Non-invasive imaging of the anti-tumor immune response

点击翻译按钮获取中文摘要

Project Summary Immunotherapy using checkpoint blockade has revolutionized cancer treatment. The outcome of therapy directly results from changes imposed on the tumor microenvironment (TME) by checkpoint blockade. However, only a subset of patients respond. What controls this disparity is poorly understood. We propose to develop and apply tools that can help differentiate responders from non-responders in different pre-clinical tumor models soon after the start of treatment. We have shown that immuno-positron emission tomography (Immuno-PET) can be used to monitor infiltration status of specific subsets of immune cells, namely T cells and myeloid cells. We use small (~15 kDa) camelid-derived single domain antibodies (nanobodies) that have nM to pM affinity for their targets to perform immuno-PET imaging. Our unique chemical approaches provide imaging agents of unprecedented quality and sensitivity. Even cells that display proteins of relatively low abundance such as CTLA-4 can be clealrly imaged. We have shown in several (syngeneic) tumor models that monitoring the dynamics of cytotoxic T cells in the TME can be used to distinguish early responders from non-responders. This observation has allowed us to stratify animals into responders and non-responders, then excise their tumors, isolate the immune infiltrating cells, and subject these to single-cell RNA sequencing. These data show that the myeloid compartment and the cytokines and chemokines it produces, plays a major role in determining the outcome of anti PD-1 treatment. We propose to expand these initial findings to additional mouse tumor models, given the distinct cells of origin that give rise to them and their differences in susceptibility to immune intervention. This project is aimed at bringing to light key changes that take place in the TME early on, using immuno-PET. Our complementary molecular analyses will help design more effective therapies. Macrophages and DCs in the TME of responders produce CXCL9, a chemoattractant for cytotoxic T cells that helps maintain their activated state. This chemokine is therefore a key player in the outcome of anti-PD-1 therapy. We thus propose to re-engineer the TME by using chemistry to make novel CXCL9-fusion proteins and deliver them to the TME. Single-domain antibodies are perfect candidates for such fusions. Their small size allows excellent tissue penetration and their high affinity ensures efficient delivery to, and retention in, the TME. Imaging the distribution of CXCL9 or its receptor will shed further light on the anti-tumor immune status. We will therefore generate nanobodies as imaging agents specific for such cytokines and their receptors.