Development of Humanized Antibodies Against Vasoactive Intestinal Polypeptide for Increased T Cell Anti-Leukemia Activity

Introduction: AML is a prevalent hematological malignancy in adults, with a 5-year survival rate of only 32%. While approximately 32,000 AML patients are diagnosed annually in the US today, projections suggest this will rise to 36,000 by 2027. Current treatments, such as allogeneic bone marrow transplantation, have limited applicability due to high morbidity and treatment-associated mortality, and current immune check-point therapies using anti-PD1 and anti-PDL1 antibodies are ineffective in AML patients. Vasoactive intestinal peptide (VIP) is a highly conserved peptide hormone with immunosuppressive properties. VIP is frequently upregulated in AML and expressed by activated T cells as a co-inhibitory ligand that downregulates T cell anti-leukemia activity. We have developed a peptide antagonist of VIP/VIP receptor signaling, ANT-308, and demonstrated increased T cell anti-tumor activity in murine leukemia and pancreatic cancer models. Herein, we report on developing humanized anti-VIP monoclonal antibodies as a new therapeutic to treat AML.

Methods: Yeast display methods were used to screen a human scFv antibody library. Multiple rounds of FACS sorting using fluorochrome-bound VIP yielded three top candidate clones with scFv fragments having high VIP binding affinity: A6, C2, and D5. The scFv fragment VL and VH sequences were cloned into the Abvec 2.0 plasmid and co-transfected into the Freestyle 293F human expression cell line to generate humanized monoclonal antibodies. Purified anti-VIP mAbs were tested for VIP binding using biotinylated VIP bound to M-280 streptavidin Dynabeads, using an anti-human FITC-conjugated secondary antibody in a flow-cytometric assay, and also using surface plasmon resonance (SPR) analysis. In vitro testing for increased T cell activation used a pool of MACS-purified T cells from three healthy donors, incubated overnight with 50 IU/ml IL-2, followed by addition of A6, C2, or D5 mAbs at 10, 50, or 100 ug/mL and activation with a sub-maximal concentration of anti-CD3/CD28/CD2. After 48 hours, T cells were analyzed for CD69 and 4-1BB activation marker expression. In an initial in vivo test for anti-leukemia activity, groups of 10 DBA/2 mice were inoculated s.c. with 1E5 P815 myeloid sarcoma cells, followed by treatment on day 7 using a single injection of 3 nmol of a pool of the 3 anti-VIP mAbs, control mAb, ANT-308, control peptide, or PBS. Mice were followed for tumor burden and survival.

Results: Analysis of anti-VIP mAb binding on VIP-coated Dynabeads demonstrated near-equivalent VIP-binding for A6, C2, and D5, with MFI 3.5-fold higher than control mAb. SPR analysis also showed VIP-binding for all three mAbs, with better binding kinetics for the D5 mAb. In the in vitro activation assay, the anti-VIP mAbs increased CD69 and 4-1BB expression in CD4+ and CD8+ T cells compared to anti-CD3/CD28/CD2 activation alone. Of note, activation levels with anti-VIP antibodies were higher than those observed with ANT-308, a peptide antagonist of the VIP receptor. In the in vivo myeloid sarcoma model, all mice in the PBS, control peptide, and control antibody groups succumbed to tumor burden by 30 days post-inoculation. The single day 7, 3 nmol dose of combined anti-VIP mAbs resulted in improved survival equivalent to that achieved with 3 nmol ANT-308 (p<0.001 compared to PBS control). Moreover, treatment of P815-bearing DBA/2 mice with a single injection of the pool of anti-VIP antibodies led to initial regression of established s.c. P815 tumors, and sampling blood for cancer-specific T cells using a P815-specific tetramer showed 5-fold expansion of tetramer+ CD8+ T cells at 14 days following anti-VIP antibody administration.

Conclusions: Humanized anti-VIP antibodies are an attractive alternative to peptide-based VIP-receptor antagonists due to their improved pharmacokinetics. We demonstrate the effective use of yeast-display single-chain antibody technology, serial selection for VIP binding, and expression in a producer cell line to identify and generate novel anti-VIP human antibodies. Our initial results show the three mAbs identified here bind VIP in bead-based flow cytometry and SPR assays, increase human T cell activation in vitro, and improve survival in a mouse mastocytoma model. We are continuing to characterize the mAbs and test their ability to improve the anti-tumor activity of T cells in additional pre-clinical AML models using repeated dosing strategies.