ECM degrading bacteria boost CAR T cell therapy in solid cancers
Institution: Barts Cancer Institute, Queen Mary University of London
Corresponding Researcher: Oliver Pearce
Data Link(s): NA
Keyword(s): NA
Summary
BACKGROUND. Triple negative breast cancer (TNBC) is the most deadly subtype of breast cancer. However, only about 20% of TNBC patients respond to immunotherapy. Poor immune cell infiltration into the tumour epithelium has been shown to drive resistance to immunotherapy. The extracellular matrix (ECM) has been shown to inhibit effector immune cell infiltration into the tumour epithelium and is thus a main driver of immunotherapy resistance.1 Targeting of the tumour ECM has been limited by on-target, off-tumour toxicities. To increase tumour specificity, Neobe Therapeutics have engineered a tumour-homing bacteria strain to degrade several ECM-molecules. In this study, we will employ a CAR T cell product developed by Leucid Bio to characterise the effect of these bacterial therapeutics on anti-cancer immunity. To this aim, we have developed a patient sample-based, decellularized tissue model of TNBC, which maintains the immunosuppressive properties of the tumour and supports complex co-cultures of cancer cells, CAR T cells and bacteria. In this model we can measure ECM target-degradation as well as CAR T cell cytotoxicity, localisation and movement.
METHODS. We first characterised effector immune cell infiltration, ECM architecture and the deposition via immunohistochemistry and immunofluorescence. Patient samples are cut into 3D discs, decellularized and repopulated with established TNBC cell lines. These tissue models are sequentially treated with bacteria therapeutics and CAR T cells.
RESULTS. Our decellularised tissue model retains the tissue architecture and ECM deposition of the patient sample it was derived from. In a library of 30 primary TNBC tissues, we then show that the localization of a series of ECM-molecules associate with poor immune cell infiltration. We then use a series of engineered bacteria strains to selectively degrade our ECM-targets in our 3D tumour model. When we co-treat our 3D tumour model with therapeutic bacteria and CAR T cells, we observe an increase in CAR T cell killing relative to the levels of our ECM-targets in the tissue. In addition, we observed an increase in CAR T cell infiltration when our 3D tumour model is co-treated with therapeutic bacteria.
CONCLUSIONS. We have developed a decellularized tissue model of TNBC which recreates the immunosuppressive features of the ECM in patients. In this model, we are able to characterise the cytotoxicity, localisation and motility of T cells. We have also shown that degrading ECM-molecules associated with immune exclusion increases CAR T cell infiltration and cytotoxicity in our model.
