Forschungsprojekt am INB

Development of a prostate cancer DNA vaccine

PCa is the second leading cause of cancer-related deaths among men in the USA.
Approximately 29.000 men died due to PCa in 2003 in the United States [Jemal, A. et al., CA Cancer J Clin 53 (2003) 5-26]. In Germany, nearly 50.000 new PCa per annum were detected and PCa represents with 22.3% the most common localization of malignant tumors in men []. The risk factors are quite unclear – probably overweight, diet rich in fat- und calories, lack of movement and smoking are associated with the development of PCa. Currently, a previously unknown virus (XMRV, related to murine leukaemia virus) has been linked to prostate cancer (PCa) [American Society for Clinical Oncology – Prostate Cancer Symposium, San Francisco, CA, USA, Feb 24-26, 2006]. Surgery and/or radiotherapy are the standard therapies for local tumor treatment. On the other hand, about one-third of PC patients will develop progressive or metastatic disease within 10 years [Oefelein, M.G. et al., J Urol 158 (1997) 1460-1465]. In early metastatic disease androgen ablation is effective, but in most cases androgen-independent tumors develop. Subsequently, no effective treatment for androgen-independent disease is available. A promising possibility to render a therapeutic treatment more effective could be the development of a prostate-specific vaccine.

Development of a therapeutic vaccine against PCa is a promising approach for an adjuvant therapy
Vaccine-based strategies are excellent adjuvant treatment options to eradicate micrometastatic disease [McNeel, D.G. et al., Immunol Lett 96 (2005) 3-9 and McNeel, D.G. et al., Cancer Chemother Biol Response Modif 22 (2005) 247-261]. It was suggested that the doubling time of serum PSA (prostate specific antigen) in stage D0 PCa (patients after definitive therapy and increasing PSA serum level) is associated with the time to the detection of metastasis and death from PCa [Freedland, S.J. et al., JAMA 294 (2005) 433-439]. As a consequence, patients are in D0 PCa stage identifies a population at high risk of developing micrometastatic disease and should benefit from adjuvant vaccine therapy.

Specific immune therapy depends on a target antigen that is ideally expressed exclusively in tumor tissue
Immunotherapy of PCa is hampered by the lack of validated tumor antigens. Different potentials prostate antigens have been identified [Tricoli, J.V. et al., Clin Cancer Res 10 (2004) 3943-3953]. Tumor antigens used for therapeutic vaccination have to fulfill al least two essential criteria. First, the antigen should be restricted to non-vital organs (here: prostate tissue). Second, the antigen should be expressed on target cells at a sufficient amount in order to provide cytotoxic efficiency. Indeed it was shown, that the induction of an immune response against the self-antigen PSA is possible [Wei, C. et al., Proc Natl Acad Sci 94 (1997) 6369-6374]. At least in rats, the immunological tolerance can be broken by immunization with a non-optimized DNA vaccine encoding for rat PAP as shown by measurement of the immune response against PAP [McNeel, D.G. et al., manuscript in preparation].
Here, we have focused on prostate specific antigen (PAP) as target for the development of therapeutic DNA vaccine. The PAP antigen is highly expressed in prostate tissue [Cunha, al., Cancer Letters (2005) 1-10] but not in any other tissues investigated [Sinha, A.A. et al., Anticancer Res 18 (1998) 1385-1392 and Solin, T. et al., Biochim Biophys Acta 1048 (1990) 72-77] and is expressed in rodents as well as in humans; hence, this antigen is of outstanding interest for preclinical testing. On the other side PAP is a secreted molecule - but in general cell surface and intracellular molecules are thought to represents the best tumor targets. For this reason we also have generated a signal-peptide deleted antigen.


The TRAMP model represents a system which mimics the natural situation of PCa development
To date the most preclinical studies has been performed in transplantable tumor models. These models are not fully adequate, because spontaneous tumor development models better mimics the natural situation of disease progression. The TRAMP (transgenic adenocarcinoma of the mouse prostate) system (C57BL/6 background) represents a model of chronic PCa development [Greenberg, N.M. et al., Proc Natl Acad Sci 92 (1995) 3439-3443]. These mice express the SV40 large T antigen (Tag) under the control of a prostate-specific androgen-dependent rat probasin-promotor leading to PCa in males during development. In this model PAP is expressed in the thymus in sufficient (low) amounts [Zheng, X. et al., J Immunol 169 (2002) 4761-4769] to enable negative selection of high-avidity T cell clones and is in the periphery selectively expressed under the influence of sexual hormones [Greenberg, N.M. et al., Proc Natl Acad Sci 92 (1995) 3439-3443]. During puberty (after week 4) animals progressively develops prostate intraepithelial neoplasia resulting in progression to invasive carcinoma of epithelial origin [Shappel, S.B. et al., Cancer Res 64 (2004) 2270-2305] and consequently metastasis [Huss, W.J. et al., Semin Cancer Biol 11 (2001) 245-260], very similar to the human pathology [DeMarzo, A.M. et al., Lancet 361 (2003) 955-964].

