Human breast carcinoma patients develop clonable oncofetal antigen-specific effector and regulatory T lymphocytes.

Oncofetal Ag (OFA) is a 44-kDa glycoprotein expressed during early to mid-gestation fetal development and re-expressed as a surface Ag by tumor cells soon after transformation. The Ag is detectable on all types of human and rodent tumors tested, but is undetectable on normal cells. In experimental animals it is autoimmunogenic and induces potentially protective T cell responses both after experimental immunization and during tumor development subsequent to carcinogenic insult. To determine whether this tumor-associated Ag is also immunogenic for human T lymphocytes, breast carcinoma patients' peripheral blood mononuclear leucocytes were stimulated in vitro with autologous tumor cells in the presence of IL-2, gamma-IFN, and IL-6 for 2 wk. The tumor-reactive cells were then restimulated and cloned by limiting dilution, and the clones were analyzed. We established 24, 19, 11, and 16 tumor-reactive clones from the four respective patients. Of those, 4, 6, 4, and 7, respectively, proliferated specifically to purified OFA. Both CD4 and CD8 OFA-specific clones were established, which responded equally well to purified OFA or 32- to 44-kDa immature laminin receptor protein. All were CD3+, TCR-alpha beta+. All CD4 clones secreted gamma-IFN, but neither secreted IL-4 nor IL-10. Both IFN-gamma-secreting cytotoxic CD8 clones and IL-10-secreting inhibitory CD8 clones were established. Thus, during human cancer development, the same types of OFA-specific effector and regulatory T cells are induced as during murine T lymphomagenesis.

37 kiloDalton oncofetal antigen protein and immature laminin receptor protein are identical, universal T-cell inducing immunogens on primary rodent and human cancers.

Based upon positive immunologic comparisons, protein and cDNA sequencing, and in vitro and in vivo immunogenicity studies we propose that carcinomas, sarcomas and lymphomas/leukemias of rodents and humans share 37 kDa Onco-Fetal Antigen [OFA] as a T and B-lymphocyte stimulating, universal tumor specific transplantation antigen [UTSTA]. In the past four years, biochemical studies from several laboratories studying laminin receptor protein and immunological studies of OFA from our laboratories independently converged. OFA protein and immature or precursor Laminin Receptor Protein [iLRP] are > 99% identical proteins based on amino acid and cDNA sequencing and immunobiology studies summarized here. Acquired expression of 37 kDa OFA/iLRP enables malignant tumor cells to penetrate laminin tissue and vessel barriers. 37 kDa OFA/iLRP activates precursor thymic anti-OFA/iLRP specific cytotoxic T cell which kill emerging pretumor cells. Our reported findings also demonstrate that OFA/iLRP can function to induce specific immunoregulatory CD8 T-suppressor cells secreting IL-10 which impair effector T-cell killing of emerging tumor cells non-specifically and thereby facilitate tumor progression. Potential applications of OFA/iLRP detection in early cancer formation, for monitoring patient T cell subclass responses to OFA/iLRP as a predictor of tumor progression, and the use of OFA/iLRP peptides for specific anti-tumor immunotherapy are presented.

A new immunobiological view of radiation-promoted lymphomagenesis.

Whole-body irradiation produces T-cell leukaemias/ lymphomas (TCL) in some strains of inbred mice in an X-ray dose-related manner. Radiation biologists have related the rapid "initiation' and early appearance of preleukaemic cells in these mice to unrepaired DNA damage inflicted by radiation. Following initiation, radiation-altered thymic differentiation fosters multi-step transformation changes in proto-oncogenes and suppressor gene expression in individual clones of non-invasive preleukaemia cells as they progress to malignancy. The malignant clones arising from small numbers of initiated preleukaemia thymocytes become fully transformed only after several more months to a year after irradiation in those strains of mice which develop T-cell lymphomas. When the RFM mouse was subjected to sublethal whole-body X-ray, only 50% of the mice developed TCL by 6 months, yet nearly all developed preleukaemia thymocytes. The T-cell-mediated immune response of the irradiated host has never been substantiated to contribute to malignant TCL development. Until recently, X-ray-induced TCL were not known to carry common tumour rejection antigens TATA. However, several studies have revealed that both preleukaemia cells and fully malignant TCL express an immunogenic, common oncofoetal glycoprotein, termed 44kD OFA. OFA-activated memory CD4 Tm and CD8 Ten. T-effector cells in irradiated mice expressing OFA. As most irradiated RFM mice exhibit preleukaemia thymocytes yet only half develop tumours, this finding implicates active host T-cell effector responses in X-ray-initiated tumorigenesis. Further, the recent discovery of OFA-specific CD8 Ts clones in irradiated mice, which inhibited cytotoxicity of CD8 clones to OFA or TSTA, may explain which mice develop T-cell lymphomas.

