Interactions with Innate Immunity Determine the Immunogenicity of Adenovirus Vectors Types 5, 28, and 35: Implications for Vaccine Design

Abstract

Recombinant adenovirus vectors (rAd) have proven to be very effective at inducing antigen-specific, polyfunctional T cell responses. Recombinant adenovirus serotype 5 (rAd5)-based vectors have been extensively studied as potential HIV/AIDS vaccines and tested in phase I and phase II clinical trials. The results of these trials, in conjunction with studies in rhesus macaques, have revealed that pre-existing immunity against the rAd5 vector can reduce the immunogenicity of the vaccine and limit the memory response to the HIV-antigen insert. Since 40-80% of the world's population is seropositive for Ad5, the usefulness of a rAd5-based vaccine may be compromised. To circumvent preexisting immunity, alternative adenovirus vectors from serotypes with much lower seroprevalence, such as Ad28 and Ad35, are under development. However, some vectors constructed from low-seroprevalence adenoviruses have shown poor immunogenicity in vivo. This presents a paradox whereby rAd5, which induces a good immune response, is limited due to widespread preexisting immunity while rAd28 and rAd35, to which there is low pre-existing immunity, are inherently less immunogenic.

My results show that, in stark contrast to rAd28 and rAd35, rAd5 fails to induce significant changes in DC mRNA expression or maturation. These changes are independent of infectivity and primarily involve the activation of innate immunity, particularly type I interferon signaling. The production of type I IFN and the activation or innate immunity by rAd28 and rAd35, but not rAd5 correlates well with the loss of the insert expression due to upregulated apoptosis of insert positive cells.

This paradigm was consistent in mouse DCs, with rAd28 and rAd35, but not rAd5 inducing significant IFNα production and that intact IFNα signaling reduces vector expression during vaccination. The induction of type I interferon in mice vaccinated with rAd28 or rAd35 lead to a decrease in the magnitude, but an improvement in the long-term memory potential (CD127 expression) and cytokine polyfunctionality of the insert-specific CD8 T cell response. The presence and impact of innate immune activation, type I interferon signaling, and insert loss should therefore be taken into account when designing future rAd vectors.