Natural killer (NK) cells are cytotoxic innate lymphoid cells that protect the host from pathogens. NK cells were classically considered part of the innate immune system, participating as a first-line defense against tumor and virus infected cells. However, recent studies have discovered a subset of NK cells with adaptive immune features. This novel subset of NK cells, characterized by downregulation of signaling molecules FcR and expression of NKG2C receptor on cell surface, demonstrated adaptive immune features such as clonal expansion, long-term persistence, and altered effector functions. While recent research has highlighted the significance of adaptive NK cells, there is still limited understanding about how this adaptive NK cell pool is generated. Epidemiological analyses have indicated an association between the presence of adaptive NK cells in humans and seropositivity for human cytomegalovirus (HCMV). However, it has been demonstrated that not all HCMV seropositive individuals have adaptive NK cells, and the co-infection of humans with other viruses makes it difficult to discern the precise contribution of HCMV infection to the development and expansion of the adaptive NK cell pool. In Chapter 2, I investigated the impact of primary CMV infection on the generation of adaptive NK cell pool using rhesus macaque as an animal model. I found that a subgroup of rhesus macaques with naturally acquired rhesus cytomegalovirus (RhCMV) infection possessed FcR-deficient NK cells that resemble phenotypic and functional characteristics of human adaptive NK cells. These cells were not detected in specific pathogen free (SPF) animals, however, experimental RhCMV infection of the SPF animals led to the induction of FcR-deficient NK cells in a strain-specific manner. Serological analysis of non-SPF animals indicated that subclinical infections by other common viruses can contribute to the expansion of this adaptive NK cell pool. NK cells have been shown to be a potential mediator during viral infections, including HIV. While their exact role during the course of HIV infection requires further investigation, recent research has provided insights regarding the impact of HIV infection on NK cell population. Studies have shown that HIV infection alters the distribution and function of NK cell subpopulations, particularly leading to the expansion of unusual CD56neg NK cells. In Chapter 3, I examined the CD56neg NK cell population in HIV-viremic patients, which was reported to be hypo-functional. I found that the majority of CD56neg NK cells found in HIV patients were deficient in the signaling adaptor FcR and were specialized for antibody-dependent effector functions. These FcR-CD56neg NK cells shared characteristics similar to the adaptive NK cells and were hypo-responsive to tumor or cytokine stimulation but highly responsive to HIV-infected cells in the presence of anti-HIV antibodies. This study suggests that despite the previous understanding of CD56neg NK cells as dysfunctional in HIV patients, they may be cytotoxically active during HIV infection and contribute during disease progression. NK cells mediate their cytotoxicity by releasing lytic granules containing granzymes and perforin. Granzymes, expressed in both NK cells and cytotoxic T cells, are a family of serine proteases, which play a crucial role during the elimination target cells by inducing apoptosis. NK cells release several different types of granzymes during their degranulation, thus the killing of a target cell is likely a result of a combined action of different granzymes. Granzyme B (GrB) is the most abundant and well-characterized granzyme, and they were shown to cleave and activate caspases and other cellular proteins that regulate apoptosis. Although numerous studies have revealed the mechanism of GrB-induced cytotoxicity, it is still uncertain whether NK cell cytotoxicity absolutely requires the presence of GrB. In Chapter 4, we generated NK cells that lack the expression of GrB and perforin using CRISPR Cas9 to knockout their genes. These GrB KO and perforin KO NK cells showed stable expression of activating surface receptors, however, GrB KO NK cells showed reduced expression of GrH and GrM compared to control KO. Importantly, I demonstrated that absence of GrB does not affect cytotoxicity of NK cells, while absence of perforin completely disabled NK cell cytotoxicity. These results imply that the role of GrB could be substituted by other types of granzymes. Furthermore, I propose using CRISPR Cas9 to generate specific knockouts of granzymes in human NK cells as a promising model for delineating the specific roles of each granzyme during the NK cell cytotoxicity.