CD8 T cell clones inhibit antitumor T cell function by secreting IL-10.

We have reported that in irradiated, long-term surviving RFM strain of mice there is enhanced kinetics of tumor development upon challenge with RFM lymphoma cells. We reported that we cloned splenic oncofetal (OFA)-specific, noncytotoxic CD8+ T cells from such mice. These noncytotoxic CD8+ T cell clones secrete a factor upon Ag stimulation that inhibits the ability of OFA-specific RFM cytotoxic T (TC) cell clones from killing 5T RFM lymphoma cells in vitro. These supernatants do not inhibit the tumor cell-induced proliferation of the TC cell clones however. We report here that OFA-stimulated, RFM-noncytotoxic CD8 T cell clone culture supernatants also inhibit IFN-gamma-secretion by stimulated CD4 and CD8 RFM anti-OFA effector T cell clones in a dose-dependent manner. The inhibitor in those culture supernatants acts in neither an Ag-specific nor MHC-restricted manner. We find that the culture supernatants of OFA-stimulated, noncytotoxic CD8 T cell clones contain IL-10, while those from OFA-stimulated, RFM OFA-specific TC cell clones do not. We show that monoclonal anti-IL-10 Ab specifically blocks the inhibition of cytotoxic activity and IFN-gamma secretion by OFA-specific CD8 and CD4 effector T cell clones in a dose-dependent manner in vitro. Incorporation of anti-IL-10 Ab into the cytotoxicity assays of the OFA-specific, noncytotoxic CD8+ T cell clones against 5T tumor cells restores their cytotoxic activity. This may suggest that one way of inducing anergic T cells is by induction of IL-10 secretion.

Characterization of RFM mouse T lymphocyte anti-oncofetal antigen immunity in apparent tumor-free, long-term survivors of sublethal X-irradiation by limiting dilution T lymphocyte cloning.

Fractionated sublethal x-irradiation induces thymic lymphomas in up to 60% of RFM mice by 6 mo after irradiation, but no thymomas arise after 6 mo. All radiation-induced tumors expressed oncofetal Ag (OFA) and thymic OFA expression significantly preceded detectable thymomas. To determine whether lymphoma-free, radiation survivor mice had anti-OFA T cell immunity, we analyzed their clonable 5T lymphoma-reactive T lymphocytes and determined the resistance of such mice to challenge with RFM lymphoma cells. RFM mice that were irradiated 6 to 6.5 mo earlier, but which had no apparent tumors, had no more resistance to OFA+ 5T lymphoma cell challenge than nonirradiated age-matched mice. These mice actually developed maximal tumor incidence significantly faster. Analyses of 5T lymphoma-reactive T cell clones established from RFM mice 11 mo after irradiation, however, suggest that activation of anti-OFA immunity occurred subsequent to irradiation. Previously irradiated RFM mice yielded 257% more tumor-reactive T cell clones/mouse than non-irradiated controls. All clones from both sets of mice specifically proliferated to purified OFA. Each irradiated mouse yielded some clones that proliferated to fivefold lower doses of OFA than did any T cell clones from nonirradiated mice. Only these low-dose OFA-reactive clones responded equally to irradiated 4T and 5T cells by proliferation, IFN-gamma secretion, and target lymphoma cell killing. 4T cells express less OFA/cell than 5T cells. Some noncytotoxic CD8 T cells that inhibited cytotoxic T cell function were cloned only from irradiated RFM mice.

Identification of partial complementary DNA clones encoding a 59-kd protein with characteristics of a unique oncofetal antigen.

