Expanding the repertoire of Modified Vaccinia Ankara-based vaccine vectors via genetic complementation strategies

Modified Vaccinia virus Ankara (MVA) is a safe, highly attenuated orthopoxvirus that is being developed as a recombinant vaccine vector for immunization against a number of infectious diseases and cancers. However, the expression by MVA vectors of large numbers of poxvirus antigens, which display immunodominance over vectored antigens-of-interest for the priming of T cell responses, and the induction of vector-neutralizing antibodies, which curtail the efficacy of subsequent booster immunizations, remain as significant impediments to the overall utility of such vaccines. Thus, genetic approaches that enable the derivation of MVA vectors that are antigenically less complex may allow for rational improvement of MVA-based vaccines.

We have developed a genetic complementation system that enables the deletion of essential viral genes from the MVA genome, thereby allowing us to generate MVA vaccine vectors that are antigenically less complex. Using this system, we deleted the essential uracil-DNA-glycosylase (udg) gene from MVA and propagated this otherwise replication-defective variant on a complementing cell line that constitutively expresses the poxvirus udg gene and that was derived from a newly identified continuous cell line that is permissive for growth of wild type MVA. The resulting virus, MVADeltaudg, does not replicate its DNA genome or express late viral gene products during infection of non-complementing cells in culture. As proof-of-concept for immunological 'focusing', we demonstrate that immunization of mice with MVADeltaudg elicits CD8+ T cell responses that are directed against a restricted repertoire of vector antigens, as compared to immunization with parental MVA. Immunization of rhesus macaques with MVADeltaudg-gag, a udg(-) recombinant virus that expresses an HIV subtype-B consensus gag transgene, elicited significantly higher frequencies of Gag-specific CD8 and CD4 T cells following both primary (2-4-fold) and booster (2-fold) immunizations as compared to the udg(+) control virus MVA-gag, as determined by intracellular cytokine assay. In contrast, levels of HIV Gag-specific antibodies were elicited similarly in macaques following immunization with MVADeltaudg-gag and MVA-gag. Furthermore, both udg(-) and udg(+) MVA vectors induced comparatively similar titers of MVA-specific neutralizing antibody responses following immunization of mice (over a 4-log range: 10(4)-10(8) PFU) and rhesus macaques. These results suggest that the generation of MVA-specific neutralizing antibody responses are largely driven by input MVA antigens, rather than those that are synthesized de novo during infection, and that the processes governing the generation of antiviral antibody responses are more readily saturated by viral antigen than are those that elicit CD8+ T cell responses.

Our identification of a spontaneously-immortalized (but not transformed) chicken embryo fibroblast cell line (DF-1) that is fully permissive for MVA growth and that can be engineered to stably express MVA genes provides the basis for a genetic system for MVA. DF-1 cells (and derivatives thereof) constitute viable alternatives, for the manufacture of MVA-based vaccines, to primary CEFs -- the conventional cell substrate for MVA vaccines that is not amenable to genetic complementation strategies due to these cells' finite lifespan in culture. The establishment of a genetic system for MVA, as illustrated here to allow udg deletion, enables the generation of novel replication-defective MVA mutants and expands the repertoire of genetic viral variants that can now be explored as improved vaccine vectors.