Antigen-specific redirection of immune effector cells against GD2-expressing tumors

Extensive research in the last decades has revealed the dynamic role of the immune system in surveillance, recognition and elimination of cancer cells. However, resistant variants can rise and overcome these immune responses by various escape mechanisms. Therefore, huge efforts have been invested in developing of immunotherapies that can overcome these hurdles and allow the specific activation and retargeting of immune cells toward tumor cells. Immunotherapy includes diverse strategies ranging from cytokines to antibodies and their derivatives reaching to engineered immune cells.
One of the very promising immunotherapeutic approaches is based on the engineering of T cells to express chimeric antigen receptors (CARs) which can redirect immune cells to specific antigens leading to T cell activation and subsequent killing of target cells. The CAR technology has shown a strong therapeutic potential in targeting of cancer both in preclinical as well as clinical studies, especially with hematological malignancies, which lead to FDA approval of CD19-specific CAR T cells for the treatment of leukemia. Despite the success of CAR T cells, clinical trials have also revealed various toxicities and adverse events that can be life-threatening for patients. Moreover, the CAR technology still faces many hurdles to achieve an effective targeting of solid tumors due to the structure, immunosuppressive microenvironment and heterogeneity of the disease. Therefore, a novel modular and switchable CAR platform, called the UniCAR system, was developed in the group of Prof. Bachmann in order to address these obstacles. The UniCAR system consists of UniCAR T cells that cannot bind surface antigens. Instead, UniCARs recognize an epitope (E5B9) which is derived from the nuclear protein La-SS/B. Therefore, UniCAR T cells can only be redirected via target modules (TMs) that have the E5B9 tag on one hand and an antigen-binding domain on the other hand. These TMs provide a bridge with tumor cells that allow the activation of UniCAR T cells. Once these TMs are eliminated, the UniCAR T cells can no more be activated, and thus they are switched off. This approach provides not only a safety switch but also a flexible platform for multi-targeting of diverse antigens by using various TMs with different antigen specificities. Alternatively, the UniCAR system can be applied on NK cells or NK cell lines, which have a natural anti-tumor response. Cell lines like NK-92 are of especial interest because they can be used allogenically, and thus provide an off-the-shelf therapy that might reduce the cost and time of therapy development.
For my thesis work, disialoganglioside GD2 was selected as a target for the UniCAR system. GD2 is overexpressed on many tumors including neuroblastoma, Ewing’s sarcoma, melanoma, osteosarcoma and others. In fact, GD2 is considered as one of the priority antigens to be targeted for cancer therapy according to a pilot project of the national cancer institute. Moreover, targeting GD2 with CAR T cells has shown very positive clinical outcome in previously published clinical trials. However, as other tumor-associated antigens, GD2 has also some limited expression on normal tissues including some regions of the central nervous system and peripheral nerves. Therefore, using a safety switch like the UniCAR system emerges as a necessary step to assure better controlling of any on-target/off-tumor side effects.
Driven by these facts, several formats of anti-GD2 TMs were developed in order to redirect UniCAR T cells to GD2-expressing tumors. Three TMs were designed based on a single chain fragment variable (scFv) connected to the E5B9 epitope. However, they differed in the orientation of the variable light and variable heavy chain domains and their linker components in order to find the best functional conformation. In vitro functional assays showed that all of the three TMs were able to redirect UniCAR T cells toward tumor cells leading to efficient tumor cell killing and release of cytokines in an antigen-specific and TM-dependent manner. Furthermore, the TMs showed dose-dependent killing with half-maximal effective concentration (EC50) in the picomolar range. As all the TM formats showed comparable results in vitro, further in vivo studies were restricted to one TM. This anti-GD2 TM was able to activate UniCAR T cells to eradicate GD2-positive tumor cells in experimental mice. Furthermore, the TM was modified and radiolabeled with 64Cu, in order to investigate its pharmacokinetic properties and biodistribution in tumor-bearing mice. PET analysis of the radiolabeled anti-GD2 TM showed enrichment in the GD2-expressing tumors with blood elimination half-life of less than one hour, which makes it a suitable key for a fast safety switch of UniCAR T cells.
In contrast to UniCAR T cells, the UniCAR NK-92 cell line provides an-off-the shelf therapy that can be expanded for allogenic use. Since these cells are usually irradiated before infusion into patients, their life-span as well as the possibility of side effects is reduced. Therefore, we have further created an IgG4-based anti-GD2 TM with an extended half-life that fits to the life-span of irradiated NK-92 cells. PET imaging of the radiolabeled IgG4-based TM showed an increase in the half-life of about 24 folds in comparison to the scFv TM format. Further testing has shown that UniCAR NK-92 cells are functional with both scFv- and IgG4-based TM formats leading to specific killing of GD2-expressing tumor cells as well as secretion of pro-inflammatory cytokines in vitro. In addition, UniCAR NK-92 showed specific killing of tumor cells in vivo when combined with the anti-GD2 TM.
In summary, we have shown that the UniCAR system can be used to redirect both T cells and NK-92 cells against GD2-expressing tumors in vitro as well as in vivo in an antigen-specific and TM-dependent manner. The UniCAR system allows an on/off safety switch as well as fine controlling of the activity of UniCAR T/NK-92 cells via titration (dosing) of the anti-GD2 TMs. Furthermore, it provides a flexible platform that allows the use of several antibody formats for an effective and safe targeting of cancer.