BACKGROUND: Oncofetal antigens (OFAs) are conserved tumor-associated autoantigens or transplantation antigens present on the surface of all major classes of rodent and human tumors and on midgestational fetal cells but not on normal neonatal or adult human and rodent tissues. A syngeneically derived monoclonal antibody, MAb-115, recognizes murine OFAs of 44 and 200 kd in molecular mass. PURPOSE: Our goal was to clone and characterize the complementary DNAs (cDNAs) that encode these murine OFAs. METHODS: Rabbit antiserum raised against purified 44-kd OFA glycoprotein was used to screen a mouse embryo cDNA-lambda phage expression library. Recombinant phage clones positive for the expression of OFAs were detected by immunohistochemical staining, then isolated and plaque purified. The presence of an OFA-encoding sequence in the recombinant phage was confirmed by specific reaction of the expressed protein with MAb-115. Recombinant fusion protein was purified from the extracts of corresponding lysogens. Rabbit antiserum against purified recombinant fusion protein was raised, and the capacity of this antiserum to detect the expression of OFA on rodent tumor and fetal cells was determined by flow cytometry. In addition, immunoreactivity of tumor bearer and hyperimmune murine sera to bacterially expressed recombinant OFA protein was evaluated by enzyme-linked immunosorbent assay. The OFA-expressing insert DNA from plaque-purified lambda clones was subcloned into phagemid vectors for sequencing analysis. RESULTS: Antiserum derived against the isolated recombinant mouse embryo polypeptide mimicked MAb-115 in its specific binding to all OFA-positive rodent tumor and fetal cell lines tested and likewise did not show reactivity to normal adult tissues. This antiserum specifically recognized the native 44- and 200-kd OFAs in extracts of murine lymphocytic lymphoma. Furthermore, sera of tumor-bearing mice or mice immunized with purified OFA or intact, irradiated OFA-positive lymphocytic lymphoma cells also reacted with the recombinant fusion protein. The characterization of the isolated clone included nucleotide sequence information followed by analysis of the deduced primary structure of the protein. CONCLUSIONS: These data suggest that the isolated cDNA clones encode a distinct gene product which is widely expressed on the surface of tumor and fetal cells and represents the first characterized sequence of a true OFA. IMPLICATIONS: The availability of this cDNA, encoding a protein expressed only on tumor and fetal cells, provides a direct means to assess biological characteristics of malignant tissue which can be assayed by biochemical, histochemical, and molecular methods.

Increased oxygen tensions influence subset composition of the cellular immune system in aged mice.

In acute and chronic experiments, each of eight groups of aged mice were assigned separately to different pressures of oxygen to which it was to be exposed. Lymphocytes from spleen, thymus, and peripheral blood were analyzed following oxygen exposure. Subset populations changed depending on the oxygen tension. Variable changes were observed in total numbers of lymphocytes, lymphocyte subsets, B cells, and macrophages depending on the organ studied and the oxygen pressure to which the mice were exposed. There were differences between acute and chronic exposure suggestive of adaptation to environmental stressors. The suggestion is made that the immune system has a reserve capacity that can be influenced by oxygen and, thereby, theoretically capable of being pharmacologically manipulated to assist patients with altered immune systems to promote defense mechanisms or, under certain circumstances, reduce autoimmunity. It is hypothesized that an underlying hypoxia may be involved in the age-associated decline in the immune system.

Differential recognition of murine tumor-associated oncofetal transplantation antigen and individually specific tumor transplantation antigens by syngeneic cloned BALB/c and RFM mouse T cells.

We have previously demonstrated in several species that sarcomas, lymphomas, and carcinomas express a common Ag that cross-reacts with midgestation fetal cells. We also produced a mAb to that protein and characterized it as a 44-kDa glycoprotein. The cross-reactive immunity induced by immunization with tumor or fetal cells expressing the oncofetal Ag (OFA) can be adoptively transferred with cell populations containing T lymphocytes. The experiments discussed within this paper describe the establishment and characterization of two types of T lymphocytes induced by immunization with syngeneic tumor cells in two mouse strains. We find that five of the eight cloned T cells derived from spleens of BALB/c mice that had been immunized with MCA1315 fibrosarcoma cells are specific for an Ag shared by MCA1315 and MCA1321 cells. The other three clones are specific for an Ag present on MCA1315 but not on MCA1321. Also, none of the clones were reactive with the BALB/c plasmacytoma MOPC-315, which does not express OFA. We also find that 75% of the RFM T cell clones from spleens of RFM mice immune to the RFM thymoma 5T show a 5T-specific proliferative response. One of the four clones, however, responds to both 4T and 5T RFM thymoma cells. The BALB/c and RFM cross-reactive clones specifically respond to purified 44-kDa OFA derived from MCA1315 fibrosarcoma cells in the presence of syngeneic irradiated spleen cells and IL-2. All of the clones from both strains of mice, be they tumor-specific transplantation Ag specific or OFA specific, are CD4+, CD3+, alpha beta TCR+ T cells that secrete IFN-gamma on Ag stimulation.

