Eliezer Gorelik

Selection of 3LL tumor subline resistant to natural effector cells concomitantly selected for increased metastatic potency

SummaryStudies were carried out to test whether the selection of 3LL tumor cells resistant to the cytotoxic activity of normal spleen cells will concomitantly select for increased metastatic capacity. A total of 0.5×106 3LL tumor cells and 50×106 normal spleen cells were admixed and inoculated SC in C57BL/6 mice. This procedure was repeated for eight transplant generations. The tumor cells which were selected in this manner (3LLN) were relatively resistant in vitro to the cytotoxic effect of normal spleen sells. The 3LLN cells were less susceptible than the original 3LL tumor cells to the hybrid resistance mechanisms, as determined by their growth capacity in F1(BALB/c×C57BL/6) or F1(C3H/DiSn×C57BL/6) mice. After three selection generations, 3LLN tumor cells were more efficient in the production of spontaneous lung metastases. These results support our suggestion that natural cell-mediated immunity may be involved in the control of metastatic spread.

Interactions between the local tumor and its metastases.

The generation of metastases by tumor cells is an end result of a sequence of processes, each of which is controlled by distinct properties of the metastatic cell. Thus, the detachment of carcinoma cells from the local tumor graft, their binding to and penetration through capillary endothelia, their dissemination within the blood circulation, their homeotypic aggregation in the blood vessels of target organs, their recognition and hence their binding to normal endothelial or other cells of their metastatic sites, their capacity to induce angiogenesis at these new anatomical sites, each of these steps might be a function of properties of metastatic cells which nonmetastatic tumor cells may not possess. Indeed, many of the experimental approaches to the mechanisms associated with the control of metastasis were based on the demonstration that a) tumor cells constitute a diverse population with regards to their metastatogenic potential (1-3), and b) metastatogenic cells are endowed with specific properties, distinct from those which characterize cells of the local growth. These properties should control each of the sequential stages leading to the growth of metastases. Thus, studying the metastatic Lewis lung carcinoma (3LL), we demonstrated that individual clones of the 3LL tumor differ in their metastatic potency (4). Cells of a clone capable of generating metastases manifested a significantly lower susceptibility to NK cells than cells of a nonmetastatogenic clone (4). This property could determine their survival while disseminating in the circulation (5-7). Since the selection for NK resistance concomitantly selected for increased metastatogenic potency (6), and since the inoculation of a nonmetastatic clone to NK-deficient beige (Bg/Bg) mice resulted in the generation of metastases (4), it appeared that this cell-surface property might be causally related to the metastatic capacity of the cells. We also demonstrated (7, 8) that cells from lung metastases (M-3LL) differ in their cell surface antigens from the cells of the local 3LL (L3LL), thus enaNing immunoselective processes to operate in tumor dissemination. It was also found that the metastatic 3LL cells are capable of enzymatic degradation of the collagen type II which characterizes the basement membrane of blood vessels. Yet, although such and other differences between the metastatic and the nonmetastatic 3LL cells were demonstrated, it appears that once the two populations segregate in vivo, thus forming a growing local tumor and progressively growing lung metastases, regulatory interactions operate between them. The interactions which we shall review stem from the observation that the surgical removal of the local tumor is followed by an accelerated growth and an increased incidence of lung metastases (9). This observation thus suggested that the local tumor exerts an inhibiting

Modulation of formation of tumor metastases by peritoneal macrophages elicited by various agents

