Adoptive Tumor Immunotherapy Mediated By T Cells Expressing Chimeric T Cell Receptor

Effective tumor immunity requires immune recognition of tumor antigens and the tumor eradication by immune effector cells. Over the past decades, immunologists and oncologists have been searching for tumor-specific antigens. However, most antigens that have been identified so far are self-antigens, many of which are either exotically expressed or over-expressed than in normal tissues. Tumor cells have the abilities to escape from the immune response through mechanisms such as anergy, apoptosis and suppression. Although tumor antigens can be recognized, they are not sufficient to enable tumor eradication until the primed cells expand to sufficient numbers, migrate to tumor sites, mature into effector cells and overcome the tumor's abilities of escaping from the immune surveillance. The infusion of tumor-specific T cells into the tumor-bearing host referred to as adoptive immunotherapy , such as lymphokine-activated killer (LAK) cells and tumor-infiltrating lymphocytes (TIL) can be highly effective in the destruction of tumor burdens in some patients with melanoma and renal carcinoma. However, this therapy has been unsuccessful in the treatment of more prevalent cancers such as ovarian, breast cancers and colon carcinoma. Three problems have commonly precluded more universally effective adoptive immunotherapy including ①difficulties in isolating and culturing tumor specific T cells from all tumor types; ② poor tumor localization of ex vivo-modified lymphocytes upon adoptive transfer; and ③ IL-2 toxicity. On the other hand, the slow tumor penetration and short half-life of exogenously administered tumor-specific monoclonal antibodies have provided major obstacles for an effective destruction of tumor cells by the humoral effectors of the immune system. Attempts to improve the efficacy of adoptive immunotherapy have led to the development of novel strategies that combine advantages of T cell-based and antibody-based immunotherapy by gene engineering technology to construct the chimeric artificial T cell receptor which comprises antibody fragments specific for tumor-associated antigens and a cellular activation motif. The artificial chimeric antigen receptors (CARs) will be transduced into T cells by retroviral transduction system and the antigen recognition is thereforenot restricted to major histocompatibility complex. This strategy will be of great value for adoptive tumor immunotherapy. On the other hand, increasing evidence from animal studies and clinical trials indicated that CD4+ T cell plays a central role in orchestrating host immune responses against cancer. Adoptive transfer using tumor-specific Thl and Tel cells is a promising therapeutic strategy for tumor immunotherapy. However, its application has been hampered because of difficulties in generating tumor-specific Thl cells from tumor patients. To overcome this problem, we transfected CD4+ and CD8+ T cells with CARs to generate both helper and CTL responses in this projects.ErbB2 or Her-2/neu, a member of the epidermal growth factor receptor family with tyrosine kinase activity, encodes an 185kDa transmembrane protein receptor. HER-2/neu is ubiquitously expressed in many epithelial tumors and known to be over-expressed in approximately 30% of all ovarian and breast cancers, 35-45% of all pancreatic carcinomas and up to 90% of colorectal carcinomas and this overexpression was shown to correlate with aggressiveness of malignancy and poor prognoses. Because erbB2-specific antibody and T cells are detected in breast and ovarian cancer patients, erbB2 is recognized as a target of immunotherapy.In part I , we first constructed a retroviral vector which expressed the chimeric scFv-anti-erbB2-CD28- ?> receptor by gene engineering technology. For detecting purpose, we cloned a c-myc tag epitope (EIKLISEEDL) at the C terminus of VL. Anti-erbB2 scFv was obtained from VH and VL by overlapping PCR, CD28 molecule and CD3 ￡ cytoplasmic domain were obtained by PCR, respectively. An appropriate spacer of outer cytoplasmic CD28 was inserted between scFv and membrane to keep its right folding, and the intracellular domain of CD3 ￡ chain and CD28 domain as the activation motif and costimulation motif, respectively, could redirect the specificity of a transfected mouse T lymphocytes towards erbB2-expression tumor cells. The chimeric antigen receptors(CARs) were composed of scFv-anti-erbB2> CD28 molecule and cytoplasmic regions of ￡ chain. A three-plasmid expression system was used to generate MLV-derived retroviral vector particles by transient transfection. Viralsupernatant was harvested at 48h after transfection, passed through a 0.45 v m membrane, and refrigerated. Concentrated viral stocks were prepared by centrifugation of viral supernatant in an SA-300 rotor at 50000g, 4°C, for 1.5h. Virus was resuspended in an appropriate volume of 1640 and stored at -80°C. Viral titers were estimated by transduction of NIH3T3 cells.In part II, we engineered the chimeric antigen receptor on mouse T lymphocytes and conducted experiments in vitro. Firstly, the CD4+ and CD8+ T cells were sorted by MACS separators and the purity was up to 90%. After stimulated by ConA or antibodies, the cells were infected by condensed virus. The transduction efficiency of retroviral vectors containing scFv-CD28- 4 or a mock gene into CD4+ and CD8+ T cells were determined by measuring the percentage of anti-c-myc-FITC cells using flow cytometry, ranged from 35% to 45%, and CD4+ and CD8+ T cells subsets were transduced at similar efficiencies. 1 > We detected the antigen-specific binding ability of gene modified T cells. We stained T-CARs and T-mock cells with Her2/Ig, respectively. T-CARs conjugated with Her2/Ig detected by flow cytometry. 