Refine
Year of publication
- 2004 (2)
Document Type
- Article (2)
Language
- English (2)
Has Fulltext
- yes (2)
Is part of the Bibliography
- no (2)
Institute
The receptor tyrosine kinase ErbB2 (HER2) is overexpressed in multiple human tumors of epithelial origin. High ErbB2 expression is functionally involved in tumorigenesis and correlates with poor clinical prognosis. For immunotherapy of ErbB2 expressing tumors, we developed a strategy to supply the tumor cells with costimulatory activity. A bispecific fusion protein was constructed (BIg5), containing the IgV-like domain of huCD86, the CH2/CH3 domain of huIgG1 and the ErbB2-specific single chain antibody fragment scFv(FRP5). A similar fusion protein lacking the CD86 domain (Ig5) was used as a control. Upon binding of BIg5 to ErbB2 on tumor cells, these cells display CD86 on their surface and thus can deliver costimulatory signals for T-cell activation. In addition, NK cells could be activated by CD86 binding to CD28. BIg5 is secreted by eukaryotic cells as a homodimer with increased stability compared to monomers and possibly enhanced costimulatory activity due to crosslinking of CD28 on effector cells. By FACS analysis, specific binding of the scFv(FRP5) domain to ErbB2 as well as CD86 IgV binding to CTLA-4 could be demonstrated. Together with anti-CD3 antibody, BIg5 stimulates proliferation of human CD2-purified lymphocytes in vitro. After binding to ErbB2 on murine Renca-lacZ/ErbB2 tumor cells, about 50% of initially bound BIg5 is still present on the cell surface after 4 hours. For delivery of chimeric fusion proteins in vivo, we used syngeneic, stably transfected HC11 mammary epithelial cells continuously secreting the proteins. Inoculation of these bystander cells close to subcutaneously growing Renca-lacZ/ErbB2 tumors should provide a long-lasting source to achieve high local concentrations of BIg5 at the tumor site. In vivo HC11-BIg5 cells proved to be non-tumorigenic and secreted BIg5 for several weeks, causing a strong anti-BIg5 antibody response. Treatment of established Renca-lacZ/ErbB2 or ErbB2-negative Renca-lacZ tumors by peritumoral inoculation of either HC11-BIg5 or HC11-Ig5 cells led to rejection of all Renca-lacZ/ErbB2, but none of the Renca-lacZ tumors. HC11neo control cells had no effect on tumor growth. Rejection of ErbB2+ tumors led to long-term protection also against subsequent challenge with intravenously injected ErbB2- tumor cells. Intraperitoneal injection of bystander cells secreting the fusion proteins did not lead to tumor regression suggesting that high local concentrations at the tumor site are necessary to target ErbB2 on tumor cells and to overcome elimination of BIg5 or Ig5 by neutralizing antibodies. The CD86 IgV domain of BIg5 did not play a major role in the observed antitumoral immune response suggesting NK-cell mediated ADCC as the initial effector mechanism followed by activation of tumor specific T cells. Targeting of ErbB2 on tumor cells with antibody fusion proteins that interact specifically with the host immune system could be an efficient and specific approach for therapy of solid ErbB2+ tumors.
Tumor-specific T lymphocytes can be regarded as a highly effective mechanism for tumor rejection. A substantial number of T-cell defined tumor antigens including mutated oncoproteins and differentiation antigens have been identified. However, while most spontaneous tumors appear to be antigenic, few are immunogenic. Activation of tumor-specific cytotoxic T cells (CTL) requires presentation of tumor antigens by professional antigen presenting cells (APCs) via MHC I molecules. Due to their crucial role in T-cell activation, APCs are being exploited for active cancer immunotherapy. Present experimental strategies include the incubation of dendritic cells with synthetic, tumor specific peptides to achieve uptake of tumor antigens and presentation in the context of MHC molecules. Alternatively, gene therapeutic approaches are aimed at the endogenous expression of tumor antigens in APCs upon transfer of suitable vector constructs. Our strategy for the presentation of tumor antigens by APCs is based on the intracellular delivery of tumor antigens as part of a fusion protein specifically targeted to APC cell surface receptors. We have constructed prototype molecules that contain a soluble fragment of CTLA-4 for cell binding via interaction with B7 molecules, genetically fused to a protein fragment derived from the tumor-associated antigen ErbB2. To improve uptake and direct the antigenic determinant preferentially to the MHC class I pathway, in one of these protein vaccines also the translocation domain of the bacterial Pseudomonas exotoxin A has been included. In the parental toxin this protein domain facilitates escape from the endosomal compartment to the cytosol upon receptor mediated endocytosis. Here we have investigated the in vitro cell binding activity of such reagents and their antitumoral activity in immunocompetent murine model systems. Specific binding to B7 molecules and uptake of bacterially expressed protein vaccines could be demonstrated. Ex vivo restimulation with an ErbB2-derived peptide of splenocytes from Balb/c mice injected with the fusion proteins resulted in enhanced IFN-gamma production by T cells. Protective and therapeutic effects of ErbB2 protein vaccines were also investigated. Vaccinated animals were protected against subsequent challenge with syngeneic ErbB2 expressing tumor cells. Likewise, s.c. injection of ErbB2 protein vaccines in the vicinity of established tumors resulted in tumor rejection and long lasting protection indicating that immunological memory was induced. Our results suggest that chimeric proteins combining a tumor antigen and specific recognition of APCs in a single molecule are suitable for targeted delivery of antigens to professional APCs and might become valuable tools for cancer immunotherapy.