First Evidence of Dysfunctional Antigen-Specific T Cell Responses in Experimental CLL As a Model for Studies of Autologous T Cell-Based Therapies

Background - Tumor immunosuppression is a major cause for treatment failure and disease relapse, both in solid tumors and in leukemia. New T cell mediated therapeutic modalities (such as CAR T cells, bispecific antibodies and immune checkpoint blockade) can be highly effective, but depend on autologous T cell functionality. In chronic lymphocytic leukemia (CLL), ex vivo patient studies have shown that T cells are skewed toward a terminally differentiated, dysfunctional phenotype, which may contribute to the disappointing results of T cell therapies in this disease. Recently, it has been shown that for CAR T cell efficacy in CLL, the presence of memory subsets both prior to CAR T cell generation and in the infusion product, is instrumental for CAR T cell persistence and complete responses (Fraietta et al., 2018). However, it is currently unknown how the presence of CLL impacts T cell skewing and whether it affects acute antigen-specific immune responses in vivo. As dynamic immune studies are limited in the human setting, we investigated whether acute antigen-mediated T cell responses are affected by CLL using the TCL1 adoptive transfer mouse model. Earlier T cell studies in this model revealed similar changes as seen in the human disease, both with respect to gene expression profiles as well as phenotypic and functional characteristics.

Methods - C57BL/6J (CD45.2⁺) mice were injected i.p. with 20x10⁶ TCL1 transgenic splenocytes (CD45.2⁺) or PBS. During development of CLL mice were monitored for T cell subset differentiation. When CLL mice reached over 70% CLL in blood, 50.000 OT-I cells (CD45.1⁺), which specifically recognize ovalbumin) were adoptively transferred (i.v.) into all mice, directly followed by infection with 100.000 PFU mCMV-OVA. This model allowed us to study the endogenous (CD45.2⁺) T cell compartment during CLL development, as well as the acute response of naïve, antigen-specific (CD45.1⁺) T cells that had not been influenced by CLL prior to antigen exposure. Seven days after infection the mice were sacrificed and splenocytes were analyzed by flow cytometry either directly or after re-stimulation with OVA-peptide to assess cytokine production.

Results - Using this adoptive transfer model of CLL, we found that CLL mice showed a decrease in naïve CD8 T cells, an increase of antigen-experienced cells, and a reversed CD4/CD8 ratio compared to control mice, as reported earlier (McClanahan et al., 2015). At day 7 after injection of naïve OT-I cells and infection with mCMV-OVA, percentages of OT-I cells were similar in CLL and control mice, and all OT-I cells showed an effector phenotype (CD44⁺CD62L^-). However, in this effector OT-I pool CLL mice showed an increased frequency of KLRG1⁺CD127^- short-lived effector cells (SLEC), while KLRG1^-CD127⁺ memory precursor effector cells (MPEC) were decreased compared to control mice. This was associated with an enhanced expression of the effector-associated transcription factor T-bet, and reduced expression of the memory-associated transcription factor Bcl-6 in total OT-I cells and within the OT-I SLEC and MPEC populations of CLL mice. Since poor responses to CAR T cell therapy are associated with effector phenotypes and higher rates of glycolysis (Fraietta et al., 2018), and Bcl-6 directly represses glycolysis (Oestreich et al., 2014), we analyzed glucose uptake by OT-I cells. In line with reduced expression of Bcl-6, we found increased levels of glucose uptake in CLL derived OT-I cells. Despite the skewing towards a more short-lived effector phenotype of the CLL derived OT-I cells, in vitro re-stimulation with OVA peptide resulted in decreased production of IFN-γ and decreased degranulation as measured by CD107a (LAMP-1) expression compared to control derived OT-I cells.

Conclusion - Our findings show that in this mouse model CLL development skews the T cell compartment towards more antigen-experienced cells. In addition to this, our results suggest that there is a CLL-antigen independent effect on acute antigen-specific immune responses. This CLL mediated effect drives T cells towards an effector phenotype, that is functionally impaired. This study provides clues for better understanding T cell responses in CLL, and may lead to new strategies to improve T cell mediated therapies in CLL, which are currently being exploited.