Effects of specific active immunization on tumor recurrence following primary tumor resection in WF rats with 1,2-dimethylhydrazine-induced bowel cancer

Primary gastrointestinal tumors were induced in male WF rats by 16 weekly sc injections of 1,2-dimethylhydrazine [(DMH) CAS: 540-73-8; 20 mg/kg/wk]. Twenty-four to 28 weeks after the start of DMH injections, all rats were surgically explored and gastrointestinal tumors were resected. Rats with no remaining microscopic disease after operation were immunized with one of four tumor isografts. The first isograft, DMH-W163, is a poorly differentiated mucinous adenocarcinoma explanted from a colon cancer in a DMH-treated animal. It has been shown to possess antigens that cross-react with other DMH-induced bowel adenocarcinoma isografts. The second isograft, DMH-W49, is a carcinosarcoma explanted from a DMH-treated primary colon cancer. It has intermediate antigenic cross-reactivity with other colon adenocarcinoma isografts in the WF model. The third isograft, DMH-W15, is a sarcoma explanted from a DMH-induced colon cancer that does not possess antigens cross-reactive with other DMH-induced colon adenocarcinomas. The fourth isograft, SPK, is a spontaneous (non-DMH-induced) renal cell carcinoma that is immunogenic but should not contain tissue-type-specific antigens cross-reacting with the bowel cancers. Immunized rats received three sc weekly injections of 1 X 10(3) irradiated cells. Concomitant control rats received no immunization after resection of the primary tumor. Within 24 weeks of primary tumor resection, 12 of 16 (75%) rats not immunized had tumor recurrence. Only 8 of 24 (34%) rats immunized with DMH-W163 had tumor recurrence (P less than .025 compared to controls). Fifty percent of animals (10/20) immunized with the carcinosarcoma DMH-W49 had a recurrence. Animals immunized with the non-cross-reacting DMH-W15 sarcoma isograft had a recurrence rate similar to that of controls (16/20, 75%). The rats immunized with SPK were not protected from recurrence. Twelve of 19 (63%) had a recurrence at or near the suture line within 24 weeks following primary tumor resection. These results confirm that adjuvant immunotherapy can decrease the rate of recurrence following primary tumor resection in this model. In addition, immunogens that possessed tissue-type-specific antigens were more effective in preventing tumor recurrence than those that did not.     

Alloantigen expression of a rat Moloney sarcoma.

Cells of a brown Norway (BN) rat Moloney sarcoma (MST) failed to express certain BN alloantigen specificities and bound only about 30-50% the amount of labeled alloantibody bound by normal BN spleen cells. MST cells lacked antigen specificities shared by BN and WF rats, but expressed some of the antigen shared by BN and AUG rats. Further loss of alloantigens that occurred with prolonged in vitro culture was associated with reduced virulence of MST cells for syngeneic hosts and with increased expression of tumor-associated antigens. The LEW rats, which are resistant to MST cells, might have rejected the tumor on the basis of factors other than Ag-B antigens.     

Genetic role of rat macrophage cytotoxicity against tumor.

Macrophages from the Lewis (Le) rat strain are significantly more cytotoxic to a Moloney sarcoma tumor both in vivo and in vitro, than are macrophages from the Brown Norway (BN) strain. Activity of macrophages from (Le x BN)F1 rats that are histocompatible with the Moloney sarcoma tumor is directed toward tumor and/or virus-associated antigens and is expressed as a dominant genetic trait. Experiments with backcross rats suggest that the genetic factors are unrelated to the major histocompatibility locus (AgB) of the rats. BN microphages, although not active against tumor and/or viral antigens, can become cytotoxic to cells displaying Le alloantigens.     

Inheritance and site of expression of genes controlling susceptibility to mammary cancer in an inbred rat model