44-kd oncofetal transplantation antigen in rodent and human fetal cells. Implications of recrudescence in human and rodent cancers.

OBJECTIVE: This article summarizes the phase-specific nature of a cell surface, 44-kd tumor-associated transplantation antigen glycoprotein expressed during early and middle gestation in a portion of rodent and human fetal cells during normal fetal tissue development and illustrates how this glycoprotein is consistently recrudesced in primary and established human squamous cell carcinomas and other human and rodent tumors. The oncofetal antigen was not detectable in any human or rodent term fetal tissue or normal adult tissues tested. The tumor-associated transplantation antigen was tumor specific, yet not germ-line specific (expressed in lymphomas, sarcomas, and carcinomas) in human or rodent cancers. Rodent model tumor studies have shown 44-kd oncofetal antigen can act as a tumor-associated autoantigen of potential use in cancer detection and therapy. DESIGN: The oncofetal antigen was detected by immunogenicity, flow cytometry, and Western blotting in syngeneic rodent tumor recipients and by the last two methods in humans with progressive cancer. Syngeneically derived mouse monoclonal antibody (MoAb 115) was used to identify 44-kd oncofetal antigen. Early to middle gestation, oncofetal antigen-positive, mouse embryo/fetal cells used to stimulate the hybridoma were tested for immunogenicity as a tumor-associated transplantation antigen in syngeneic hosts. SETTING AND PATIENTS: Patients presenting with head and neck squamous cell carcinoma (N = 25) and other carcinomas at the University of South Alabama Medical Center, Mobile, underwent a biopsy, and the tumors were mechanically dispersed and were then tested for oncofetal antigen expression directly in flow cytometry. The tumors were also cultured and tested as squamous carcinoma cell lines. Growing squamous carcinoma cells and uncultured tumor cells were stained with MoAb 115 or control MoAb. Extracts of the cells were banded by electrophoresis in gels, Western blotted, and reacted with MoAbs and enzyme-linked immunosorbent assay second antibody. Time-mated mouse fetus and human fetal cells were also stained with MoAb 115 or control antibody and analyzed in the flow cytometer. RESULTS: Eight- to 13-day mouse fetal cells conferred protection against syngeneic tumor challenge. Term 18- to 21-day fetal or neonate or adult mouse cells were nonprotective. All head and neck squamous cell carcinomas tested expressed 44-kd oncofetal antigen by flow cytometric analysis and in Western blots as did ATCC cell lines of these tumors, whereas normal control tissues were negative. Second trimester human fetal cells were 44-kd oncofetal antigen positive. A large spectrum of rodent sarcomas and lymphomas express the OFA. CONCLUSIONS: Shared 44-kd oncofetal antigen OFA offers promise as a tumor detection marker in human squamous cell carcinoma and other human carcinoma development, and syngeneic mouse tumors are good model systems to explore oncofetal antigen antigenicity.

Purification and partial characterization of 200 kDa oncofetal antigen from radiation induced murine lymphocytic lymphoma.

1. A 200 kDa glycoprotein (gp200) oncofetal antigen was purified from solubilized membranes of a radiation-induced murine lymphocytic lymphoma cell line (XR11-5T), grown in syngeneic RFM mice, by successive gel chromatography of the active fraction on lentil lectin agarose, Q- and S-Sepharose and Superose-12 using an FPLC system. 2. A murine monoclonal antibody 115, produced by the syngeneic immunization of adult male C57BL/6N mice with 12-day mouse fetal cells, was used in a slot blot antibody assay to follow up the active fractions. 3. The purified glycoprotein has a pI of 5.4. 4. Treatment of radiolabeled gp200 with neuraminidase caused a slight reduction in size due to the removal of sialic acid groups and a shift in pI to 6.3. 5. Treatment of gp200 with different glycosidases shows that gp200 is susceptible to N- and O-glycanase but not to endoglycosaminidase H. 6. On extraction of gp200 with Triton X-114 it partitions exclusively into the detergent-rich fraction consistent with being an integral membrane protein.