SummaryWe have studied the formation of experimental B16 melanoma metastases in the lungs of mice inoculated IV with tumoricidal or nontumoricidal peritoneal macrophages elicited by various agents. IV inoculation of peritoneal Mϕ elicited by Brewers thioglycollate medium (TG-Mϕ) 1 day before the injection of B16 melanoma cells dramatically increased the number of metastatic foci in the lungs. NIH thioglycollate broth and proteose peptone each elicited a relatively low number of Mϕ, which were morphologically distinguishable from TG-Mϕ and did not influence the yield of B16 melanoma colonies in the lungs. Resident or C. pravum-elicited Mϕ also did not augment metastatis formation. TG-Mϕ became highly tumoricidal after IP stimulation with poly I: C. However, tumoricidal TG-Mϕ inoculated IV 1 day before IV inoculation of B16 melanoma cells did not have an antimetastatic effect. On the contrary, both tumoricidal and nontumoricidal TG-Mϕ augmented metastasis formation. Poly I: C treatment had a substantial antimetastatic effect in the normal mice, but not in mice with adoptively transferred TG-Mϕ. Histological analysis revealed that IV-inoculated TG-Mϕ (tumoricidal or nontumoricidal, either viable or disrupted) induced severe intravascular reaction in the lungs, but not in the liver or kidney. This reaction manifested in the aggregation of the various blood cells, preferentially neutrophils. These reactions were not observed after IV inoculation of PMϕ or NIH TG-Mϕ.Intravascular inflammatory reactions induced by TG-Mϕ may be responsible for the augmentation of metastasis formation, partly by suppression of NK reactivity and mostly by the acceleration of the processes of tumor cell extravasation. These data may provide some insight into the failure to achieve systemic adoptive immunotherapy using activated peritoneal TG-Mϕ.

In vivo resistance of secondary antitumor immune response to cyclophosphamide: effects on T cell subsets

SummaryWe have analyzed the effects of high doses of cyclophosphamide (Cy) on primary and secondary antitumor immune response against immunogenic (tum−) variants of Lewis lung carcinoma (3LL) treated in vitro with UV light. Normal mice and mice previously immunized with tum− clones were inoculated i.p. with Cy (200 mg/kg body weight) and 24 h later challenged intrafootpad with tum− or parental 3LL cells. Cy treatment suppressed the primary immune response of normal animals and allowed the growth of tum− cells. In contrast, Cy-treated immune mice rejected the tumor challenge. The in vivo treatment with Cy decreased the total number of lymphoid cells in the spleens, as well as the proportion of B lymphocytes; however, it increased the percentage of both Lyt2+ and L3T4+ lymphocytes. Thus, the immunosuppressive effects of Cy on the primary antitumor response could not be attributed to elimination of major T lymphocyte subpopulations. Although the treatment of immune mice with Cy did not significantly impair their antitumor resistance, nor the proportion of Lyt2+ and L3T4+ lymphocytes in their spleens, the in vitro generation of cytotoxic T lymphocytes (CTL) was markedly reduced.After Cy treatment, the proliferative ability of spleen cells in response to interleukin-2 (IL-2) was substantially impaired. Using monoclonal antibodies to the IL-2 receptor, we found that Cy-treated T lymphocytes failed to fully express the IL-2 receptor following in vitro stimulation with irradiated tumor cells. In line with these findings, the in vitro generation of CTL was not restored by addition of recombinant IL-2 to the cultures. In vivo experiments using purified functional subsets of immune T cells showed that Lyt1+, but not Lyt2+ lymphocytes were able to transfer antitumor immunity in normal irradiated recipients.Therefore, since Ly1+ T lymphocytes were responsible for the antitumor resistance in vivo, the Cy-induced impairment of CTL generation did not affect the ability of immune mice to reject a secondary tumor challenge.

Effect of splenectomy on the progression of postoperative pulmonary metastases of the 3LL tumor

Surgical excision of the local intrafootpad tumor of the 3LL lung carcinoma is followed by accelerated growth of its lung metastases. When, however, splenectomy was performed concomitantly with tumor excision, the acceleration of lung metastases was prevented. In cases where excision of the local tumor took place when it reached large sizes, concomitant splenectomy did not prevent the accelerated growth of the lung metastases. If, however, at these stages of tumor growth splenectomy was performed 3 days prior to the excision of the tumor, it did prevent the accelerated growth of metastases. Intrafootpad reinoculation of tumor cells following tumor excision and splenectomy caused further reduction in metastatic growth. The results suggest that existence of two possible distinct mechanisms which control metastatic growth: the local tumor might exert non-immunologically, an inhibitory effect on its lung metastases, and the spleen, possibly via suppressor lymphocytes, may suppress an immune effector activity against the tumor metastases, an activity which is manifested following splenectomy.