An inbred rat model for genetically controlled susceptibility to chemically induced mammary cancer has been established. Wistar-Furth (W/Fu) rats were found to be more susceptible to 7,12-dimethylbenz(a)anthracene-induced mammary tumors than were Fischer (F344) rats. The susceptibilities of various F1, F2, and backcross generations of these strains were examined for susceptibility to 7,12-dimethylbenz(a)anthracene-induced mammary tumors. The data suggest that susceptibility is inherited as a dominant trait. Both a single locus autosomal model and an X-linked model have been ruled out. However, the data support the hypothesis that complete susceptibility is controlled by any one of a group of independently segregating genes; i.e., any one gene of this group is both necessary and sufficient to induce maximal susceptibility. It is not known if these genes are identical or different. In order to identify the role of these genes we asked if they were expressed in the mammary epithelial cells themselves or elsewhere in the rat. Chimeric animals were produced by transplanting mammary cells from either W/Fu or F344 rats into the white interscapular fat pad of female W/Fu X F344 F1 rats. One month after transplantation the animals were treated with 7,12-dimethylbenz(a)anthracene and then palpated weekly for tumor development at the graft site. Tumors developed more rapidly and in greater total frequency at sites grafted with W/Fu mammary cells. This result suggests that the genes controlling inherited susceptibility are expressed in the mammary cells. The role of these genes is now under investigation. We have thus far shown that they do not control carcinogen metabolism or activation.     

Genetic control of immune responses to Moloney leukemia virus in rats.

When BN and LEW rats were immunized with untreated or with inactivated Moloney murine leukemia virus (M-MuLV), BN rats produced high antibody responses to the p15, p30, and gp70 antigens of the virus, whereas LEW rats were low responders to these antigens. BN rats also exhibited a high response and LEW rats a low response when the two strains were immunized with purified p30. Studies of (LEW X BN)F1 and backcross rats suggested that factors associated with AgB exerted major influences on responses to antigens of M-MuLV and that other factors were also important. When other rat strains representing 5 AgB alleles were tested, some were high and some were low responders to M-MuLV, and responses to p15, p30, and gp70 were not always parallel. Since M-MuLV replication was greater in cells of BN rats than in cells of LEW rats, replication of M-MuLV may have influenced the levels of responses to some viral antigens. Control of virus replication appeared to be due to cell mechanisms rather than to the environment of the host.     

Gross-virus-induced lymphoma in the rat. IV. Cytotoxic cells in normal rats

Lymphoid cells from normal W/Fu rats are cytotoxic in vitro for syngeneic Gross-virus-induced lymphoma cells in a 4½ h ⁵¹Cr release test, and protect against tumour growth in vivo when adoptively transferred to syngeneic recipient rats. Cytotoxicity by normal lymphoid cells was greater with in vitro rather than in vivo passaged target cells, and was increased by preincubation of the lymphoid cells at 37° C for 3 h in vitro before their addition to the target cells. Cytotoxic activity, which was localized predominantly in the spleen, was absent at birth but increased until adulthood with some decline in older age. Histocompatibility at the major Ag-B locus was not required for the killer cell—target cell interaction; normal spleen cells of many Ag-B genotypes were cytotoxic for W/FuG-1 target cells, although BDIX and (BDIX × W/Fu)F₁ were less active and BN were inactive. The specificity of cytotoxicity was studied by the techniques of direct lysis, competitive inhibition or adsorption to cellular monolayers, using a variety of cell lines. Selectively of lysis or binding was observed and was restricted to rat target cells releasing exogenous or endogenous C-type viruses.     

Interleukin generation in experimental colon cancer of rats: effects of tumor growth and tumor therapy

The capacity of inbred W/Fu rats bearing syngeneic colon carcinomas to generate interleukin(s) (IL) was studied during primary tumor growth, after tumor resection, and during postresection immunotherapy. During local tumor growth, there was a significant decrease in the capacity of the host's adherent mononuclear cells to generate IL-1 and of peripheral blood mononuclear cells to generate IL-2 (16.6 and 23%, respectively, when compared to control animals; P less than .01). The presence of regional metastases or large primary tumor burden resulted in a further sharp fall in IL generation (0.9 and 10% for IL-1 and IL-2, respectively, when compared to control animals; P less than .01). With the use of three different doses of tumor inoculum, inhibition of IL generation was shown to occur when tumors were barely palpable. Decrease in IL correlated with tumor growth and not with the initial number of tumor cells injected. Tumor resection resulted in a rise in IL-2 generation from 36 to 64% of control animals' levels. Postresection immunotherapy with the use of an active specific immunization protocol successfully modulated IL-2 production to normal in animals protected from tumor recurrence. Animals that developed recurrent tumors despite immunization exhibited a continued inability to generate IL (mean values of IL-2 production compared to controls: 184% in animals free of recurrence after immunotherapy, 1% in animals developing recurrent tumors after immunotherapy; P less than .01). These results suggested that alterations in IL generation may lead to immune unresponsiveness during tumor growth. Active specific immunotherapy protecting animals from recurrence after primary tumor resection may be predicated on the successful modulation of IL level generation by host immunocytes.     