The major hurdle in naked DNA vaccination is the targeting from the extra cellular milieu into the nucleus
Plasmid DNA must cross the plasma membrane, either by endocytosis and subsequent endosomal escape or via direct penetration (see fig. 5). There are investigations documenting that exogenous DNA is routed through the endosomes and lysosomes [Coonrod, A. et al., Gene Ther 4 (12) (1997) 1313-1321]. It is believed, that passive diffusion of cytoplasmic DNA is extremely inefficient [Seksek, O. et al., J Cell Biol 138 (1) (1997) 131-412]. Because DNA does not bind to cytoplasmic cytoskeletal motor proteins, the plasmid DNA probably could use intermediary proteins. Similar models for DNA-associated protein-mediated transport from plasmid DNA through the cytoplasm after release from endosomes/lysosomes supports this possibility [Dean, D.A. et al., Exp Cell Res 253 (1999) 713-722 and Wilson, G.L. et al., J Biol Chem 274 (1999) 22025-22032].
The last challenge is the targeting into the nucleus, where transcription of the encoded antigen occurs (see fig. 5). Because the nuclear envelope remains in non-dividing cells (like muscle cells) intact, it was reported that only 1-10 % of the plasmid DNA reaches the nucleus [Labat-Moleur, F. et al., Gene  Therapy 3 (1996) 1010-1017 and Tachibana, R. et al., Adv Drug Deliv Rev 52 (2001) 219-226]. If large numbers (1.000.000) of plasmids lacking a nuclear localization signal is delivered to tissues, some of the plasmids enter the nucleus - resulting in low expression [Utvik, J.K. et al., Hum Gene Ther 10 (2) (1999) 291-300]. A smaller number of the same plasmid (100.000) was not able to induce detectable gene expression in the same study. The presence of a nuclear localization signal in the SV40 enhancer greatly increases the gene transfer and expression up to 200-fold [Li, S. et al., Gene Ther 8 (2001) 494-497].


Fig. 5: Intramuscular DNA application. Most likely muscle cells (and to a minority pAPCs, like DCs) are transfected after intramuscular DNA application. It is known, that in muscle cells expressed proteins can be transferred to neighbouring cells, e.g. pAPCs (cross-priming). The pAPCs moves to the draining lymph node and activates naive T-lymphocytes.

Therapeutic gene vaccination
Besides the adjuvant genes (see 1.6), here we will take advantage of further features in order to enhance gene vaccine immunogenicity: 

