Successes and Challenges of NK Immunotherapy: Breaking Tolerance to Cancer Resistance

By Benjamin Bonavida

Successes and Challenges of NK Immunotherapy: Increasing Anti-tumor Efficacy describes the unique therapeutic applications of NK cells to fight cancers and eliminate the bulk and subset of cancer stem cells responsible for metastasis, relapse and recurrences. The book provides information on the development, engineering, mechanisms of action, response to various preclinical models, and applications in various clinical trials. Sections cover the development of highly engineered cytotoxic NK cells, their mechanisms of action, preclinical and clinical applications, the development and application of CAR-NK cells, and new NK-drug conjugates, also emphasizing that activated NK cells can target and kill highly resistant cancer stem cells.

Written by the leading experts on NK immunotherapy worldwide, this is a valuable resource for researchers, clinicians and members of the biomedical field who are interested in understanding novel and efficient therapies to fight cancers.

• Discusses the unique developmental applications of NK immunotherapy against cancers, which differs greatly from other types of immunotherapies
• Provides up-to-date and highly relevant information through chapters written by the leading researchers in the field
• Presents a significant number of schematic diagrams for easy understanding and reproducibility

• Language: English
• Publisher: Academic Press
• Release date: Jun 23, 2021
• ISBN: 9780128243961

Book preview

Part I

NK cells: General properties

Chapter 1: NK cells and CD8 T cells in cancer immunotherapy: Similar functions by different mechanisms

Cordelia Dunaia; Craig P. Collinsa; Isabel Baraoa; William J. Murphya,b    a Department of Dermatology, University of California, Davis, Sacramento, CA, United States

b Department of Internal Medicine, University of California, Davis, Sacramento, CA, United States

Abstract

NK cells and CD8 T cells are cytotoxic lymphocytes that have critical protective roles against pathogens and cancers. NK cells are conventionally regarded as rapid-acting cells of the innate branch of the immune system, while CD8 T cells act later in infections and in long-term memory responses. Both cells have similar cytotoxic pathways and respond to the same cytokines. We discuss how NK cells and T cells break the mold of innate and adaptive immunity with evidence that NK cells exhibit memory-like responses and memory bystander T cells rapidly respond to cytokine signals in heterologous challenges. A great deal is still unknown about coregulation of NK cells and T cells and the factors that are necessary for preserving long-term memory responses; however, light is being shed on the importance of inhibitory receptors for preventing cells from activation-induced cell death (AICD). Above all, there is a fine balance between having a protective immune response at the time of challenge and preventing immune-mediated pathology, while also generating long-lived memory cell populations with enhanced functions for the future. There are many exciting new therapeutic approaches aimed at utilizing NK cells and T cells to their full advantage for potent and long-term antitumor responses.

Keywords

NK cells; Bystander T cells; Cancer immunotherapy; Adaptive NK cells

Abbreviations

ADCC 

antibody-dependent cell-mediated cytotoxicity

AICD 

activation-induced cell death

AML 

acute myeloid leukemia

APC 

antigen presenting cell

CML 

chronic myeloid leukemia

CMV 

cytomegalovirus

GVHD 

graft-versus-host-disease

GVT 

graft-versus-tumor

HSCT 

hematopoietic stem cell transplant

ILC 

innate lymphoid cell

KIR 

killer-cell immunoglobulin-like receptor

LCMV 

lymphocytic choriomeningitis mammarenavirus

MAIT 

mucosal-associated invariant T

MCMV 

murine cytomegalovirus

MHC 

major histocompatibility complex

NHL 

non-Hodgkin’s lymphoma

NHP 

nonhuman primate

NK 

natural killer

PBMC 

peripheral blood mononuclear cell

TiNK 

tumor infiltrating NK cells

TME 

tumor microenvironment

Acknowledgments

Many thanks to Dr. Sean Judge, Michelle Bagood, Dr. Lam Khuat, Logan Vick, and the Murphy lab journal club group for helpful input and discussions. Work was supported by R01-HL-140921.

