Intra-pinna anti-tumor vaccination with self-replicating infectious RNA or with DNA encoding a model tumor antigen and a cytokine

To optimize polynucleotide vaccinations for protective antitumor immunity we used a self-replicating RNA vaccine in which Semliki Forest virus replicase drives RNA expression of the lacZ gene coding for beta-galactosidase as model tumor-associated antigen (TAA). This was compared with replicase-deficient control RNA and with lacZ DNA plasmids with respect to gene expression in vitro and in vivo and for vaccination using the mouse ear pinna as an optimal immunization site. In vitro, the highest expression was observed with self-replicating RNA. Gene expression following pinna inoculation of either non-replicating DNA plasmids or self-replicating RNA was similar, lasting for 2-3 weeks. Higher antibody responses were obtained with RNA than with DNA. beta-Gal peptide specific CTL memory responses to lacZ DNA or RNA lasted for more than 6 weeks while respective responses induced by lacZ-transfected tumor cells lasted for only 2 weeks. To achieve a protective response against lacZ tumor cells with self-replicating RNA about a 100-fold lower dose of polynucleotide was sufficient in comparison to DNA. The extent of protective antitumor immunity not only depended on the gene dose used for vaccination, but also on the aggressiveness of the lacZ-transfected tumor line used for challenge. In comparison to lacZ-transfected tumor cells as vaccines, polynucleotide vaccination also demonstrated superiority with regard to cross-protection. Protective antitumor immunity could be strongly increased upon co-inoculation of lacZ DNA with IL-2 DNA or IL-12 RNA. IL-2 DNA, but not IL-12 RNA, also augmented the CTL response while IL-12 RNA, but not IL-2 DNA, reduced the antibody response. These results demonstrate efficient protective antitumor immunity after intra-pinna lacZ TAA polynucleotide vaccination and show additional immunomodulatory effects by co-administration of cytokine polynucleotides.