Immunological memory is the cornerstone of vaccine development and many cellular immunotherapy strategies and is largely attributed to the responses of conventional T and B cells of the adaptive immune system. However, recent evidence has pushed the boundaries of the classic dogmas, which has prompted a reevaluation of what delineates “immune memory”. Among the cells that have pushed these boundaries is the natural killer (NK) cell, an innate lymphocyte whose capabilities span the borders of innate and adaptive immunity. Most notably, NK cells that recognize cytomegalovirus (CMV) in an antigen-specific manner can undergo prolific clonal expansion and sustain a pool of long-lived memory cells after viral clearance, much akin to adaptive lymphocytes. In addition, NK cells pre-activated with inflammatory cytokines can retain a nonspecific, enhanced effector capacity weeks after the initial activation, a phenomenon more broadly known as trained immunity. This flexibility has contributed to NK cells being a promising form of adoptive cell therapy. Both forms of memory each offer their own advantages, including longevity and potential for heterologous immunity. Both forms of NK memory are also epigenetically programmed, which has become a common defining feature for both trained and adaptive immune memory generally.

Up until now, comparing this common defining feature has been obfuscated by underlying differences in cellular settings (e.g. innate versus adaptive cells). We aim to leverage the unique ability for a single cell type, NK cells, to acquire both antigen-specific and trained immune memory in order to answer two open and related questions in the field: How are the fundamental molecular mechanisms that generate trained immune memory different from classic adaptive immune memory, and how do we use this knowledge to enhance current immunotherapy? We believe that these fundamental mechanisms are reflected in the epigenome, where memory potential is encoded, and therein lies opportunities to reprogram an NK cell in a way that optimizes its therapeutic potential. However, in order to fully co-opt desirable memory NK cell traits, we must fill in the knowledge gap of how epigenetic reprogramming naturally manifests, and ultimately how these changes result in a cellular and functional outcome. As such, our work up until now has provided important insight on antigen-specific NK cell epigenetic responses during mouse CMV (MCMV) infection, which provides a crucial starting point to answer the questions above. Based on chromatin accessibility profiling in antigen-specific NK cells and T cells, we identified the AP-1 transcription factor JunB as a potential common mediator of immune memory formation and/or responses. We hypothesize that common and distinct mechanisms underly antigen-specific and trained NK cell memory programs, and that JunB plays a key role in shaping the epigenetic landscape in both memory NK cells. Using novel genetic mouse models, MCMV infection, and next generation sequencing, we propose:

Aim 1: Identify sustained epigenetic changes of trained NK cells over time. We will establish a chromatin accessibility trajectory atlas of trained NK cell memory, as we have done in antigen-specific NK cells, and bioinformatically interrogate how the stable epigenetic features found in trained NK cells behave in antigen-specific NK cells.

Aim 2: Investigate the epigenetic role of JunB in antigen-specific and trained NK cell memory. We have found that JunB has an early functional role in promoting antigen-specific effector responses, and provide evidence to suggest it may have a role in trained NK cells as well. Using a conditional knockout system that depletes JunB in mature NK cells, we will interrogate how the early absence of JunB alters the open chromatin landscape to identify key epigenomic hotspots for AP-1-mediated activity.