Immunization vector pPOE
This vector is driven by an artificial strong CMV immediate early promoter.
Because is not feasible to inject ampicilline-selected plasmid-DNA in humans, due to the relatively common
β-lactam-antibiotic allergy (small residues of the antibiotic will leave in the preparation) [Williams, S.G. et al., Nucleic Acids Res 26 (9) (1998) 2120-2124], we have introduced a kanamycine resistance gene. Moreover, β-lactam-antibiotics are commonly used in humans – after vaccination the ampicilline resistance gene could be transferred to physiological occurring bacteria, resulting in insensitivity of these bacteria against this important antibiotic group [Williams, S.G. et al., Nucleic Acids Res 26 (9) (1998) 2120-2124]. 
The number and composition of unmethylated CpG motifs within the plasmid backbone are critical to induce killer cells, to secrete IFN-
γ [Yamamoto, S. et al., Microbiol Immunol 36 (1992) 983-987] and to stimulate pAPCs to induce Th1-cytokines [Klinman, D.M. et al., Proc Natl Acad Sci USA 93 (1996) 2879-2883 and Stacey, K. J. et al., J Immunol 157 (1996) 2116-2122 and Jakob, T. et al., J Immunol 161 (1998) 3042-3049]. In draining lymph nodes DCs are activated by CpGs even if they have not been transfected by an antigen-encoding plasmid [Akbari, O. et al., J Exp Med 189 (1999) 169-177]. NK-cells are indirectly activated by CpGs via cytokines secreted by activated pAPCs [Cowdery, J. J. S. et al., J Immunol 156 (1996) 4570-4575]. Here, we have used sequences for optimal activation in mice and humans [Krieg, A.M. et al., Nature 374 (1995) 546-549 and Bauer, M. et al., Immunology 97 (1999) 699-705 and Hartmann, G. et al., J Immunol 164 (2000) 1617-1624]. The immune stimulatory effects of the CpG-cassette used here are enhanced by a TpC dinucleotide on the 5´ end and a pyrimidine-rich region on the 3´end. [Yi, A.K. et al., J Immunol 160 (12) (1998) 5898-5906 and Hartmann, G. et al., J Immunol 164 (3) 1617-1624]. The immune stimulatory properties are affected by the number and spacing of the CpG motifs and the presence of a poly G sequence [Yi, A.K. et al., J Immunol 160 (12) (1998) 5898-5906 and Hartmann, G. et al., J Immunol 164 (3) (2000) 1617-1624], therefore we have also taken advantage of these findings. On the 5´ and 3´ ends poly G sequences were added in order to enhance the cellular uptake [Kimura, Y. et al., J Biochem 116 (5) (1994) 991-994 and Ballas, Z.K. et al., J Immunol 157 (5) (1996) 1840-1845]. 
The CpG cassette within the backbone can easily be exchanged with sequences for optimal activation in humans.

Fig. 6: Immunization vector pPOE/CpG. This CMV-driven plasmid contains CpG elements optimized for the murine and human system and is kanamycine selectable.

Gene optimization
Adapting the codons for optimal use in mammalian cells has not only proven beneficial for protein expression (e.g. in case of EGFP) but was recently also shown to increase the immunogenicity after DNA immunization [Liu, W.J. et al., Virology 301 (1) (2002) 43-52, Cid-Arregui, A. et al., J of Virol 77 (2003) 4928-4937 and Steinberg, T., Öhlschläger, P. et al., Vaccine 23 (9) (2005) 1149-1157]. 
Moreover, an additional feature in order achieve more effective and TAP-independent MHC-I cross-presentation and cross-priming of CTLs were included. We also take advantage of features enhancing the nuclear entry of the plasmid vector.

The major hurdle in naked DNA vaccination is the targeting from the extra cellular milieu into the nucleus. One reason for this fact is the presence of nucleases resulting in the loss of intact plasmid DNA [Lechardeur, D. et al., Gene Therapy 6 (1999) 482-497]. Indeed it was shown, that unprotected (naked) DNA is degraded within minutes by nucleases [Lechardeur, D. et al., Gene Ther 6 (1999) 482-497]. Furthermore, the uptake of naked DNA though the intact cell membrane is quite insufficient [Petricciani, J. et al., Dev Biol (Basel, Karger) 123 (2006) 23-28]. To overcome these two challenges we additionally use electroporation technology (in cooperation with 
ichor medical systems). This technology allows an increased intracellular delivery of DNA by the application of electrical fields in the target region of the tissue (see fig. 7).

Fig. 7: Ichor’s TriGrid™ delivery system. This one-stop process delivers DNA and implants electrodes at precisely the same site.

Usage of electroporation as plasmid application method clearly improves cellular immune response as well as anti-tumour effect
Within cooperation with ichor medical systems we compare intramuscularly injected DNA with and without the impact of electrical fields. First, we generated a second generation immunization plasmid (designated pPOE), which is (I) via introduction of a kanamycin cassette suitable for the usage in humans and (II) through introduction of an optimized CpG sequence very potent in the induction of a cellular immune response. Interestingly, also animals which had received the pPOE-mCpG plasmid showed delayed tumour growth in comparison to mice which had received plasmid without CpG optimization. Immunization experiments with HPV-16 E7SH model antigen encoding pPOE immunization vector revealed dramatically enhanced cellular as well as anti-tumor responses in electroporated animals [Öhlschläger, P. 
et al., Int J Cancer 128 (2) (2011) 473-81].

FH Aachen Campus Jülich

Institut für Nano- und Biotechnologien

Prof. Dr. Peter Öhlschläger
Heinrich-Mußmann-Str. 1

52428 Jülich

T  +49.241.6009 53835

F  +49.241.6009 53834