Refine
Year of publication
- 2004 (5) (remove)
Document Type
- Article (4)
- Doctoral Thesis (1)
Language
- English (5) (remove)
Has Fulltext
- yes (5)
Is part of the Bibliography
- no (5)
Institute
- Georg-Speyer-Haus (5) (remove)
The signal transducer and activator of transcription (Stat) gene family comprises seven members with similarities in their domain structure and a common mode of activation. Members of this gene family mediate interferon induction of gene transcription and the response to a large number of growth factors and hormones. Extracellular ligand binding to transmembrane receptors causes the intracellular activation of associated tyrosine kinases, phosphorylation of Stat molecules, dimerization, and translocation to the nucleus. Prolactin-induced phosphorylation of Stat5 is a key event in the development and differentiation of mammary epithelial cells. In addition to the crucial phosphorylation at tyrosine 694, we have identified an O-linked N-acetylglucosamine (O-GlcNAc) as another secondary modification essential for the transcriptional induction by Stat5. This modification was only found on nuclear Stat5 after cytokine activation. Similar observations were made with Stat1, Stat3, and Stat6. Glycosylation of Stat5, however, does not seem to be a prerequisite for nuclear translocation. Mass spectrometric analysis revealed a glycosylated peptide in the N-terminal region of Stat5. Replacement of threonine 92 by an alanine residue (Stat5a-T92A) strongly reduced the prolactin induction of Stat5a glycosylation and abolished transactivation of a target gene promoter. Only the glycosylated form of Stat5 was able to bind the coactivator of transcription CBP, an essential interaction for Stat5-mediated gene transcription.
Safety concerns associated with the use of viral vectors in gene therapy applications have attracted considerable attention towards the development of nonviral vectors as alternatives for DNA delivery. While nonviral vectors are commonly not associated with safety problems, they are still very inefficient compared to viral vectors, and require significant improvements to approach the efficiency of their viral counterparts. Meanwhile ligands or single-chain antibody fragments that bind to cell surface receptors for increased and/or specific cellular uptake, endosome escape activities, and nuclear localization sequences (NLSs) to enhance transport of plasmid DNA into the nucleus, have become available that can be incorporated into nonviral vectors to improve their efficacy. However, as gene delivery is a multistep process, the challenge is to incorporate multiple of these functional elements into a single nonviral vector system, while retaining their specific activities. A promising method to attach such entities to plasmid DNA is the use of multifunctional fusion proteins that bind to DNA through a DNA-binding domain. In principle, two types of DNA-binding domains/proteins can be used to anchor additional functional domains or peptides to a plasmid, namely sequence-specific DNA-binding domains, described in the first part of this thesis, or those that bind DNA independent of its sequence, exemplified in the second part of this work by a derivative of the human HMGB2 protein. The first fusion protein constructed and analyzed contained the E. coli LexA repressor as a sequence-specific DNA-binding domain. In addition, this DNA-carrier protein, termed TEL, included a bacterial translocation domain as an integrated endosome escape activity, and human TGF-a for specific targeting to the EGF-receptor (EGFR). TEL was expressed in E. coli and purified under both native and denaturing conditions. Purified, denatured TEL was refolded and subsequently shown to bind specifically to EGFR-expressing cells. However, inclusion of TEL in complexes of plasmid DNA and poly-L-lysine (pL) did not lead to increased gene delivery into EGFR-expressing COS-1 cells. Most likely this was due to the absence of DNA-binding activity of the LexA moiety in TEL. In contrast, native TEL was able to interact specifically with DNA. Nevertheless, since this interaction was rather weak, and refolding of denatured TEL had not resulted in functional activity of all of its protein domains, it seemed unlikely that fusion proteins containing LexA would exhibit gene transfer capabilities superior to those of similar DNA-carrier proteins previously constructed in our group. Further work therefore focused on the use of the E2C-Sp1C protein as an alternative sequencespecific DNA-binding domain. This artificial zinc-finger protein was fused to the single-chain antibody fragment scFv(FRP5), directed against the human ErbB2 growth factor receptor. The resulting 5-E2C fusion protein was expressed in E. coli and purified under native and denaturing conditions. Refolded and native 5-E2C were found to bind specifically to ErbB2-expressing cells, indicating that scFv(FRP5) in 5-E2C was functional in both preparations. In contrast, whereas refolded 5-E2C bound DNA only weakly, significant DNA binding was observed for native 5-E2C. In addition, it could not only be shown that the interaction of native 5-E2C with DNA containing its recognition sequence was specific, but also that this protein was able to bind DNA and recombinant ErbB2 simultaneously, demonstrating the functionality of both domains in native 5-E2C. Despite these encouraging results, the inclusion of native 5-E2C in pL- or polyethyleneimine (PEI)-DNA complexes did not lead to an (5-E2C-specific) enhancement of gene transfer efficiency, irrespective of the presence of the endosome-disruptive reagent chloroquine during transfection. In the second part of this thesis an alternative approach for the development of DNA-carrier proteins for nonviral gene delivery is described, based on human HMGB2, a DNA-binding