The human MAGE-3 gene is expressed in many tumors of several histological types but it is silent in normal tissues, with the exception of testis. Antigens encoded by MAGE-3 may, therefore, be useful targets for specific anti-tumor immunization of cancer patients. We reported previously that MAGE-3 codes for an antigenic peptide recognized on a melanoma cell line by autologous cytolytic T lymphocytes (CTL) restricted by HLA-A1. Here we report that the MAGE-3 gene also codes for another antigenic peptide that is recognized by CTL restricted by HLA-A2. MAGE-3 peptides bearing consensus anchor residues for HLA-A2 were synthesized and tested for binding. T lymphocytes from normal individuals were stimulated with autologous irradiated lymphoblasts pulsed with each of three peptides that showed strong binding to HLA-A2. Peptide FLWGPRALV was able to induce CTL. We obtained CTL clones that recognized not only HLA-A2 cells pulsed with this peptide but also HLA-A2 tumor cell lines expressing the MAGE-3 gene. The proportion of melanoma tumors expressing this antigen should be approximately 32% in Caucasian populations, since 49% of individuals carry the HLA-A2 allele and 65% of melanomas express MAGE-3.

Although murine tumor cells have been transfected to express a multitude of different cytokines and shown to be rejected in vivo, it is unclear which of these factors might be useful to facilitate tumor Ag immunization schemes. A study of the normal immune mechanisms involved in tumor rejection when it naturally occurs should reveal critical signals for generation of antitumor CTL in vivo. The highly transfectable variant of P815, P1.HTR, was found to be rejected in the hind footpads by approximately one-third of syngeneic DBA/2 mice. Analysis of draining popliteal lymph nodes revealed a large influx of CD4+ and CD8+ T lymphocytes in all mice, indicating that a failure to reject was not due to the complete absence of an inflammatory response. However, although IL-2 and IL-3 were produced by lymph node cells from all mice, only approximately one-third generated a high IFN-gamma response. IL-4 was not detected. To explore a role for IL-12 in the induction of the IFN-gamma-producing phenotype, a histidine-tagged IL-12 fusion protein was expressed in mammalian cells and purified by nickel-chelate chromatography, and a rabbit antiserum was produced. Neutralization of IL-12 in vivo eliminated the high IFN-gamma response and prevented rejection of P1.HTR tumors and also of a more immunogenic tum- variant of P815, P198. Conversely, exogenous IL-12 delivered early during challenge with P1.HTR cells induced high IFN-gamma production and resulted in tumor rejection in most mice. Therefore, endogenous IL-12 is vital for the rejection of these tumors when it naturally occurs, supporting a role for exogenous administration of this cytokine to favor a Th1-like phenotype in the immunotherapy of cancer.

The requirements for generating CD8+ CTLs against the mastocytoma P815 were first defined by using an allogeneic mixed lymphocyte tumor culture (MLTC). Both the expansion of effector lymphocytes and the acquisition of lytic activity were dependent on the presence of accessory cells, but not CD4+ lymphocytes. Several factors were examined for their ability to replace accessory cell function. Expression of B7-1 by P815 was sufficient to induce IL-2 production by CD8+ cells, but substantial proliferation was achieved only if IL-6 was provided as well. Although IL-12 had little effect on the net proliferation of developing effector cells, it increased specific lytic activity 10-fold, acted synergistically with B7-1 in induction of IFN-gamma production during primary stimulation, and resulted in a shift toward a Th1 cytokine profile following secondary stimulation. These costimulatory factors were then studied in a primary syngeneic MLTC by using splenocytes from nonimmunized DBA/2 mice, an approach that had never before succeeded in generating specific CTLs. The combination of B7-1, IL-6, and IL-12 was sufficient to induce P815-specific CTL activity after a 5-day MLTC, which was expanded optimally following secondary stimulation in the presence of B7-1, IL-2, and IL-7. The irrelevant syngeneic tumor L1210, constructed to express B7-1 and the P815-derived tumor Ag gene P1A, also generated CTLs that lysed the parental P815. The combination of B7-1, IL-6, and IL-12 should be useful in the derivation of other tumor-specific CTLs in vitro, and may constitute the optimal stimuli for rapid proliferation and differentiation of helper-independent, tumor-specific CTLs in vivo.

No abstract

The protein tyrosine kinase Fyn has been shown to be involved in signal transduction through the TCR and the glycosyl-phosphatidylinositol-linked surface molecule Thy-1 expressed on T cells. In this study, we examine the requirement for Fyn expression in signaling through the TCR or Thy-1 using a panel of Ag-specific T cell clones derived from fyn-/- mutant mice. These clones do not express normal Fyn protein, as measured by immune-complex kinase reaction using anti-Fyn Ab. Stimulation through the TCR, either by APC bearing relevant Ag or by immobilized anti-CD3 mAb, resulted in comparable levels of proliferation, lymphokine production, and cytolysis by clones from both wild-type and fyn-/- mice. In contrast, stimulation through Thy-1, using soluble (or cross-linked) anti-Thy-1 mAb, was deficient, as measured by these responses. Thus, Fyn expression is selectively required for functional activation through Thy-1 in these T cell clones.