Conflict of interest

No potential conflicts of interest were disclosed.

Introduction

NK cells and T cells are two key lymphocytes whose evolutionary lineages diverged with the emergence of somatic rearrangement of antigen receptors [1]. NK cells form a distinct lineage of the recently discovered family of innate lymphoid cells (ILCs, type 1), and are often described as the cytotoxic arm of the ILCs, or the innate immune counterpart to CD8 T cells [2]. The cytotoxic activity of NK cells was first observed in a phenomenon termed as hybrid resistance—this involves rejection of hematopoietic parental cells by F1 offspring which was different than the previously observed acceptance of parental tissue grafts. The results showing that hybrid resistance is mediated by a radiation-resistant cell type were published in 1971 [3]. The discovery of NK cells and the coining of their name were published in 1975 with evidence of their spontaneous (nonsensitized) killing of tumor cells [4,5]. Currently, our understanding of ILCs and memory T cells is expanding as these lymphocytes blur the line of the innate and adaptive immunity. ILCs elicit appreciation for the immune system’s ability to fill a wide range of functional and anatomical niches and to station fast-responding sentinels to protect against pathogens [1].

Both NK and CD8 T cells are cytotoxic lymphocytes armed to fight cancer and pathogens, but their recognition, specificity, sensitivity, education, and memory mechanisms have key differences. Currently, cancer immunotherapy is mainly focused on CD8 T cells (anti-PD-1, anti-CTLA-4, CARs), but issues of drug toxicity and efficacy occur [6,7]. Among other factors, cancer heterogeneity in the patient population and immune escape often result in lack of response to therapies and relapse [8]. NK cells have unique immunological features, and their clinical use is an attractive alternative to T cells. In contrast to the single dominant T cell receptor (TCR) on T cells, NK cells present a diverse repertoire of activating and inhibitory receptors [9] that act in balance to regulate functional activity. In addition, NK cells undergo an education process (or licensing) involving major histocompatibility complex (MHC) I molecules during development and maturation [10], which confers self-tolerance and also impacts the strength of their responses against cancer. NK cells can also acquire target recognition and functional memory after encountering target cells (i.e., viral infection in preclinical and clinical models) [11–14]. A goal of the ongoing research is to develop patient-customized approaches that merge the power of NK cells (innate immunity) and T cells (adaptive immunity), along with refined targeting and successful delivery of activated NK cells and specific memory CD8 T cells for tumor eradication. The focus of this review is on the distinct but complementary roles of cytotoxic lymphocytes and applications in immunotherapy. We propose that developing clinical applications that play to the strengths of each cell type and use a combination of approaches (as happens in a natural immune response) is a promising strategy for cancer immunotherapy.

Blurred line of innate and adaptive immunity: Adaptive NK cells and bystander T cells

NK cells and cytotoxic CD8 T cells have many traits in common and optimizing their antitumor cytotoxicity has been a focus of cancer immunotherapy for decades. Both cell types respond to type I interferons, IL-2, IL-12, IL-15, IL-18, and IL-21, and kill target cells via natural killer group 2 member D (NKG2D)-mediated activation and degranulation of perforin and granzyme. One of the best-studied scenarios for NK cell and T cell interaction is during viral responses—both NK and CD8 T cells specialize in killing virally infected cells, complementing each other with NK cells being able to kill MHC I-negative cells which evade T cells. Another major distinction between the cell types is that NK cells are fast responders, whereas naïve CD8 T cells need at least 1–4 days to become primed by antigen-presenting cells (APCs) in lymphoid tissue, then expand, traffic to, and kill targets. In these ways, there is a division of labor in terms of timing and method of killing virally infected cells. However, bystander T cells (memory T cells responding to inflammatory signals, i.e., cytokine stimulation in the absence of TCR engagement) have been shown to facilitate cytotoxicity and play a role in heterologous protection in pathogen responses or collateral tissue damage in autoimmune diseases [15] (Fig. 1). Also acting outside of the classic innate immunology definition, NK cells can be adaptive—in the sense that they have receptors against specific pathogens and preserve a memory pool with enhanced secondary responses [i.e., cytomegalovirus (CMV)-specific receptors Ly49H and NKG2C in mice and humans, respectively] [11].

