Abstract
Current approaches to in vivo bioluminescent imaging (BLI) of T cells utilizing luciferase enzymes may be compromised by poor transduction efficiencies in primary T cells, and low photon emissions. Here we describe a novel and widely applicable approach to in vivo BLI of primary T cells utilizing a membrane-anchored form of the naturally secreted Guassia luciferase (GLuc) enzyme, termed exGLuc. We constructed exGLuc by fusion of the GLuc gene to the coding region of the CD8 transmembrane domain. The resulting exGLuc enzyme is anchored to the cell surface of retrovirally transduced cells. In vitro, cells which expressed the exGLuc enzyme demonstrated a markedly (>9 fold) increased bioluminescent signal when compared to cells which expressed the native GLuc (GLuc), the related Rhenilla luciferase (RLuc), a membrane-anchored form of RLuc (exRLuc), or a green fluorescent protein (GFP)-fire fly luciferase fusion protein. Following injection into SCID-Beige mice, MHC-mismatched C57BL6 T cells transduced to express exGLuc were detected by BLI generating graft versus host disease at an earlier time point and demonstrated a >10 fold increased bioluminescent signal when compared to infused C57BL6 T cells expressing GLuc, RLuc, or exRLuc. We further demonstrate homing of OVA-specific DO11.10 exGLuc+ T cells to A20(OVA) but not A20(GFP) subcutaneous tumors in both SCID-Beige and Balb/c mice. We further successfully applied this strategy to the study of in vivo human T cell homing using a xenogeneic SCID-Beige tumor model. Specifically, we demonstrate that human T cells, retrovirally co-transduced to express both a CD19-specific chimeric antigen receptor (CAR), 19z1, and exGLuc, could be monitored over time, and that these T cells quite rapidly home to subcutaneous CD19+ acute lymphoblastic leukemia (NALM-6) tumors when compared to T cells co-expressing an irrelevant CAR with exGLuc. Furthermore, we were able to demonstrate for the first time specific homing of 19z1/exGLuc+ human T cells to systemic NALM-6 tumor, detecting exGLuc+ T cells in deep tissues including the bone marrow, spleen, and liver. We conclude that the exGLuc enzyme emits a superior bioluminescent signal when compared to other commonly utilized luciferase enzymes; that primary T cells are readily transduced to express the exGLuc enzyme; and that exGLuc+ T cells may be accurately monitored in vivo over time by BLI. Based on these data, we believe that this novel approach to primary T cell in vivo imaging will significantly enhance the study of adoptive T cell immunotherapy in murine models of leukemia and other malignancies.
CARTmath—A Mathematical Model of CAR-T Immunotherapy in Preclinical Studies of Hematological Cancers