-treated immunized donors had increased phagocytic activity as compared to that of controls. Increased phagocytosis also developed in cells nonadherent to plastic.

Differential expression of murine leukemia antigen on L1210 parental and drug-resistant sublines.

The different expression of surface antigens on L1210 leukemia DBA/2 and drug-resistant L1210 sublines was investigated. Indirect cytotoxic test, the anti-L1210/v alloantiserum reacted more strongly with subline cells than with parental cells. Absorption of the antiserum with Gross cellular surface antigen-positive AKR leukemia (AKSL-4) cells led to a much greater difference in this reactivity. Quantitative absorption experiments revealed that the drug-resistant sublines had 5 times higher absorption capacity than did the parental line. After complete absorption of antibodies against murine leukemia virus-related antigens, the anti-L1210/v alloantiserum still reacted with L1210 cells. This cytotoxicity could be removed after absorption with C3H mammary tumor (MAC-1) cells but not with normal C3H lymphocytes. These results provide evidence that the major cytotoxic activity of the antiserum against L1210 and L1210 subline cells was due to antibodies against murine mammary tumor virus-related antigen and that the drug-resistant sublines of leukemia L1210 have higher quantitative expression of mammary leukemia antigens.

Effect of recombinant tumor necrosis factor on tumoricidal activation of murine macrophages: synergism between tumor necrosis factor and gamma-interferon.

The activation of tumoricidal murine macrophages by recombinant human tumor necrosis factor (rH-TNF) alone or in combination with recombinant murine gamma-interferon (rM-IFN-gamma) was examined. When used alone, rH-TNF (10(-1)-10(5) units/ml) did not induce macrophage tumoricidal activity against TNF-insensitive P815 mastocytoma cells. Combining rH-TNF with rM-IFN-gamma resulted in the synergistic induction of tumoricidal activity in resident peritoneal macrophages. This synergistic effect was not due to contaminating bacterial lipopolysaccharide. A comparative study using recombinant murine tumor necrosis factor (rM-TNF) showed that rM-TNF alone also could not stimulate murine macrophages and there was no significant difference between effects of rM-TNF and rH-TNF on macrophage activation in the presence of rM-IFN-gamma. In experiments comparing sequential to simultaneous exposure of macrophages to rH-TNF and rM-IFN-gamma, it was found that: (a) when macrophages are primed with rM-IFN-gamma, rH-TNF serves only as a very weak triggering signal for tumoricidal activation; and (b) marked activation is obtained only when macrophages are exposed to the two cytokines simultaneously. These results suggest that TNF has an autocrine regulatory function in concert with lymphokines in macrophage-mediated host defense against tumors.

Effect of antimyeloma cell antiserum on immunological enhancement.

Growth of Sarvoma 180 in AKR mice was enhanced by immunization with frozen-thawed homogenates of this tumor. Treatment of host mice with rabbit antimyeloma cell antiserum, either furing immunization or shortly after tumor implantation, resulted in a decreased incidence of tumor rejection or an increased rate of tumor growth. A slow-growing subline from a spontaneous, DBA/2Ha-DD mammary tumor is rejected after initial growth in DBA/2J mice. The incidence of rejection was reduced by treatment with the antiserum studied.

Modification of host antitumor defense mechanisms in mice by progressively growing tumor.

The EL4 lymphoma in C57BL/6 mice was used as a model to examine the effect of progressive tumor growth on a variety of cell mediated cytolytic effector functions which have been shown in other systems to have antitumor potential. The functions examined were those of cytolytic T-lymphocyte, lymphokine activated killer cells, natural killer cells, and tumoricidal macrophage (MO). The kinetics of each function displayed a unique pattern as a consequence of tumor growth, but all were inhibited in animals bearing large tumors (late tumor bearers). In cell mixing experiments it was shown that spleen cells from individual late tumor bearers were suppressive for cytotoxic T-lymphocytes, lymphokine activated killer cells, and splenic MO but not peritoneal MO or splenic natural killer cells. The suppression was nonspecific and was mediated primarily by nonadherent cells and/or their soluble products. Suppression appeared to be mediated, in part, by tumor cells in the spleen since the degree of suppressor activity associated with a particular spleen cell preparation correlated with the number of tumor cells present. Furthermore, the direct addition of viable ascites EL4 cells to response cultures or assays had similar suppressive effects as late TBM spleen cells, i.e., inhibited cytotoxic T-lymphocytes, lymphokine activated killer cells, and splenic MO but had no effect on natural killer cells or peritoneal MO. The mechanism of suppression by ascites EL4 was not determined but it was mediated by viable cells only and not due to contaminating viruses or other microorganisms.

