Engraftment Fitness of Human Central Memory-Derived Effector CD8⁺ T Cells In Immunodeficient Mice

Abstract 1019

Development of T cell products that have engineered specificity for CD19 has broad application to adoptive transfer therapy for B-lineage lymphoma and leukemia. Clinical studies have demonstrated the safety and feasibility of T cell transfer as a therapy for patients. But the potency of this strategy has proven challenging, primarily due to issues relating to a lack of persistence of the adoptively transferred cells in patients. The repertoire of memory T cells is heterogeneous with respect to phenotypic, functional, and epigenetic attributes. Memory T cells are divided into sub-populations of 1) effector memory (T_EM) cells that distribute to tissue beds and exhibit immediate cytolytic effector functioning, and 2) central memory (T_CM) cells that home to lymph nodes based on CD62L/CCR7 expression and are capable of extensive proliferative activity upon re-encountering antigen. Thus the cell-intrinsic programming of distinct memory T cell subtypes, such as T_EM and T_CM, likely dictate divergent fates of their derived effector cells. To address this important issue, a clear functional dichotomy between T_CM- and T_EM-derived CD8⁺ CTLs was recently delineated in a nonhuman primate model, where it was found that virus-specific CD8⁺ CTL clones derived from T_CM, but not T_EM precursors, establish persistent and functional memory following adoptive transfer. Here, we extended these studies to human effector T cells using CMV as antigen model system to investigate the engraftment of human CMVpp65-specific CD8⁺ effector T cells derived in vitro from either sort purified CD45RO⁺CD62L⁺ T_CM or CD45RO⁺CD62L^- T_EM precursors in NOD/Scid IL-2RγC^null (NOG) mice. T_CM-derived effector cells (T_E(CM)) and T_EM-derived effector cells (T_E(EM)) were adoptively transferred (i.v) into NOG mice reconstituted with human IL-15 and T cell levels in circulation were evaluated at different time points by FACS. 20% CD8⁺ T_E(CM) and 3% CD8⁺ T_E(EM) were detected on day 14. Then after, engraftment of the CD8⁺ T_E(CM) remained at a steady state of approx 2% of circulating mononuclear cells for 100 days while T_E(EM) remained at or below the level of detection, indicating that T_E(CM) were superior in their ability to engraft in response to IL-15 as compared to T_E(EM) after adoptive transfer (P<0.05). The long-term (100 days) persisting CD8⁺ T_E(CM), harvested from primary recipient mice were found to be capable of engrafting secondary recipients. TcR Vβ analysis of persisting cells demonstrated that CD8⁺ T_E(CM) engraftment was polyclonal, suggesting that homeostatic engraftment fitness is a general feature of these cells_.

To delineate the mechanism(s) by which T_E(CM) exhibit superior in vivo engraftment, T_E(CM) and T_E(EM) were first labeled with CFSE before in vivo administration. CFSE profiles appear that the T_E(EM) proliferated more extensively than T_E(CM) early after adoptive transfer as indicated by the percent of cells which diluted CFSE on day 9 (i.e., 80% vs. only 25%, respectively). However, using D₂R cleavage as a measure of caspase activity as a surrogate for apoptosis, 5.8% of engrafting T_E(CM) were positive for activated caspase activity compared to 31.6% of T_E(EM), suggesting that in NOG mice both CD8⁺ T_E(CM) and T_E(EM) proliferate in response to IL-15 whereas T_E(CM) are intrinsically resistant to caspase activation and apoptosis.

We also evaluated the antigen specific responsiveness of engrafted cells. Weekly infusions of irradiated pp65⁺/A2⁺ LCL as antigen significantly augmented the levels of circulating CD8⁺ T_E(CM) as compared to no antigen stimulation (P<0.05), whereas CD8⁺ T_E(EM) did not respond to antigen challenge. Moreover, when CMVpp65 specific CD8⁺ T_E(CM) or T_E(EM) were infused into CMVpp65⁺ tumor bearing mice, tumor cells progressed in mice receiving T_E(EM) at a rate similar to untreated control mice over a ten day observation period, whereas T_E(CM) significantly controlled tumor progression (P<0.05), indicating that CD8⁺ T_E(CM) but not T_E(EM) are able to mediate an anti-tumor response. Together these studies confirm that human CD8⁺ effector T cells derived from T_CM precursors are capable of persistence after infusion, can proliferate in in vivo in response to antigen, can mediate an anti-viral or anti tumor response, and are likely the preferred T cells for antigen specific anti-tumor adoptive T cell therapy .