Specific enhancement of tumor growth and depression of cell-mediated immunity following sensitization to soluble tumor antigens.

W/Fu rats inoculated s.c. with less than or equal to 5 x 10(7) syngeneic (C58NT)D (Gross virus-positive) lymphoma tumor cells normally develop a palpable tumor which reaches its maximum size (12 to 14 mm) at 6 to 8 days and is subsequently rejected by 10 to 12 days. However, rats previously sensitized with soluble tumor antigens from (C58NT)D cells prior to (C58NT)D tumor inoculation demonstrate a significant enhancement of tumor growth (the tumor reaches up to 26 mm and is rejected by 16 to 18 days). This enhancement persisted in antigen-treated rats that continued to receive soluble antigen after tumor inoculation. The in vivo enhancement coincided with a significant in vitro depression of cell-mediated cytotoxicity [assessed with 51Cr-labeled (C58NT)D target cells and peripheral blood leukocytes]. The observed tumor enhancement was specific, inasmuch as presensitization to either soluble tumor antigens from WR6 (Gross virus-negative) tumor, syngeneic to W/Fu rats, or to soluble antigen from W/Fu spleen cells had no enhancing effect on (C58NT)D tumor growth. Interestingly, sensitization to soluble tumor antigen alone did not elicit detectable cell-mediated immunity, cytotoxic antibody, or serum-blocking activity to the (C58NT)D tumor. We conclude that sensitization to soluble tumor antigens specifically impairs the immune apparatus normally acting in tumor rejection. This impairment appears to act primarily at the induction phase of the immune response.     

Increased immunogenicity of low-antigenic rat tumors after superinfection with endogenous murine C-type virus in nude mice.

Four chemical carcinogen-induced and two polyoma-virus-induced rat tumors were repeatedly passaged through nude mice. A methylcholanthrene-induced tumor in BDIX rats (MBDB) and a polyomavirus-induced tumor in Wistar/Fu rats (PW41) became infected with endogenous mouse virus (EMV), as judged by the expression of murine C-type virus-associated gp71, p30 and p12 antigens on their cell surface. Two ethylnitrosourea-induced tumors in BDIX rats (290T and GE3A) were exposed in vitro to the supernatant of EMV-infected PW41. Subsequently, 290T but not GE3A converted to murine gp71, p30 and p12 positivity. All these successfully infected rat tumors (EMV-MBDB, EMV-PW41 and EMV-290T) became less transplantable to and more rejectable in otherwise susceptible syngeneic rats. To compare the immunogenicity of the virus-infected and non-infected tumors, syngeneic rats were immunized three times with irradiated cells, and challenged with the non-infected tumor. Wistar/Fu rats immunized with irradiated EMV-PW41 showed no improvement in PW41 rejection, compared to rats immunized with irradiated, non-infected cells. On the other hand, BDIX rats immunized with EMV-MBDB or EMV-290T rejected MBDB or 290T, respectively, with no cross-immunity, while the rats immunized with irradiated but non-infected tumors showed no significant rejection. These results indicate that EMV infection augmented the immunogenicity of non-immunogenic or only low-immunogenic rat tumors.     

Genetic association of the humoral and cellular immune responses of rats to moloney sarcomas

Brown Norway (BN), Lewis (Le), F¹ hybrids of Le × BN (LBN) and parent-to-LBN backcross rats were tested for cellular and humoral responses to a BN Moloney sarcoma. Regardless of AgB phenotype, BN backcrosses produced low levels of cell-mediated cytotoxicity (CMC) that were comparable to those of BN parents. Le backcrosses developed high levels of CMC similar to those produced in Le parents. An inverse relationship between levels of CMC and serum antibodies (cytotoxic for tumor cells and anti-p30 of MuLV) was observed; BN parents and backcrosses produced high levels of serum antibodies whereas levels in Le parents and backcrosses were low. LBN hybrids developed relatively high levels of CMC and serum antibodies. An additional finding was that the CMC response in Le parents and back-crosses was directed primarily against tumor-associated antigens rather than histocompatibility antigens expressed on the tumor cells. The results suggest that humoral and cellular responses to Moloney sarcoma in rats are not determined solely by the major AgB histocompatibility locus but do have a genetic association. This genetic association was detected with a ⁵¹Cr release assay which detects T-cells, suggesting that select populations of effector T-cells may be genetically regulated.     