Fig. 1

Fig. 1 Blurred lines of adaptive and innate immunity.

Adaptive NK cells have been described in chemical hypersensitivity and viral responses, as well as being generated by cytokine activation [16–19]. Mouse cytomegalovirus (MCMV)-specific Ly49H+ NK cells have been shown to exhibit the classical components of immune memory: antigen specificity, expansion, contraction, and secondary responses [17]. In DAP12-deficient neonatal mice (lacking Ly49H signaling) challenged with lethal dose MCMV, adaptive NK cells were more protective than naïve NK cells when given in equal numbers [17]. It took 10 times as many naïve NK cells to provide the same level of protection, which is evidence of a phenotypic advantage [17]. There is significant research on long-term adaptive NK cell responses in nonhuman primate (NHP) and humans with interest in utilizing their antigen-specificity and enhanced effector function in clinical settings [20–22]. Fascinatingly, NK cells mediated antigen-specific killing of dendritic cells up to 5 years after vaccine challenge in NHP [20]. In humans, an extremely diverse repertoire of NK cells has been discovered by examining subsets with mass cytometry. NKG2C+ NK cells are increased in CMV-seropositive healthy individuals and aviremic HIV-1 patients [23–25]. It has also been discovered that NKG2C+ NK cells expand during CMV reactivation posthematopoietic stem cell transplant (HSCT) and this repertoire change is long lasting (elevated frequency detected at 1 year) [22]. Activated NKG2C+ NK cells also contain a significant mature (CD57+) subset, which preferentially expands during acute CMV infection and can be detected years later [26]. It is not known if CD57 is associated with senescence in vivo or if it is just a marker of cells that have divided multiple times. This CD56dimCD57+ NKG2C+ NK cell subset has been termed adaptive and its expansion has been associated with reduced leukemia relapse suggesting a cross-protective effect of these activated NK cells. However, the risks of CMV reactivation as a cause of nonrelapse mortality remain [21,27].

Bystander T cells have been shown to play beneficial roles in heterologous infections and antitumor responses by exhibiting NKG2D-mediated cytotoxicity, proliferation, and IFN-γ production [28–30]. It is postulated that because memory T cells express distinct chemokine receptors, they are posed to respond to inflammatory signals, unlike naïve cells which remain in circulation and pass through lymph nodes [31]. It has been shown that memory, but not naïve, T cells specific for ovalbumin peptide (therefore nonantigen-specific and acting as a bystander) can be found in virus-infected lungs [32]. The extent of cross-reactivity of TCRs for different MHC-presented peptides versus pure non-TCR-mediated responses is the subject of ongoing research. Surprisingly, virus-specific memory T cells have been shown to exhibit antitumor effects in response to checkpoint blockade [33]. This has interesting implications for the extent of potential protection provided by previously generated tissue-resident memory T cells. It is very informative to utilize preclinical models that include bystander T cells (e.g., previously infected and/or aged mice) to model the human scenario of having a large pool of tissue-resident memory T cells that are capable of rapid TCR-independent responses.

