Finding the Keys to the CAR: Identifying Novel Target Antigens for T Cell Redirection Immunotherapies

Abstract

Oncology immunotherapy has been a significant advancement in cancer treatment and involves harnessing and redirecting a patient's immune response towards their own tumour. Specific recognition and elimination of tumour cells was first proposed over a century ago with Paul Erlich's 'magic bullet' theory of therapy. In the past decades, targeting cancer antigens by redirecting T cells with antibodies using either bispecific T cell engagers (BiTEs) or chimeric antigen receptor (CAR) T cell therapy has achieved impressive clinical responses. Despite recent successes in haematological cancers, linked to a high and uniformly expressed CD19 antigen, the efficacy of T cell therapies in solid cancers has been disappointing, in part due to antigen escape. Targeting heterogeneous solid tumours with T cell therapies will require the identification of novel tumour specific targets. These targets can be found among a range of cell-surface expressed antigens, including proteins, glycolipids or carbohydrates. In this review, we will introduce the current tumour target antigen classification, outline existing approaches to discover novel tumour target antigens and discuss considerations for future design of antibodies with a focus on their use in CAR T cells.

Keywords: Bi-specific T cell Engager (BiTE); antigen selection; antigenic screen; cell surface antigen; chimeric antigen receptor T cells (CAR T); glycomics; immunotherapy; lipidomics; oncology; phage display; proteomics; target antigen.

Figure 1

The generations of chimeric antigen receptors (CAR). The CAR designs differ based on the intracellular signalling tail. First generation CARs feature only the transmembrane domain fused to CD3ξ, these proliferated poorly in vivo. Second and third generation CARs differ in the inclusion of one (second generation) or two (third generation) costimulatory domains—these are commonly CD28 or CD137 (4-1BB).

Figure 2

Common Antibody and antibody fragments which can be generated to validate target antigens. (A) Upon antigenic challenge, full sized dual chain antibodies are produced in model systems such as rodents and humans. The antibody fragment generated is a single chain variable fragment (scFv). (B) Camelids and sharks produce single, heavy chain only antibodies, with a nanobody antibody fragment. (C) The antibody fragments discussed in this review include diabodies—two fused scFvs or nanobodies of the same antibody, and bi-specific antibodies made of two fused scFvs with different specificities.