Gut-associated lymphoid tissue and dimethylhydrazine-induced colorectal carcinoma in the wistar/furth rat

Although gut-associated lymphoid tissue in the form of discrete lymphoid patches (LP-GALT) in mammalian intestine is most prominent in the distal ileum, appendix, and, in some species, the cecal appendage, LP-GALT can be found throughout the intestinal tract. LP-GALT appears as single or multiple subepithelial lymphoid follicles covered by a specialized, structurally unique epithelium. In the colon of the Wistar/Furth(W/Fu) rat, LP-GALT appears as aggregates of follicles, or lymphoid patches, that can be detected macroscopically. We studied the relationship between 1,2-dimethylhydrazine- (DMH) induced colon carcinomas and the lymphoid patch associated epithelium in these animals. In addition, we defined the normal distribution of colonic lymphoid patches in both DMH-treated and control rats. Patches were found macroscopically and confirmed by histologic examination at five constant sites: lower pole of cecum, proximal ascending colon, the major colonic flexure, mid descending colon, and the rectosigmoid. These are also the predominant sites of DMH induced carcinomas in W/Fu rats. In 120 DMH-treated animals, 109 colon carcinomas were found. Eight percent were in the lower pole of the cecum, 56% in the proximal ascending colon, 16% at the major flexure, 15% in the mid descending colon, and 5% in the rectosigmoid. Lymphoid patches could often be detected histologically in association with DMH-induced tumors. The depth of tumor invasion was found to correlate inversely with our ability to identify tumor-associated lymphoid patches suggesting that tumors arising at the anatomical sites where lymphoid patches occur progressively destroyed them. Of colon tumors confirmed histologically to be associated with lymphoid patches, 88% were superficial lesions confined to the submucosa and 12% were more extensive but confined to the bowel wall. No lymphoid patches could be found associated with tumors that extended through the bowel wall. Thus, DMH-induced colon carcinomas in W/Fu rats arise at sites containing preexisting LP-GALT with associated specialized epithelium.     

Reduced tumorigenicity of murine tumor cells secreting gamma-interferon is due to nonspecific host responses and is unrelated to class I major histocompatibility complex expression

Spontaneous SP1 murine adenocarcinoma cells transfected with the murine gamma-interferon (IFN-gamma) gene expressed IFN-gamma (SP1/IFN-gamma) failed to grow in syngeneic hosts and grew in nude mice. The rejection of SP1/IFN-gamma cells was related to the amount of IFN-gamma produced and appeared to be mediated primarily by nonspecific cellular mechanisms, although some role for T-cells in the afferent arm of this response is possible. SP1 cells are H2-Kk negative but express class I antigens when producing IFN-gamma. However, class I major histocompatibility complex (MHC) expression, while likely necessary, was insufficient in itself to prevent tumor growth since secretion of greater than 64 units/ml IFN-gamma was needed to inhibit tumorigenicity while only 8 units/ml IFN-gamma could induce class I antigens. Similar results were obtained with the murine colon carcinoma CT-26, a tumor that constitutively expresses class I MHC antigens, further supporting the contention that class I MHC expression is not essential for the rejection response induced by IFN-gamma. The failure of SP1/IFN-gamma cells to protect against a challenge with parent SP1 cells argues that factors other than IFN-gamma production or class I MHC expression are needed to induce a protective response against weakly or nonimmunogenic tumor cells.     

Induction of transplantation immunity to rat colon carcinoma isografts by implantation of intact fetal colon tissue

The objective of this study was to demonstrate whether intact fetal colon tissue can induce immunity to large-bowel cancer isografts. Adult rats were immunized against fetal antigens by implantation of mitomycin-C-treated intact fetal colon tissue beneath the renal capsule. Control animals received implants of pieces of intact adult colon tissue. Kidneys containing the implants were removed after 3 or 5 weeks and the rats were challenged with isografts of a 1,2-dimethylhydrazin (DMH)-induced colon carcinoma. Significant inhibition of tumor growth was observed in immunized groups compared to controls. The present results indicate that some embryo-fetal antigens expressed in intact colon tissue of 13 to 15-day-old fetuses are also expressed in colon carcinoma cells and can act as transplantation antigens.     