Immunobiological diversity of metastatic cells

The analysis of antigenic and immunogenic properties of tumor cells and of mechanisms of antitumor response should furnish the basis for the investigation of methods aimed at immunotherapy of malignant diseases. One may expect that such methods will be especially effective in the prevention of metastatic progression following surgical removal of the primary tumor mass. Immunotherapeutic approaches are based either on nonspecific stimulation of the host’s immune system or on the specific immunization of the organism against tumor cells originating in the surgically removed tumor. These immunization procedures are expected to stimulate the development of cytotoxic lymphocytes or antibodies which are capable of inhibiting tumor growth and destroying metastatic tumor cells. Another experimental approach to immunotherapy is based on the adoptive transfer to the diseased host of lymphocytes sensitized in vitro against tumor cells [1].

Organ-Specific Homing of B-Cell Hybridomas

In an adult animal, the physiological rule for red and white blood cells is continuous traffic (1). Conversely, the physiological rule for parenchymatous and endocrine organ cells is not to travel. Traffic of cells from organs is generally associated with malignant transformation. Malignant cells can leave primary tumor mass, enter the blood or the lymph circuits and are found to settle and proliferate in sites distant from the original tumor. Hence the acquisition of traffic properties by malignant cells could be a basic mechanism of metastatic dissemination. In the present work the metastatic capacity and homing characteristics of plasmacytoma-B lymphocyte hybridization was studied. The results indicated that somatic hybridization of non-metastatic tumor cells with cells which express high traffic poentialities (i.e. lymphocytes) conferred metastatic properties on a non-metastatic tumor cell and modified the homing properties of the parental tumor cell.

Immunogenetic Determinants Controlling the Metastatic Properties of Tumor Cells

The generation of metastases by tumor cells is preceeded by a sequence of processes, each of which is controlled by different properties of the invading cells. Thus, the detachment of tumor cells from its local population might be associated with alterations in cell-to-cell recognition patterns, its metastatic penetration through the basement membrane of blood capillaries might require the synthesis of specific enzymes capable of degrading membrane proteins (Liotta et al., 1977; Strauli and Weiss, 1977), its “homing” to specific organs might be determined by specific adherance to the normal cells of the target organ (Kramer and Nicolson, 1979), whereas its growth at its new site might require the secretion of angiogenic factors (Folkman and Hochberg, 1973; Folkman and Handeschild, 1981).

Immunogenetic Determinants of Metastatic Cells

The progression of metastases is controlled by a series of cellular interactions between host and tumor cells. The generation of blood-bourne metastases is initiated by tumor cells capable of adhering to capillary basement membranes, followed by penetration, apparently via enzymatic degradation, of its proteins (1–3).

The control of tumor metastasis.

Most carcinomas and sarcomas in man are metastatic neoplasms. On the other hand, most solid tumors in mice, whether spontaneous or experimentally induced, are nonmetastatic. In the absence of mouse metastatic tumors, experimental models for the study of tumor metastasis have been rather limited, hence, the limitations imposed on the study of the mechanisms controlling metastatic spread and progression. In fact, some very basic questions have not been fully investigated, e.g., the question of whether a population of tumor metastasis cells is a random representative of the population of the local growth or whether it is a selected subpopulation possessing cell surface properties which might determine the metastatic properties. Recent studies have indicated that metastatic cells may differ from the primary cell population in a number of properties, such as susceptibility to drugs (Trope 1975), affinity to different organs (Fidler and Nicolson 1976; Nicolson et al. 1976) and chromosome number (Rabotti 1959; Chu and Ulmgren 1961).