Abstract 3519: Appropriate crosslinking is required for co-stimulatory activity of human CD27 antibody in a transgenic mouse model

Abstract

CD27 is a member of TNFR superfamily. It constitutively expresses on the majority of T cell and a subset of NK cells, playing key roles in T cell activation and survival and in NK cell proliferation and cytotoxicity upon interaction with ligand CD70. Some antibodies to mouse CD27 have been reported that display agonistic and anti-tumor activities while other mAbs had less anti-tumor activity and were depleting. We hypothesized that differences in these antibodies may be due to Fc receptor engagement, as has recently been shown for the adjuvant and anti-tumor activities of agonistic mouse CD40 mAbs, which is also member of TNFR superfamily. We have developed and previously described a human anti-human CD27 antibody (1F5) and a human CD27 transgenic mouse model (hCD27-Tg) to explore the therapeutic potential of targeting CD27. In this study, we examined the effect of modifying the constant regions of the 1F5 mAb on its ability to enhance antigen specific T cell responses. With the original 1F5 hIgG1 as template, a panel of 1F5 variants was made including 1F5 mIgG1, 1F5 mIgG2a, 1F5 mIgG1D265A and 1F5 hIgG1N297S using molecular cloning techniques. All of the variants retained equal binding to hCD27 as shown by ELISA and flow cytometry studies. In addition, Biacore analysis confirmed the expected pattern of binding to human and mouse FcαRs. Co-injection of 1F5 or its variants with ovalbumin enhanced antigen-specific CD8 T cell response to different extents, as detected by SIINFEKL-specific IFNα-ELISPOT and ICS. The 1F5 mIgG1 induced the highest number of IFNα-producing CD8+ cells, whereas 1F5 mIgG2a was very weak at enhancing the CD8 T cell response. The hIgG1 version of 1F5 was intermediate in activity. Introduction of the D265A mutation that disrupts FcαRs binding into the mIgG1 eliminated the co-stimulatory function of 1F5. Similarly, the 1F5 hIgG1N297S also showed reduced activity compared to the original 1F5 hIgG1. The isotype-specific effects on our anti-hCD27 mAb are surprisingly consistent with the findings described for the agonistic anti-mCD40 mAbs, and imply that engagement of the inhibitory Fcα receptors (FcαRIIb) is driving the co-stimulatory activity in this model. Interestingly, the 1F5 hIgG1 triggered a significant T cell response, despite the lack of FcαRIIb binding by Biacore analysis. The effect of these variants on anti-tumor activity in hCD27 transgenic mice is currently being investigated. The 1F5 hIgG1 mAb (CDX-1127) is currently undergoing clinical evaluation in a phase 1 trial of patients with advanced cancers. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 3519. doi:1538-7445.AM2012-3519