Engineering of next-generation chimeric antigen receptors based on alternative binding domains

Abstract

Exogenous expression of chimeric antigen receptors (CARs) on T lymphocytes (CAR-T) and NK cells (CAR-NK) has allowed the design of effective therapies against lymphoproliferative diseases and has recently achieved success against solid tumors. Given their success, several researchers set out to improve the process of engineering these proteins to extrapolate to other types of diseases, as well as to study cell signaling events that occur after antigen recognition. Additionally CAR-like constructs expressing the natural receptor of an antigen of interest bound to the intracytoplasmic region of a CAR, instead of a scFv that recognizes it, yields cytotoxicity results similar to those observed in CAR-T cells. These CAR-like molecules allow cells to recognize antigens in its native conformation, without the restriction of the single antibody chain (scFv) of an archetypal CAR, which is limited to a single epitope. For this purpose, we synthesized lentiviral vectors containing the coding sequences of these molecules, to subsequently transduce T cell lines. Our research group has recently been working on the design of different CAR-like molecules for the treatment of various pathologies, characterized by the existence of well-defined biomarkers that could be used as CAR targets. In this regard, we have developed two different second-generation CAR-like constructs (CD28/zeta), basing the design of the target antigen-binding extracellular domain on the human angiotensin-converting enzyme 2 (ACE2), which is the receptor for the SARS-CoV-2 virus Spike protein, a protein with such high variability that the design of scFv-based CARs, which might not be able to recognize different variants of this protein makes it difficult for CAR-T cells to recognize new variants of this protein.. Our results indicate that these molecules successfully promote T cell activation by binding to the Spike protein, and the strength of stimulation is directly dependent on both the affinity against the antigen and the level of CAR expression. Additionally, we recently developed a novel antigen recognition system based on modular scFv, allowing T cells to target multiple antigens using a single signaling CAR backbone. This approach enabled targeting CAR-T cells to tumor biomarkers such as CD19 and CD20, commonly found in leukemia and non-Hodgkin lymphomas, and produced an activation pattern in T cells similar to that of ACE2-based CAR-like constructs. Furthermore, we compared signaling in these CAR-T cells with that of TCRzeta-CD28 co-stimulation. Phosphoproteomic assays revealed that signaling events following CAR-antigen engagement are notably diminished compared to a physiological TCR response, both in the short and long term, potentially explaining some known side effects of this therapy. Our goal was to establish a new allogenic cell platform for easy-to-handle T and NK CAR therapies. We focused on an unexplored NK-like cell line, termed YT, which is easy to maintain in vitro without special cytokine supplementation, enables high lentiviral transduction, and exhibits a phenotype similar to that of primary peripheral NK cells (CD3- CD16± CD56+). Although YT cells display natural cytotoxicity against B cell lymphoma cells, this activity is further enhanced by transduction of anti-CD19 CARs, highlighting the suitability of this novel cell platform for next-generation CAR-NK cell therapies.