Cell population kinetics of fast- and slow-growing transplantable tumors derived from spontaneous mammary tumors of the DBA/2 Ha-DD mouse.

The proliferation kinetics of a fast-growing spontaneous mouse mammary tumor subline (SMT-F) and a slow-growing spontaneous mouse mammary tumor subline (SMT-S) tumor have been determined autoradiographically at 2 different stages of tumor growth. The length of the cell cycle and the growth fraction for SMT-F were 11.2 hr and 0.85, respectively, on Day 14 and 12.1 hr and 0.78 on Day 28. For SMT-S these same parameters were 15.6 hr and 0.50 on Day 14 and 16. 1 hr and 0.45 on Day 28. On Days 14 and 28 the mitotic indices were 1.3 and 1.0%, respectively, for SMT-S, compared to 2.2 and 1.9% for SMT-F. The cell loss rate, cell loss factor, and cell loss were significantly higher for SMT-F than for SMT-S. The difference in the growth rates for these 2 tumor lines was attributable to a slight prolongation of the length of the cell cycle and a reduction in the growth fraction of SMT-S.

Correlation between adriamycin-induced augmentation of interleukin 2 production and of cell-mediated cytotoxicity in mice.

Based on the observation that spleen cells from Adriamycin-treated mice could develop augmented levels of cytotoxic T-lymphocyte activity in response to heat-treated and/or X-irradiated alloantigens, it was postulated that modulations in soluble mediators could be involved in this phenomenon. In fact, in this study Adriamycin-induced increases in the levels of prostaglandin E2 and interleukin 2 activity have been observed with isolated cells. The "interleukin 2-like" activity was indistinguishable from that of partially purified interleukin 2 in terms of ability to restore responsiveness to experimentally inhibited primary alloantigen response cultures and to maintain long-term cultures of activated T-cells. Furthermore this latter activity was completely ablated by antiinterleukin 2 monoclonal antibody. While the modification in prostaglandin E2 production did not appear to play a role in determining augmentation of cytotoxic T-cell activity, the modification in interleukin 2 production was consistent with the possibility that this is a primary mechanism of Adriamycin-induced augmented cell-mediated cytotoxicity.

Selective modulation by alpha-difluoromethylornithine of T-lymphocyte and antibody-mediated cytotoxic responses to mouse tumor allografts.

alpha-Difluoromethylornithine (DFMO), an inhibitor of polyamine biosynthesis, has been shown to be growth inhibitory in a wide variety of normal and tumor cell systems. Since cells of the host defense system are among the most rapidly proliferating cells in the body, DFMO may inhibit certain components of this system. In order to assess this possibility, four randomized groups of C57BL/6 mice were maintained either on water throughout (controls) or on 2% DFMO in drinking water for various periods of time. Mice were given DFMO from Days -4 to 0, Days 0 to +4, or Day 0 to day of assay. On Day 0 randomly selected mice from each group received P815 tumor allografts. Daily from Days +3 to +14, pools of spleen cells from three mice per group were assessed for allospecific cytolytic T-lymphocyte, antibody formation, natural killer cell, and phagocytic cell activities. While natural killer cell and phagocytic cell activities remained essentially unchanged under all conditions, both cytotoxic T-lymphocyte and antibody responses were modified. Somewhat similar effects were seen with both responses and involved to varying degrees: (a) a delay of the initiation of rapid increase in the response but not in the onset of first detectable response; (b) delay in the time of peak response; (c) increased level of maximal response; (d) two peaks of maximal response. The data indicate that DFMO treatment of whole animals, dependent upon schedule of administration and time of assay, induces very selective effects on both cytotoxic T-lymphocyte and antibody responses, without apparent modification of nonspecific host defense mechanisms, with the overall effect being a prolongation of the period of specific response.

