Idiotype mRNA therapeutic vaccination in a mouse model of B cell lymphoma

Abstract B cell lymphoma is a clonal disease where each tumor cell expresses the same, unique variable region sequence that makes up the B cell receptor (BCR), i.e. idiotype. Previously, we discovered that peptides derived from the variable regions of these receptors complex with their MHC class II molecules and are presented by the tumor B cells (Khodadoust et al. Nature, 2017, 543: 723-727). Therefore, targeting this region and the tumor-specific antigens within with a vaccine could provide a polyclonal B and T cell specific anti-tumor response, which would not affect the other cells in the body. We cloned the BCR heavy and light chain variable region from A20, a murine B cell lymphoma, and produced mRNA for a tumor-specific vaccine. We added the MHC class I trafficking domain (MITD) sequence for better antigen presentation and Flag sequences to confirm translation. We formulated them with a polymeric mRNA delivery system composed of multiple lipid and cationic components forming stable electrostatic complexes of nucleic acids. These charge-altering releasable transporters (CARTs) deliver the mRNA to antigen-presenting cells within the immune system where it is released and translated, producing the encoded tumor specific antigen. Previous work in our lab has shown that antigen-coding mRNA delivered by CARTs can generate a robust immune response against the ovalbumin model antigen, curing large, established tumors in a mouse model (Habaath et al. PNAS, 2018,115: 9153-9161). To optimize our A20 idiotype mRNA vaccine strategy, we took advantage of the robust responses elicited by OVA using OVA mRNA formulated in CARTs. We confirmed that our vaccine induced a T cell immune response in the spleen against cells expressing OVA as measured by intracellular interferon-g (IFN-γ), and that addition of an adjuvant TLR9 ligand as part of the vaccine formulation was necessary. We also determined that intravenous (IV) administration was better than intramuscular (IM) injection to induce a measurable T cell immune response. Immunizing Balb/c mice with the A20 idiotype mRNA + CpG formulated in CARTs, we demonstrated that this vaccine induced a specific T cell response in splenic T cells as measured ex vivo against A20 cells. Co-formulating the mRNA from the HV and LV chains with the CARTs proved better than either alone or when linked together as a scFv. Also, increasing the amount of idiotype mRNA to a maximum of 25ug in the vaccine improved the T cell response. We confirmed that this response required translation and protein expression since mice immunized with the same mRNA in which a stop codon had been introduced to prevent translation resulted in no immune response. In a therapy experiment where the mice were inoculated with 1x106 cells prior to vaccination, we could show that the A20 idiotype mRNA vaccine reduced tumor growth significantly over mice immunized with the STOP mRNA by day 20, p=0.0010. T cells from these mice expressed IFN-γ and CD137 in vitro in response to A20 cells. Granzyme B was also expressed, suggesting the induction of a cytolytic CD8 T cell population. In conclusion, we demonstrated the feasibility of an idiotype vaccine using mRNA, TLR9 ligand and CARTs as a delivery system. Immunized mice showed benefit from the vaccine and were able to mount a measurable immune response against cells that expressed the idiotype. Though the path forward requires optimization with respect to vaccine route, timing and possibly the actual mRNA construct itself, we believe this is the first step in developing a customized therapeutic vaccine for B cell lymphoma in humans using an mRNA approach.