The zinc glycinate marker in human colon carcinoma.

In an attempt to identify new tumor markers in human colon carcinoma, we produced antisera in rabbits tolerant to normal human tissue antigens and immunized with zinc glycinate-treated extracts of liver metastases from a colon carcinoma. These antisera reacted with carcinoembryonic antigen and with an additional component present in the tumor extracts but not detected in the extracts of normal tissues. The new component, the zinc glycinate marker (ZGM), had an alpha2 mobility on immunoelectrophoresis, was soluble in 1 M perchloric acid, and had a molecular weight of approximately 2X10(6), as indicated by its elution pattern on Sepharose 6-B chromatography. It differed from alpha fetoprotein, nonspecific cross-reacting antigens (NCA, NGP, or CCA III), ferritin-like molecules, and blood group substances A, B, H, Lewis a, and Lewis b. The ZGM was similarly identified in saline or zinc glycinate extracts of 11 of 23 carcinomas of the colon. With routine hematoxylineosin staining, no histologic differences were apparent between tumors bearing the antigen and those without it.     

Regression of rat primary mammary tumors following dietary d-limonene

The effects of d-limonene on rat mammary tumors were investigated. Following the appearance of mammary tumors induced by 7,12-dimethylbenz[a]anthracene in female (W/Fu X F344)F2 rats, animals were assigned to pairs in which 1 rat was fed a diet containing 10% d-limonene and the other was pair fed an isocaloric diet containing 10% cellulose. There was a highly significant increase in the regression of the first tumors in the rats fed d-limonene. In addition, d-limonene given at this time inhibited the formation of subsequent multiple tumors in these rats.     

Spleen cells of whole body X-irradiated W/Fu rats enhance tumor growth in vivo and non-specific cytotoxicity in vitro

This study was designed to investigate the influence of spleen cells of normal Wistar/Furth (W/Fu) rats obtained after whole body X-irradiation (WBI) upon mammary carcinoma growth in vivo, and cell mediated cytotoxicity against several target cells in vitro. The ME/H mammary carcinoma employed here originally arose spontaneously in a W/Fu rat, metastasizing to the retroperitoneal lymph node and lungs. It was found that surviving non-adherent spleen cells taken two days after 500R WBI cause enhanced tumor growth and metastases development in a Winn assay compared with nonadherent spleen cells from unirradiated controls. These cells were also enriched in granulocytes compared with controls. While the level of nonspecific cell mediated cytotoxicity was variable, it increased significantly following WBI of the spleen cell donor. Our results indicate that there are apparently two opposing effects shown by non-adherent spleen cells surviving WBI of normal W/Fu rats: enhancement of in vivo tumor growth, and enhancement of in vitro cell mediated cytotoxicity. A possible mechanism to explain these contrasting results is suggested.     

Membrane immunofluorescence studies on cells producing rat C-type virus particles

The WF-l rat cell line was investigated for the presence of virus-associated cell membrane antigens similar to those found in the membranes of cells infected with murine leukemia viruses. These cells transform spontaneously in tissue culture from embryos derived from the Wistar-Furth ( W/Fu) strain of rats and produce large amounts of rat C-type virus particles. When sera from W/Fu rats bearing WF-l tumors were tested for antibodies directed against the WF-l cells by indirect membrane fluorescence, it was found that 20–35% of the cells reacted strongly. The sera did not react with normal W/Fu embryonic fibroblasts nor did absorption of a pooled positive serum with these cells remove antibodies reactive with the WF-l cell line. Interestingly, however, these sera also reacted with two other virus-positive cell lines (R-35 and RMTL-8) derived from mammary tumors originating in Sprague-Dawley rats but not with virus-negative cell lines. Absorption experiments verified that the R-35 and RMTL-8 cells contained cross-reacting antigens with the WF-l cells. These results suggested that (a) virus-associated antigens were expressed in the membranes of cells infected with rat C-type particles and (b) C-type virus particles found in cell lines originating from different Sprague-Dawley rats as well as from the Wistar-Furth strain were antigenically closely related if not identical.     

