Metabolic Abnormalities Associated with T-Cell Exhaustion in CLL Eµ-TCL1 Murine Model

Abstract

Introduction. T-cell differentiation and immune responses are associated with major metabolic changes that control cell fate upon antigen exposure. Chronic lymphocytic leukemia (CLL) is characterized by a marked T-cell dysfunction that plays a major role in disease progression and failure of adoptive cell therapy. A handful of recent studies have demonstrated mitochondrial and metabolic defects in the T-cell compartment in CLL patients. However, to date, no studies have been done to validate such metabolic abnormalities in the Eµ-TCL1 model, the most established and reliable murine model for studying CLL. Therefore, we hypothesized that CD8+ T-cells derived from the Eµ-TCL1 adoptive transfer model exhibit metabolic and mitochondrial disturbance that might contribute to the overall dysfunctional state of these cells. Methods. To investigate the influence of leukemic burden on the differentiation phenotypes, metabolic state and gene expression profile of CD8+ T-cells, C57BL/6 mice adoptively transferred with Eµ-TCL1 leukemic splenocytes were euthanized at varying time points. Once disease was established, spleens were collected for analysis. Mice were divided into 3 groups; early, intermediate, and late disease, based on the peripheral blood leukemic burden at endpoint. Multicolor flow cytometry was used for identification of T-cell phenotypes and measurement of mitochondrial mass, membrane potential, reactive oxygen species (ROS) and cellular glucose uptake. For T-cell activation studies, we cultured splenocytes in the presence of CD3/CD28 dynabeads and 30IU/ml hIL-2 for 72 hours. Nanostring metabolic pathways gene expression analysis was done on splenic CD8+ T-cells isolated from early and late-stage Eµ-TCL1 adoptive transfer mice via fluorescence-activated cell sorting. Differential gene expression was measured using ROSALIND cloud-based software platform. Results. Unsupervised t-distributed stochastic neighborhood embedding (tSNE) and Xshift clustering analyses revealed a gradual enrichment of specific CD8+ T-cell populations along disease progression. These populations upregulated exhaustion-related markers (EOMES, TOX, PD-1 and KLRG-1) and significantly downregulated TCF-1 expression, a transcription factor responsible for cell self-renewal capacity. Further analysis demonstrated a significant downregulation of PGC-1α expression, a master regulator of mitochondrial function, in Eµ-TCL1 CD8+ T-cells with high EOMES and low TCF-1 expression levels (EOMEShi TCF-1lo) compared to EOMEShi TCF-1hi cells. At late disease stage, we noticed significant mitochondrial depolarization, an increase in mitochondrial ROS and marked reduction in glucose uptake capacity in the Eµ-TCL1 CD8+ T-cells. Post in vitro activation, there was a significant increase in Eµ-TCL1 CD8+ T-cell mitochondrial biogenesis, membrane potential and ROS production compared to wild type (WT) controls. However, Eµ-TCL1 cultures showed significant enrichment in EOMEShi TCF-1lo CD8+ T-cell population, that was not detected in the activated WT cells. Interestingly, this population exhibited marked downregulation of PGC-1α, SOD-2 and NRF-2 expression compared to the populations maintaining TCF-1 expression. Moreover, we found that Eµ-TCL1 CD8+ T-cells had increased frequency of effector memory phenotypes compared to more central memory cells in WT cultures. These effector memory populations accumulated significant levels of mitochondrial ROS and depolarized mitochondria compared to the central memory counterparts. Nanostring analysis revealed disease stage-dependent changes in gene expression, where glycolysis was the top differentially expressed pathway in the late versus early-stage CD8+ T-cells as indicated by the gene set global significance score. On the other hand, negative significance scores were assigned to AMPK and autophagy pathways indicating a probable downregulation of these pathways during late phases of the disease. Conclusion. Our results demonstrate mitochondrial and metabolic abnormalities in Eµ-TCL1-derived CD8+ T-cells and point to possible defects in cellular stress signaling response that might lead to accumulation of dysfunctional mitochondria and push the cells towards exhausted states.