Increased oxygen tensions modulate the cellular composition of the adaptive immune system in BALB/c mice.

In acute and chronic experiments, each of eight groups of young mice were assigned separately to different pressures of oxygen to which it was to be exposed. Lymphocytes from spleen, thymus, and peripheral blood were analyzed following oxygen exposure. Subset populations changed depending on the oxygen tension. Blood lymphocyte populations reflected lymphocyte changes in thymus or spleen. Thus, a full understanding of the pharmacological effects of hyperbaric oxygen, requires a knowledge of simultaneous effects of increased oxygen pressures on the various compartments comprising the immune system.

Characterization of oligosaccharides on methylcholanthrene-induced murine oncofetal antigen using biotinylated lectins.

Biotinylated lectins were used to characterize oligosaccharide chains of the minute amounts of the 44 kD membrane-bound oncofetal antigen (OFA) derived from murine MCA 1315 tumor cells. Lectins of known sugar specificity were used to detect oligosaccharides on glycoprotein which had been previously separated electrophoretically on sodium dodecyl sulfate polyacrylamide gels (SDS-PAGE) and blotted onto nitrocellulose (NC). Oligosaccharide reactivity with biotinlylated lectins was also analyzed after glycosidase treatment of the NC blotted OFA. Results indicate MCA 1315 derived OFA possesses N-glycosylated oligosaccharide chains which are high in mannose. This result is in agreement with the high mannose glycosylation of OFA derived from 13-day mouse fetuses.

Expression of 44-kilodalton oncofetal antigen as a premalignancy marker in X irradiation-induced murine T-cell lymphoma.

BACKGROUND: Oncofetal antigens (OFAs) are found on the surface of murine and human midgestation fetal cells, in human and rodent tumor tissues, and on human and rodent tumor and embryonic cell lines but not in normal neonatal or adult human and rodent tissue. PURPOSE: The usefulness of OFA as an early indicator of lymphoma development was evaluated. METHODS: With the use of monoclonal antibody directed against a 44-kd glycoprotein, cells from the thymus and spleen of RFM/UnCr mice receiving whole-body, split-dose x irradiation (1.75 Gy once a week for 4 weeks) or cells from these organs from control (nonirradiated) mice were analyzed for the presence of OFA in the flow cytometer and in limited intrathymic transplant. RESULTS: OFA was detected on thymocytes from 75% of irradiated mice by 2 months after treatment, reflecting eventual lymphoma development in the irradiated controls by flow cytometry and in intrathymic transplant. In general, the number of thymuses expressing OFA and the percentage of OFA+ cells increased with time after irradiation. By 4 months, OFA+ splenocytes were present, but only in mice possessing OFA+ thymocytes. Serially tested, irradiated RFM mice that never expressed OFA in the thymus reflected the percentage of irradiated RFM/UnCr mice that never developed lymphomas. This observation was also made in irradiated C57BL/6N mice, which attests to the tumor specificity of OFA expression. Spleen immunoglobulin-positive cells were decreased, while CD4+ and CD8+ cells were greatly increased. Indirect evidence of CD4/CD8 expression on OFA+ splenocytes suggests that the newly forming lymphomas were of immature T-cell origin. Major histocompatibility antigen expression did not vary significantly. Histopathologic examination revealed radiation-induced lymphomas in OFA-positive tissues characterized by a monomorphic population of large blastic immature lymphoid cells. CONCLUSION: The early expression of OFA in radiation-induced oncogenesis was established. IMPLICATIONS: OFA expression significantly preceded clear histologic evidence of malignant T cells or clinical lymphoma in irradiated RFM/UnCr mice that went on to develop T-cell lymphomas.

Disseminated histoplasmosis in AIDS patients in Maryland.