Mouse and human differences

NK cells and T cells have many differences across species. For example, NK cells have whole receptor family differences across species that carry out similar functions, such as Ly49s (C-type lectin-like receptors) in mice and killer-cell immunoglobulin-like receptors (KIRs) in humans, which are activating or inhibitory receptors that bind MHC I molecules and govern NK cell activity. A summary of select markers used to phenotype NK cells in mice and humans is provided in Table 1. The diversity of NK cells across species could be in part due to coevolution with species-specific viruses and the lack of MHC restriction. NK cells are defined by different markers in mice and humans, however, in both species, they are CD3 − and NKp46+, so at minimum, they do not have TCR signaling and they express the activating receptor NKp46 which is conserved across mammalian species and has been shown to bind to hemagglutinin (HA) [34]. In C57BL/6 mice, NK1.1 is a dominant activating receptor and has been found to bind the MCMV-encoded protein m12 [35]. In humans, CD56 is present on all NK cells—being either highly or lowly expressed making up the bright and dim subsets discussed later. Interestingly, NK cells are found throughout tissues in humans, including lymph nodes and CD56bright NK cells are the predominate population there [36], while negligible numbers of NK cells are found in the lymph nodes of mice without infection, which could be due to their specific-pathogen free status contrasting with humans [37].

Table 1

In humans only, there are two additional activating receptors in the Natural Cytotoxicity Receptor family with NKp46: Ncr2/NKp44 and Ncr3/NKp30. The Fc receptor, CD16, is expressed by NK cells of both species and plays an important role in antibody-dependent cell-mediated cytotoxicity (ADCC). Certain Fc receptor polymorphisms have been associated with better responses to monoclonal antibody therapy—with the hypothesis that this depends on ADCC [38]. It is important to mention the concept of split anergy that is present in NK cells of both species and appears to be a trade-off in functions—they can specialize in cytotoxicity or cytokine production [39–43]. IL-12 and IL-18 stimulation leads to the loss of CD16 by cleavage by ADAM17 and a related increase in IFN-γ production and degranulation [44].

There is a difference in the family of molecules in NK cells related to education and licensing by receiving signals from MHC I molecules in mice and humans (discussed in Section Phenotype and function of NK and T cells and models for study). The Ly49 family of mice and the KIR family of humans bind MHC I molecules and can either be activating or inhibitory. Ly49H binds the MCMV-encoded glycoprotein m157 which is thought to have evolved originally as a way for the virus to mimic MHC I and inhibit NK cells, however, NK cells evolved in response to the complementary activating receptor. On human NK cells, NKG2C, also an activating receptor, has been shown to bind HLA-E stabilized by the CMV-encoded peptide UL-40 and in that way confers a CMV-specific response [45]. Human blood exhibits a broad range of NKG2C+ NK cells—with a positive correlation with seropositivity [46].

Roles of NK cells and CD8 T cells in infection

The immune system includes barriers and sensors to protect from pathogens. However, once a pathogen has invaded a cell, the best course of action is to kill that cell in a controlled manner to shut down pathogen replication, lysis, and spreading to neighboring cells. Cancer can be likened to a chronic infection where inflammation and suppressive factors persist for a long time and in a way, a new challenge emerges to keep a balance of immune invigoration without immune-mediated damage [47]. For example, the important role of NK cells in primary response to CMV has been highlighted in several case reports of immunodeficiency—in both cases, the lack of NK cells precipitated lethal CMV infections [48,49]. CMV is a highly prevalent, usually asymptomatic herpes infection, but NK cells play an important role in keeping the virus in check. The paradigm is that NK cells are able to respond faster than T cells and are most important in the early stages of viral infection, but we are currently learning more about bystander T cell rapid responses to TLR signaling, NKp30 and NKG2D signaling, and proinflammatory cytokine responses [50,51]. NK cells are also critical in antiviral responses in the lymphopenic post-HSCT environment in which exposure to CMV or viral reactivation results in the expansion of adaptive NK cells (CD56dimNKG2C+ CD57+) with potent ADCC and persistence long after resolution of the CMV infection [52].

There are numerous positive and negative feedback pathways during an antiviral response and the cellular interactions change over time as the immune system goes from early response through to resolve the inflammation and return to homeostasis. The paradigm is that the T cell and B cell receptor repertoire is altered upon exposure to pathogen as well as anatomical environment and these have long-lasting consequences. Unlike T and B cell diversity, NK cell diversity increases with immune experience and reflects an individual’s history of pathogen encounters and memory capabilities which has been correlated with future susceptibility to disease [53,54]. The concept of trained immunity is continuously being investigated as even innate cells have long-lasting changes following pathogen exposure [55,56]. Future mechanistic studies will bring a new level of understanding of the dynamics of NK and T cell diversity and their functional responses, as well as insights to design therapeutic strategies in the settings of infection and cancer.

