Chimeric antigen receptor (CAR) T-cell therapy has emerged as a promising cancer immunotherapy, but many patients still fail to respond to therapy. T-cell exhaustion is hypothesized to be a major driver of therapy failure, and strategies to non-invasively monitor this phenotype could (1) provide insight into the kinetics of its emergence and persistence (2) be used for early prediction of therapy failure and (3) predict response to interventions that reverse T-cell exhaustion. Existing strategies for blood-based monitoring can give a measure of T-cell expansion and persistence, but exhaustion often emerges at the site(s) of disease and could vary from site to site suggesting that these strategies are unlikely to provide accurate phenotypic information. Here we propose using positron emission tomography (PET) reporter genes in combination with a synthetic gene circuit for monitoring of CAR T-cell viability, location, and exhaustion in vivo. The PET reporter genes herpes simplex virus type 1 thymidine kinase (HSV1-tk) and dopamine type 2 receptor (D2R) are useful tools for imaging cell state since their expression can be non-invasively, repeatedly, and independently monitored by PET upon administration of their cognate radiolabeled substrates [18F]FHBG and [18F]FESP. We thus propose engineering CAR T-cells with both a constitutively expressed PET reporter gene (D2R) for monitoring of viability and location and a second PET reporter gene (HSV1-tk) whose expression is linked to T-cell exhaustion. Since there exists no single marker of T-cell exhaustion, we present a two-input gene circuit using the Clustered Regularly Interspaced Short Palindromic Repeats interference (CRISPRi) system to express HSV1-tk only if PD-1 is expressed and IL-2 expression is lost, a combination unique to T-cell exhaustion.
In Aim 1, we construct a doxycycline- and cumate-controlled CRISPRi circuit to test candidate guide RNAs for efficient CRISPRi.
In Aim 2, we replace the chemically inducible promoters with promoters for PD-1 and IL-2 and assess the circuit?s ability to mirror the exhaustion phenotype of mouse GD2-targeting CAR T-cells in vitro using radiotracer cellular uptake studies. Thereafter we will perform repeated PET imaging of T-cell location and exhaustion in mouse models of GD2+ osteosarcoma and melanoma to better understand the spatiotemporal kinetics of exhaustion.
In Aim 3, we assess whether monitoring of T-cell exhaustion enables early prediction of therapy failure and/or predicts response to checkpoint blockade with anti-PD-1 antibody. The training plan will provide the applicant technical skills in PET reporter genes and molecular imaging, multi- input gene circuits, and immunotherapy models as well as professional skills in oral and written communication to facilitate growth as an independent investigator. Training will take place in Stanford University?s highly collaborative and well-resourced research environment. The applicant will be mentored primarily by Dr. Sanjiv Gambhir, a world-renowned expert in molecular imaging and reporter genes, Dr. Crystal Mackall, a leader in CAR T-cell therapy and T-cell exhaustion, and Dr. Stanley Qi, inventor of the CRISPRi platform.
Adoptive T-cell therapy has emerged as a promising cancer immunotherapy with curative potential in liquid tumors, but effective treatment of solid tumors remains a challenge. Strategies to monitor T-cell behavior in vivo could shed light on the mechanisms of failure and guide rational therapy modification, but traditional blood-based monitoring is unable to provide information on T-cell persistence or activity specifically at the site(s) of disease. Here we genetically engineer T-cells with sensors that allow for spatiotemporal monitoring of T-cell location, expansion, persistence, and exhaustion using positron emission tomography with the goal of predicting therapeutic efficacy, identifying the mechanisms of treatment failure, and providing clinicians an actionable window in which to modify the therapy.