protein without sequence specificity. HMGB2 contains an acidic C-terminus that has been found to decrease the affinity of the protein for DNA. Therefore, this C-terminal tail was deleted, resulting in an HMGB2-variant consisting of amino acids 1-186. HMGB2186, purified under native conditions from E. coli lysates, was able to interact with DNA and bound to the surface of different cell lines. Importantly, after binding to plasmid DNA HMGB2186 mediated gene delivery into COS-7 cells with higher efficiency than pL. In addition, HMGB2186-mediated gene transfer was strongly enhanced in the presence of chloroquine, indicating that the endocytic pathway was involved in cellular uptake. To improve internalization and intracellular routing of HMGB2186 as a DNA-carrier, a derivative containing the TAT47-57 cell-penetrating peptide (CPP), reported to facilitate cell entry independent of endocytosis, was constructed. Since this peptide also contains an NLS, in addition an HGMB2186-variant containing the SV40-NLS was constructed to investigate the effect of a peptide that has only nuclear localizing properties. Interestingly, the resulting TAT-HMGB2186 and SV40-HMGB2186 fusion proteins displayed DNA-binding activities similar to HMGB2186, but mediated gene delivery into different cell lines clearly more efficiently than the parental molecule. Furthermore, the efficacy of both fusion proteins was enhanced markedly in the presence of chloroquine, an indication that endocytosis was involved in the transfection process mediated by these proteins. This suggests that the increased transfection efficiency observed for TAT-HMGB2186 was more likely due to the NLS function present in the TAT47-57 peptide, rather than to its ‘cell penetrating properties’. Finally, the incorporation of functional peptides derived from human proteins into HMGB2186 was investigated. An uncharged CPP originating from Kaposi-FGF, reported to facilitate efficient cellular uptake of fused protein domains in an endocytosis-independent manner, was fused to HMGB2186 together with the SV40-NLS. Interestingly, the resulting KSV40-HMGB2186 fusion protein bound DNA similarly as previously tested DNA-carrier proteins, but did not mediate enhanced transfection compared to HMGB2186. In addition, the importin-b-binding (IBB) domain derived from human importin-a2 was investigated as a component of a DNA-carrier protein. Since the IBB domain can function as an NLS, it was fused to HMGB2186 resulting in the DNA-carrier protein IBBHMGB2186. Although IBB-HMGB2186 bound DNA in a similar manner as the other HMGB2186-derivatives, gene delivery mediated by IBB-HMGB2186 was only as effective as HMGB2186 mediated transfection, suggesting no significant role of the IBB domain. However, addition of chloroquine resulted in a remarkable enhancement of IBB-HMGB2186-mediated gene transfer, which was now more efficient than with any other HMGB2186-variant tested, and not much lower than gene transfer mediated by PEI, one of the most efficient transfection reagents available to date. To enhance nonviral gene delivery even further, the HMGB2186-based DNA-carrier proteins described in this thesis might now serve as building blocks for novel fusion proteins that include additional complementing activities. In this respect it seems particularly promising that, under conditions of effective end some escape, IBB-HMGB2186, which consists entirely of protein domains of human origin, was the most efficient of all proteins tested in this work.
The continuously growing natural killer (NK) cell line NK-92 is highly cytotoxic against malignant cells of various origin without affecting normal human cells. Based on this selectivity, the potential of NK-92 cells for adoptive therapy is currently being investigated in phase I clinical studies. To further enhance the antitumoral activity of NK-92 cells and expand the range of tumor entities suitable for NK-92-based therapies, here by transduction with retroviral vectors we have generated genetically modified NK-92 cells expressing chimeric antigen receptors specific either for the tumor-associated ErbB2 (HER2/neu) antigen or the human Epithelial Cell Adhesion Molecule (Ep-CAM). Both antigens are overexpressed by many tumors of epithelial origin. The chimeric antigen receptors consist of either the ErbB2 specific scFv(FRP5) antibody fragment or the Ep-CAM specific scFv(MOC31), a flexible hinge region derived from CD8, and transmembrane and intracellular regions of the CD3 zeta chain. Transduced NK-92-scFv(FRP5)-zeta or NK-92-scFv(MOC31)-zeta cells express high levels of the fusion proteins on the cell surface as determined by FACS analysis. In europium release assays no difference in cytotoxic activity of NK-92 and transduced NK-92 cells towards ErbB2 or Ep-CAM negative targets was found. However, even at low effector to target ratios transduced NK-92 cells specifically and efficiently lysed established ErbB2 or Ep-CAM expressing tumor cells that were completely resistant to cytolytic activity of parental NK-92 cells. Similarly, ErbB2-positive primary breast cancer cells isolated from pleural effusions of patients with recurrent disease were selectively killed by NK-92-scFv(FRP5)-zeta. In an in vivo model in immunodeficient mice treatment with retargeted NK-92-scFv(FRP5)-zeta, but not parental NK-92 cells resulted in markedly delayed growth of ErbB2 transformed cancer cells. These results demonstrate that efficient retargeting of NK-92 cytotoxicity can be achieved, and might allow the generation of potent cell-based therapeutics for the treatment of ErbB2 and Ep-CAM expressing malignancies. This therapeutic approach might be applicable for a large variety of different cancers where suitable cell surface antigens have been identified.
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.