In a Chimeric Antigen Receptor (CAR), the tumor antigen-specificity of an antibody fragment is directly coupled with T cell (co)receptor intracellular domains, which trigger the cytotoxic effector functions of T cells (CD8 T-lymphocytes). T cells can be isolated from human blood buffy coats with ease and genome-engineered ex vivo, upon which they can be adoptively transferred back to patients (and modelled in the lab by mice) that carry a tumor that expresses the antigen for the CAR. This approach of antibody-mediated redirection of T cells to tumor-specific antigens expressed on the surface of tumor cells, is independent from T cell specificity being restricted to MHC-peptide recognition, which is patient-specific and very often selected to be lost in the course of tumor evolution in the face of a patient’s immune system.
However, enormous challenges remain in extending the therapeutic success of CAR-T therapy to solid tumors. The most frequent tumor types are carcinomas, and the heavily immunosuppressive environment that is selected for during cancer evolution up to the point of clinical manifestation, is fundamentally hampering success with adoptive cell therapy, incl. with CAR-T cells. Also, and possibly in part because of this, long-term persistence of CAR-T cells upon primary tumor load clearance has been very difficult to achieve, despite recent progress in mouse models. Most of the CAR-T engineering interventions to overcome this intratumoral immunosuppression that have been studied so far target only one of the multitude of immunosuppressive pathways. However, it is clear from early-stage clinical trials with such engineered CAR-T cells that multiple pathways will need to be tackled at the same time.
Manipulations of CAR-T cells that can overcome multiple key mechanisms of tumor immunosuppression with as few genetic engineering steps as possible are amongst the most sought-after medical biotechnological advances in oncology at this moment.
Glycan synthesis pathways almost always modify a large variety of cell surface proteins rather than just a single one, and the assembled properties of this ‘glycocalyx’ are often what determines its function11. Therefore, manipulation of these pathways most often has very pleiotropic effects on cell surface receptor biology. In the context situated above, we hypothesize that the pleiotropism of glycocalyx manipulation can actually be a strong advantage: one glyco-gene manipulation affects multiple receptor-triggered pathways at the same time. Glycans are known to regulate multiple key physiological aspects of T cell biology, such as T cell development and thymocyte selection, T cell receptor signaling sensitivity, sensitivity to cytokine signals, T cell differentiation, tissue homing and proliferation. We have recently reviewed the current state of knowledge in this field. The challenge of this ERC project is to investigate which CAR-T glycocalyx components can be engineered to enhance therapeutic efficacy.
We have found that MGAT5 knockout in CAR-T cells enhances their functional persistence upon tumor clearance in the highly immunosuppressive SKOV-3 carcinoma model and hence enhances their capability of clearing subsequent tumor rechallenges, with a majority of mice remaining tumor-free. (De Bousser et al., patent application ‘Glyco-engineered CAR-T cells’, WO/2023/111322). This is a very difficult to achieve result with any other known CAR-T manipulation and it provides strong proof of concept that glycocalyx engineering of CAR-T cells deserves in-depth study. Hence, we have defined a programme to build on this finding and to explore a candidate set of further glycosylation engineering concepts in CAR-T cells, to further improve CAR-T therapy of solid tumors.
