Immunoinformatic analysis of Chinese hamster ovary (CHO) protein contaminants in therapeutic protein formulations

Abstract

Over many years, therapeutic protein manufacturers have addressed the issue of product immunogenicity focusing on the active pharmaceutical ingredient. Recently, concerns about immune responses to contaminating host cell proteins (HCPs) have emerged, mainly due to concerns about their contribution to anti-drug immune responses and drug safety. The presence of Chinese hamster ovary (CHO) -derived HCPs contributed to the cancellation of two clinical trials in 2012. CHO cells are the most commonly used mammalian cell expression systems in recombinant protein production. The publication of the widely-used CHO-K1 genome and transcriptome provides an opportunity to analyze CHO HCPs and better understand their potential to drive immune responses. Of greatest concern are immune responses that may be cross-reactive with human T cell epitopes. We analyzed CHO-derived protein sequences from three overlapping datasets. These datasets consisted of (1) expressed; (2) putatively secreted; and (3) experimentally determined CHO HCPs. After arriving at a final subset of 1757 constitutively expressed CHO proteins, we used our T cell epitope prediction tool, EpiMatrix, to explore CHO HCP immunogenicity as defined by their relative epitope abundance. Each protein received an immunogenicity score reflecting T cell epitope content, and the results revealed 26% of the 1757 proteins analyzed were above a pre-defined threshold for potential immunogenicity, containing an abundance of predicted T cell epitopes. A subset of these proteins was then evaluated for epitopes that were substantially different from the human genome. This is the first step in the important process of uncovering the potential for HCPs to contribute to unwanted immunogenicity.