Role of T-cell subsets in the destruction of established metastases of 13762A rat mammary adenocarcinoma

The 13762A rat mammary adenocarcinoma metastasizes with high frequency to regional lymph nodes and lungs. The intratumoral injection of Corynebacterium parvum on day 7 followed by primary tumor excision on day 20 significantly prolonged survival and cured 10-40% of syngeneic F344 rats. Established metastases were destroyed by the treatment, and strong and specific tumor rejection immunity was induced. The purpose of the present study was to determine if T-cells were required for the C. parvum treatment to be effective and to identify the subsets of T-lymphocytes that might participate in the response. The results indicated that rats depleted by either neonatal thymectomy or a combination of adult thymectomy, 900 rad, and bone marrow reconstitution did not inhibit tumor growth after C. parvum treatment. Restoration of depleted rats with lymph node cells permitted effective treatment. The lymph node cells that were responsible for restoration expressed both W3/13 (pan-T-cell) and W3/25 (helper T-cell) membrane-associated differentiation antigens. T-cells that bore the MRC OX8 (cytotoxic-suppressor T-cell) antigen did not restore the response to C. parvum treatment. The effect of lymph node restoration was markedly potentiated by simultaneous administration of thymocytes, a T-cell population that expresses both W3/25 and MRC OX8 antigens. In conclusion, the cytotoxic-suppressor T-cells were ineffective in the restoration of T-cell-depleted, tumor-bearing rats to benefit from C. parvum but helper T-cells were highly effective, and their activity was strongly potentiated by administration of thymocyte amplifier cells.     

Prevention of peritoneal carcinomatosis recurrence with a prostaglandin synthesis inhibitor, indomethacin

Carcinomas produce large amounts of prostaglandin (PG) E2, which play an important role in suppression of non-specific cellular immune reaction in tumor-bearing individuals. PG synthesis inhibitor can restore the immune activity against tumors. The anti-tumor activity of indomethacin was investigated in CDF1 mice (BALB/c X DBA/2) implanted intraperitoneally with mouse colon adenocarcinoma 26 (5 X 10(5) or 2 X 10(5) cells) in a model study to prevent peritoneal recurrence after surgery for gastrointestinal cancers. Oral administration of indomethacin (0.002% water solution as drinking water) depressed and inhibited the disseminated tumor growth in the abdominal cavity, and prolonged the survival time, resulting in 30-50% cures of mice. The treatment combined with a small intraperitoneal dose of Picibanil (OK-432) (0.5 mg/kg twice weekly), which activates macrophages in the abdominal cavity, cured 90% of mice. An intraperitoneal dose of 16,16-dimethyl-PGE2 (5 micrograms/mouse, daily) reduced the anti-tumor activity of indomethacin. The results suggest that indomethacin treatment relieved the endogenous(tumor cell- and macrophage-produced) PGE2-mediated immunosuppression. It is postulated that PG-synthesis inhibitor in combination with chemotherapeutic agents, immunotherapeutic agents and low dose radiation, may provide a good therapeutic tool to prevent the development of peritoneal carcinomatosis, particularly in the cases having a small number of residual cancer cells or micrometastases in the abdominal cavity after surgery.     

Serum requirements for in vivo modulation of thymus-leukemia antigens on mouse leukemia cells and thymocytes.

Mouse leukemia cells and normal thymocytes bearing thymus-leukemia (TL) cell surface antigens were previously shown to acquire resistance to lysis by guinea pig complement (C) during incubation with TL alloantiserum in vitro at 37 degrees C due to heat-labile serum activity resulting in deposition of mouse C3 onto the cell surface. The role of heat-labile serum activity and C3 in modulation of TL+ cells in vivo in mice actively or passively immunized against TL antigens was investigated. Mice of the TL-/TL+ C57BL/6J (B6) strain and the B6 congenic strain B6-Tiaa possessed poorly modulating sera, and the radiation-induced A-strain leukemia RADA1 transplanted into B6 mice passively immunized with heated (56 degrees C) TL antiserum failed to modulate; thymocytes of B6-Tiaa mice immunized similarly also did not modulate. A specific requirement for mouse C3 deposition onto RADA1 cells to achieve a modulated state was demonstrated in actively immunized TL- (B6 X A-Tiab)F1 mice in which circulating C3 and modulating activity were depleted by administration of cobra venom factor. In immunized (B6 X A-Tiab)F1 mice bearing RADA1 transplants and repeatedly given injections of B6 serum, tumor cells escaped immune destruction despite a lack of modulation. Thus modulation of TL antigenicity on tumor cells in vivo, but not tumor escape, required cell-bound C3.