2> We evaluated the ability of transduced CD4+ and CD8+ T cells to mediate specific target cell lysis in a non-radioactivation cytotoxicity assay. T cells expressing scFv-anti-erbB2 receptor were able to lyse the erbB2+ tumor cells D2F2/E2 and SK-BR-3, but not erbB2" MCF-7 and D2F2/E2-mask cell lines. Mock-transduced T cells did not lyse D2F2/E2, SK-BR-3 > D2F2/D2-mask or MCF-7 cells. CD8+T-CARs cells exhibited a stronger cytotoxicity compared with CD4+ T-CARs cells, indicating the cytotoxicity was antigen-dependent and CARs induced. 3 > We also demonstrated the ability of gene-modified T cells to expand in culture after specific ligation with irradiated erbB2+ tumor target for 72 hours. Compared with antibody-mediated stimulation of endogenous CD3 and CD28 receptors expressed on the same T cells, the gene-modified T cells have an equivalent proliferation ability. 4, The capacity of CD4+and CD8+ T-CARs cells to produce cytokines in response to erbB2-expressing tumor cells were examined by measuring cytokine levels in the culture supernatant. CD4+ and CD8+ T-CARs cells produced high levels of IFN- y and GM-CSF when they were cocultured with erbB+ D2F2/E2 and SK-BR-3 target cells. T cells transduced with mock virussecret only background level of cytokines when incubated with D2F2/E2 or MCF-7 cells. In addition, These CD4+ or CD8+T-CARs cells did not produce IL-4, IL-5 or IL-10 in response to stimulation by erbB+ tumor cells(data not shown), indicating that CD4+ and CD8+ T-CARs were successfully induced from nonspecifically activated T cells in Th 1-polariaing conditions. CD4+and CD8+T-CARs cells did not produce IFN-Y and GM-CSF in response to erbB2' MCF-7 and D2F2/E2-mask cells, indicating that its production was induced in an antigen-specific manner.In part III, the antitumor efficacy of gene-modified T cells was evaluated against the D2F2/E2 tumor cell line following adoptive transfer. In the first model, 1X105 D2F2/E2 tumor cells were injected subcutaneously into 3 groups of BALB/c nude mice(n=10). Enriched CD3+ T cells from BALB/c mice(transduced with the scFv-CD28- C receptor or mock transduced T cells ) were injected intravenously into tumor-inoculated mice at 6 hours(day 0, 5X106) and 24 hours(day 1, 5X106). Only tumor cells were injected and non-T cells were transferred as control. In the second model, 1X105 D2F2/E2 tumor cells were injected subcutaneously into 3 groups of BALB/c mice(n=10). The treatment regimen was as the same as in the first model. After adoptive transfer of T cells, tumor growth in mice was measured daily by Caliper Square along the perpendicular axes of the tumor. The data were recorded as mean volume+ SEM. Tumor growth in mice was observed daily for at least 60 days. After the initial tumor challenge, tumor free mice were rechallenged with 1 X 105 D2F2/E2 tumor cells in the opposite flank. Naive mice were also injected with 1 X 105 D2F2/E2 cells to verify tumor growth. In BALB/c nude mice, although CD3+ T-CARs cells strongly inhibited tumor growth in vivo, only 2 out of 10 mice completely eradicated the tumor. In BALB/c syngeneic mice, 7 out of 10 completely rejected the tumors, and then we rechallenged them with 1 X 105 D2F2/E2 in the opposite flank of the tumor free mice. Any tumors were not palpated for 4 weeks observation in rechallenged BALB/c mice and two BALB/c nude mice all grew tumors. Spleen T cells were harvested from tumor free mice and control mice; cytotoxicity capacity of T cells was determined in non-radioactivation cytotoxicity assay. T cells coming from tumor-free mice were able to lyse not only D2F2/E2 cells but also D2F2/E2-mask cells, although the lysis functionto D2F2/E2 was stronger than to D2F2/E2-mask cells. We also evaluated IFN- Y and GM-CSF secretion when T cells cocultured with D2F2/E2, D2F2/E2-mask cells in 12 well plates for 20 hours. Supernatants were harvested and IFN- y and GM-CSF secretion by transduced T cells were dertermined by ELISA. T cell coming from tumor-free mice were able to produce GM-CSF and IFN- Y when cocultured with D2F2/E2^ D2F2/E2-mask cells. While T cells coming from control mice were unable to produce any cytokines. Histologically, tumors treated by T-CARs showed massive destruction accompanied by a large degree of lymphocytic infiltration in BALB/c mice. Immunohistochemistry assays showed that massive gene modified T cells detected by anti-c-myc-biotin staining in tumor site, lymph nodes, spleen and thymus. Tumors treated by T-mock cells were conserved with no evidence of lymphocytes infiltration and tissue destruction.To investigate whether these CARs were involved in immune synapse when T cells were stimulated by corresponding antigens. In part IV, we studied the colocalization of CARs with lipid rafts using a raft marker fluorescein isothiocyanate (FITC)-labled cholera toxin (CTx) B subunit, which binds GM1 glycosphingolipid, and anti-c-myc-CY3 binds to CARs. In unstimulated T-CARs cells, the surface distribution of FITC and CY3 appeared homogeneous. When these T cells were cultured with precoated Her2/Ig protein, within 30 min FITC and CY3 redistributed to form a dense cap. However, T-mock cells had not any changes whether stimulated or unstimulated by antigens. So we think maybe the scFv-CD28- C chimera synergistic the signaling activities of endogeneous TCR and CD28 signaling pathways. Further exploration will be made about the signal transduction by the CARs in the future.In this experiment, the artificial TCR was transduced into primary mouse T lymphocytes by retroviral transduction system and antigen recognition is therefore not restricted to MHC. CD4+ T cells in this process played a very important value. The CD8+ T-CARs cells transfused into mice could directly kill the target cells, while the CD4+ T-CARs have many anti-tumor effects mainly induced by cytokines, such as IFN-Y and GM-CSF with inflammatory and antitumor potentials. In the last but not the least, the mechanism of the function played by CARs was determined by confocol...