Role of tumor necrosis factor in macrophage activation and tumoricidal activity.

Tumor necrosis factor (TNF)-sensitive (LM) and -insensitive (P815) target cell lines were used to examine the role of TNF in both the activation and lytic phases of macrophage-mediated lysis. LM cells were lysed spontaneously by thioglycolate-elicited macrophages in an 18-h assay (media or activating agents added with targets) or 36-h assay (macrophages cultured with media or activating agents for 18 h, washed, and targets added for a subsequent 18 h). In contrast, P815 cells were lysed only in the 36-h assay by macrophages exposed to appropriate activation signals. Using antibody to murine TNF, it was shown that lysis of LM cells but not P815 cells was TNF mediated. The addition of lipopolysaccharide (LPS) to the 18-h assay resulted in augmented LM killing. This was probably due to the fact that LPS stimulates macrophages to produce TNF. Conversely, when macrophages were pretreated with LPS for 18 h, washed, and assessed for lytic activity during the subsequent 18 h, lysis of LM cells was reduced relative to the endogenous level. Although macrophage lysis of P815 was not mediated by TNF, the addition of TNF to macrophage activation cultures facilitated LPS triggering of cytolytic activity against P815. Similarly, the addition of TNF to the activation cultures partially prevented the LPS-induced reduction in macrophage-mediated LM cell lysis. Taken together, these data suggest that TNF may act as an autocrine signal during macrophage activation, in addition to being directly lytic to a select number of sensitive target cell lines.

Adriamycin-induced modulation of host defenses in tumor-bearing mice.

Using the C57BL/6/EL4 tumor model, studies were carried out to demonstrate the feasibility of administering Adriamycin (ADM) in therapeutic doses and schedules such that the host antitumor defenses would not be suppressed and in some cases might be stimulated by treatment. ADM treatment caused prolongation of survival and, in general, either stimulated host cytolytic activities above untreated control levels or had no effect. These effects by ADM were observed with the ADM-sensitive parent EL4 line as well as with an ADM-resistant subline, indicating that the effects did not result entirely from direct antitumor activity. The cytolytic activities examined were those of cytolytic T-lymphocytes, lymphokine-activated killer cells, and splenic and peritoneal macrophages. All activities were assessed against the syngeneic EL4 target line. The information obtained in this investigation provides a rational basis for the future development of curative protocols with ADM plus biological response modifiers, which would depend on a functional immune system for optimum efficacy and would also exploit synergistic immunomodulating effects of the agents used in combination.

Effect of recombinant human tumor necrosis factor on the induction of murine macrophage tumoricidal activity.

The ability of recombinant human tumor necrosis factor (rH-TNF) alone or in combination with lymphokines (LK) to induce the in vitro activation of murine macrophages was evaluated. The treatment of C57BL/6 mouse resident peritoneal exudate cells (PEC) with rH-TNF and LK was found to induce the activation of macrophages to a tumoricidal state against P815 mastocytoma cells. Neither rH-TNF nor LK alone induced macrophage cytotoxic activity. Furthermore, the macrophage activation seen was not due to small amounts of contaminating lipopolysaccharide. The TNF plus LK-mediated macrophage activation could be totally ablated by rabbit antiserum to murine gamma-interferon, thus suggesting a role for gamma-interferon in this system. Since adherent cells (greater than or equal to 95% macrophages) only marginally responded to stimulation with rH-TNF plus LK and the addition of nonadherent PEC caused a marked augmentation of rH-TNF plus LK-mediated macrophage activation, the involvement of nonadherent PEC was suggested. In addition, using antibodies and complement to deplete subsets of cells from the nonadherent PEC, the requirement for cells bearing Thy 1.2 and asialo GM1 surface markers was demonstrated. These results suggest that TNF may play an autocrine regulatory role in concert with lymphokines in macrophage-mediated host defense against malignant neoplasia.