Phenotype and function of NK and T cells and models for study

NK cells and T cells develop from common lymphoid progenitors, with NK and CD8 T cells relying on many of the same transcription factors and signaling cues to differentiate with divergence of T cells undergoing education in the thymus. During thymic involution, which occurs with aging, the production of naïve T cells dramatically dwindles. NK cells develop in the bone marrow and appear to continuously develop throughout life. These cells are classically defined by their ability to lyse virally infected or transformed target cells in a non-MHC-restricted manner and without prior sensitization. However, there is some evidence that activation in the bone marrow environment has an effect similar to negative selection and deletes cells from the circulating NK cell repertoire and leads to tolerance to antigens that were encountered early in development [57,58]. By different mechanisms, both NK and T cells have inhibitory pathways rendering them tolerant to self cells in nondisease settings. NK cells are inhibited by self-MHC molecules, whereas T cells require priming via the TCR in the presence of costimulation which occurs in an inflammatory lymphoid tissue. In fact, the degree of MHC I inhibitory signals that NK cells receive determines whether they are licensed or not—a phenomenon similar to priming. NK cells that receive a strong inhibitory signal from MHC I because they express inhibitory receptors that bind MHC I (Ly49 family in mice or KIR family in human) are more cytotoxic when they encounter an MHC I-negative target cell [59]. Once activated, NK and T cells can migrate to target tissues following a chemokine gradient. Memory T cell populations reside in tissue, in addition to innate-like T lymphocytes including mucosal-associated invariant T (MAIT) cells, and tissue-resident NK cells. The NK cells in human lymphoid tissue are predominately CD56bright, unlike in the peripheral blood where the CD56bright NK cell population is only ~ 10% with the rest being CD56dim NK cells that have been found to have higher cytotoxicity [60]. The activity of NK and T cells is governed by activating and inhibiting markers, as well as by cytokine signals and metabolic and innate receptors (Table 2). The essence of their role is to selectively kill virally infected cells without causing host damage. They can mediate target cell death via four main pathways: FasL, TNF, TRAIL, and perforin/granzyme B, all of which can be triggered by NKG2D signaling. Perforin-mediated lysis primarily induces necrosis in the target cells and it is this property that dominates short-term (4 h) in vitro cytotoxicity assays, which may lead to an underestimation of their true killing potential as long-term studies demonstrate involvement of these other pathways [61]. NK cells were originally described as large granular lymphocytes and this highlights their ability to rapidly respond to stimulation and degranulate, unlike T cells which require priming. Degranulation is the predominant pathway for killing target cells and NK cells are ready to do this spontaneously. Cytotoxic cells can kill multiple targets sequentially and without toxicity to themselves through the creation of a unidirectional synapse that protects them from perforin and granzyme, a mechanism of interest in the ongoing research [62].

Table 2

There have been many studies aiming to elucidate the lifespan of lymphocytes and it is difficult to separate a clonally expanded subset that persists over time versus the duration of a single cell’s survival just by assessing the phenotype. Further complicating studies of secondary responses, epigenetic changes can be passed on from parent cells which could bestow an enhanced protective phenotype [17,63]. Studies with deuterium-enriched glucose in humans have shown that NK cells in peripheral blood have a turnover rate of 14 days, while that of T cells in peripheral blood is ~ 100 days [64,65]. Bar-coded cell studies have revealed that NK cell clones persist for several months [66,67]. Interestingly, subpopulations of NK and T cells have been found to persist in tissue for more than a decade as discovered in the studies of liver transplants [68]. In addition to a difference in lifespan, the production of NK and T cells is very different—NK cells are continuously produced from bone marrow progenitor cells, however, as the thymus involutes with age, the production of naïve T cells is drastically reduced over time and the T cell repertoire shifts to predominantly memory cells.

