Adenovirus-derived (Ad) vectors are being developed for use as anti-cancer vaccines and as oncolytics. Additionally, Ad vectors are undergoing testing as vaccine vectors against every major human pathogen. It is now appreciated that different Ad vector serotypes induce distinct innate immune responses in vivo and in vitro, in addition to circumventing pre-existing immunity. This includes differential induction of IFN-γ by T cells. However, the processes driving this robust IFN-γ production remain unexplained. We thus hypothesized that (1) Ad vectors activate MAIT cells, and (2) that distinct serotypes of Ad vectors differ in their capacity to activate MAIT cells.

To test these hypotheses, human PBMCs were transduced with Ad serotype 5 (Ad5) vectors and ChAdOX1 vectors. After 24 hours, T cell activation was assessed by expression of CD69, Granzyme B, and IFN-γ. Consistent with our hypotheses, ChAdOX1 robustly induced IFN-γ production by MAIT cells (mean of 49.7%) at an MOI of 10³ VP. By contrast, Ad5 induced negligible production of IFN-γ (mean of 0.5%). Similar differential induction of Granzyme B and CD69 was also observed. Mechanistically, the differential capacity of ChAdOX1 and Ad5 to activate MAIT cells was due to two factors: (1) ChAdOX1 robustly induced IL-18 production by monocytes, while Ad5 induced no detectable IL-18; and (2) ChAdOX1 transduced pDCs and induced production of IFN-α/β, while Ad5 failed to transduce pDCs. Finally, we determined that Cathepsin B-mediated activation of the NLRP3 inflammasome was critical for the production of IL-18 and activation of MAIT cells in response to ChAdOX1 stimulation.

In summary, we demonstrate for the first time that a virus-derived, candidate oncolytic/vaccine vector platform can potently activate innate-like T cells, and distinct vectors display differential activation capacity. These data can inform improved, rational design of adenovirus vectors as therapeutics for cancer and infectious diseases.