These patients demonstrate the difficulty in arriving at the diagnosis of disseminated histoplasmosis. The diagnosis in two of the three patients also served as the initial AIDS case-defining opportunistic infection. In each of these patients, the clinical presentations were atypical and in only one patient was a positive exposure history elicited. Recurrent bowel obstruction was the presenting complaint in the first patient and the diagnosis was made only on pathologic exam of the resected small bowel. The second patient's diagnosis was made on biopsy of the colon via colonoscopy. The third patient's diagnosis also eluded an extensive FUO workup; he was diagnosed by bone marrow culture and silver stain of a mediastinal lymph node biopsy, despite serial negative serologic tests for histoplasmosis. The first two patients had significant gastrointestinal disease which is a relatively unusual manifestation for disseminated histoplasmosis. The third patient illustrates the limited diagnostic usefulness of serologic testing in AIDS patients and the continued usefulness of bone marrow analysis in an FUO evaluation. In conclusion, these case presentations demonstrate that disseminated histoplasmosis in patients with HIV infection can present with unusual manifestations, outside of the typical endemic arca, without a positive exposure history or positive serologic test, and may be the initial AIDS case-defining opportunistic infection in these patients. Consequently, a disseminated histoplasmosis should be considered in all AIDS patients with perplexing clinical presentations.

Isolation and partial characterization of a soluble oncofetal antigen from murine and human amniotic fluids.

A soluble (cell-free) oncofetal antigen (OFA) was detected in murine and human amniotic fluids by immunostaining with the murine monoclonal antibody (MAb 115) produced by syngeneic immunization with mid-gestational mouse fetal cells. OFA was purified from the amniotic fluids by ammonium sulfate precipitation at 30-70% saturation, followed by successive gel chromatography of the OFA-containing fraction on Sephacryl-S300 HR, Q- and S-Sepharoses and lentil lectin agarose. The fraction eluted from the lentil lectin column gave a single band on SDS-PAGE of the same molecular weight as the membrane-bound OFA found on both fetal and tumor tissues of humans and several rodents. Both soluble and membrane-bound OFAs share several chemical characteristics, including binding to lentil lectin and wheat-germ agglutinin, molecular weight (44 kDa) and pI (6.8). Mild periodate oxidation of OFA did not affect its binding to MAb 115 in an enzyme-linked immunosorbent assay, indicating that the reactive epitope is a peptide.

Differential expression of class I major histocompatibility complex determinants by lymphoblastic leukemia-lymphoma cell lines.

Two T-cell lines (XR11-4T and XR11-5T), established from radiation-induced, murine lymphoblastic lymphomas, were examined for the expression of class I major histocompatibility complex antigen and tumor induction. These cell lines expressed class I private determinants, H-2.9 and H-2.26, but not the monomorphic determinant defined by monoclonal antibody M1/42. Both cell lines produced tumors in syngeneic and allogeneic hosts. The monomorphic determinant could be demonstrated on both cell lines following growth in allogeneic (BALB/c mice) but not in syngeneic (RFM mice) hosts. The re-expressed determinant present on cells following growth in allogeneic mice was not of host origin. Thus, tumorigenic x irradiation may differentially affect the expression of class I major histocompatibility complex determinants.

Immunogenicity of a soluble partially purified oncofetal antigen from murine fibrosarcoma in syngeneic mice.

A tumor/fetal associated antigen, termed oncofetal antigen (OFA), conserved in the tumor and fetal tissue of rodents and humans, was extracted from murine fibrosarcoma cells and tumors and was fractionated on an Ultrogel AcA34 gel filtration column. A monoclonal antibody 115 specific for the OFA identified three peaks of antigenic activity in the eluted fractions, designated fractions (Fr.) I, II, and III, in the molecular weight range of greater than 160, 90, and 44 kDa, respectively. Most of the activity resided in the high molecular weight fraction, Fr. I. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis and gel scanning of Fr. I showed multiple bands, one of which, constituting about 5.6% of the total protein bands in Fr. I, was the 44 kDa oncofetal antigen, apparently present in this fraction in a soluble complex form. Lectin binding studies and isoelectric focusing showed that the 44 kDa OFA is a glycoprotein, whose pI is 6.8. Spleen and peritoneal exudate cells of BALB/c mice immunized with Fr. I protected naive syngeneic mice, in adoptive transfer experiments, from developing tumors when challenged with syngeneic fibrosarcoma tumor cells, MCA-1315. Also, immune spleen cells were cytotoxic to the tumor target cells, MCA-1315, in a 51Cr release assay at several different effector to target cell ratios. This is the first description of a conserved, true, oncofetal antigen capable of inducing tumor transplantation resistance in syngeneic rodents in a semipurified form.