Characterisation of transgenic pigs expressing a human T cell-depleting anti-CD2 monoclonal antibody.

Background: Pig islet xenotransplantation is a potential treatment for type 1 diabetes. We have shown that maintenance immunosuppression is required to protect genetically modified (GM) porcine islet xenografts from T cell-mediated rejection in baboons. Local expression of a depleting anti-CD2 monoclonal antibody (mAb) by the xenograft may provide an alternative solution. We have previously reported the generation of GGTA1 knock-in transgenic pigs expressing the chimeric anti-CD2 mAb diliximab under an MHC class I promoter (MHCIP). In this study, we generated GGTA1 knock-in pigs in which MHCIP was replaced by the β-cell-specific porcine insulin promoter (PIP), and compared the pattern of diliximab expression in the two lines.

Methods: A PIP-diliximab knock-in construct was prepared and validated by transfection of NIT-1 mouse insulinoma cells. The construct was knocked into GGTA1 in wild type (WT) porcine fetal fibroblasts using CRISPR, and knock-in cells were used to generate pigs by somatic cell nuclear transfer (SCNT). Expression of the transgene in MHCIP-diliximab and PIP-diliximab knock-in pigs was characterised at the mRNA and protein levels using RT-qPCR, flow cytometry, ELISA and immunohistochemistry. Islets from MHCIP-diliximab and control GGTA1 KO neonatal pigs were transplanted under the kidney capsule of streptozotocin-diabetic SCID mice.

Results: NIT-1 cells stably transfected with the PIP-diliximab knock-in construct secreted diliximab into the culture supernatant, confirming correct expression and processing of the mAb in β cells. PIP-diliximab knock-in pigs showed a precise integration of the transgene within GGTA1. Diliximab mRNA was detected in all tissues tested (spleen, kidney, heart, liver, lung, pancreas) in MHCIP-diliximab pigs, but was not detectable in PIP-diliximab pigs. Likewise, diliximab was present in the serum of MHCIP-diliximab pigs, at a mean concentration of 1.8 μg/mL, but was not detected in PIP-diliximab pig serum. An immunohistochemical survey revealed staining for diliximab in all organs of MHCIP-diliximab pigs but not of PIP-diliximab pigs. Whole genome sequencing (WGS) of a PIP-diliximab pig identified a missense mutation in the coding region for the dixilimab light chain. This mutation was also found to be present in the fibroblast knock-in clone used to generate the PIP-diliximab pigs. Islet xenografts from neonatal MHCIP-diliximab pigs restored normoglycemia in diabetic immunodeficient mice, indicating no overt effect of the transgene on islet function, and demonstrated expression of diliximab in situ.

Conclusion: Diliximab was widely expressed in MHCIP-diliximab pigs, including in islets, consistent with the endogenous expression pattern of MHC class I. Further investigation is required to determine whether the level of expression in islets from the MHCIP-diliximab pigs is sufficient to prevent T cell-mediated islet xenograft rejection. The unexpected absence of diliximab expression in the islets of PIP-diliximab pigs was probably due to a mutation in the transgene arising during the generation of the knock-in cells used for SCNT.