The classic assays of cytotoxicity involve measuring death of target cells; in the case of NK cells: MHC I-negative targets (Yac-1 for mice, K562 for humans), while T cells can target tumor cells and cells with NKG2D ligands. NK and T cells are studied in vivo in the context of infections and cancer. Ex vivo functional assays can be used to measure their cytokine production and degranulation (CD107a marker). In the mouse model of hepatitis, infection with lymphocytic choriomeningitis virus (LCMV) has garnered much important information about the roles of NK and T cells in infection, their coregulation, and exhaustion over time [69,70]. The development of NKp46 (Ncr1) knockout mice has opened the field to cell-specific deletion and shed light on nonredundant roles of NK cells in viral infection [71].

Regulation of NK and T cells and modifying factors

A sometimes unappreciated aspect of NK cells is their role, not just as cytotoxic cells, but as regulatory helper cells—secreting cytokines that can support antigen presentation (i.e., GM-CSF increasing costimulation markers on APCs and the T cell response)—this is fulfilled by unlicensed NK cells in mice and humans and CD56dim NK cells in humans [37]. There is evidence of NK cell and dendritic cell (DC) cross talk within lymph nodes in the T cell zone, which puts them in the right place at the right time to influence the adaptive immune response [37,72]. Depletion of the unlicensed NK cell subset, but not the licensed subset, resulted in fewer mature DCs, which led to fewer antigen-specific T cells and increased viral titers in mice influenza and CMV models [37]. There is a multitude of ways that NK and T cells interact and regulate each other. Not only do they compete for the same cytokines, but also produce cytokines that inhibit each other, namely IL-10 and TGF-β. Regulatory T cells (Tregs) are able to suppress both cell types, but there is evidence of direct competition between NK cells and CD8 T cells where depletion of one cell type lead to compensatory proliferation of the other cell type and preserved overall cytotoxic capability [73]. It has also been shown in multiple models that NK cells, in addition to being rapid first responders, play a role in culling T cells and preventing immune-mediated pathology, i.e., NK cells can kill activated CD4+ T cells via FasL and reduce lung tissue damage in LCMV infection [69]. However, a fine balance needs to be achieved because, depending on the viral load, the presence of NK cells can exacerbate pathology by affecting the dynamics of the antigen-specific T cell response [69]. In several viral and tumor models, depletion of NK cells is counterintuitively beneficial to the adaptive immune response as evidenced by increased antigen-specific tetramer+ T cells; although timing is critical—with early-stage depletion or complete lack of NK cells still being detrimental to the overall immune response [49,74,75]. This could be related to effects on the viral load and subsequent antigen presentation to T cells. Modifying factors such as age, obesity, and comorbidities affect the function of NK and T cells and it is important to include these factors to make preclinical mouse models more relevant and reflective of clinical scenarios. Aging has been associated with changes in the frequencies of NK and T cells and impaired function. Obesity also has been associated with detrimental effects on function [76]. Long-term chronic infection has been shown to drive exhaustion of T cells and can detrimentally skew the immune repertoire with inflationary clones [77].

The role of IL-15 in lymphocyte development, activation, and maintenance

While initial studies used IL-2 to expand NK cells, studies with IL-2 knockout mice demonstrated normal NK cell numbers and function [78]. However, IL-15 is one of the more critical cytokines for development, activation, and maintenance of both NK cells and memory CD8 T cells. IL-15 is trans-presented by stromal cells and DCs on the IL-15α receptor to NK and T cells bearing the beta and common gamma cytokine chain (CD122 and CD132). Studies by Caligiuri and others using IL-15 and IL-15R knockout mice elegantly demonstrated that IL-15 was critical for not only NK cell development but also NK cell survival. Subsequently, there has been intense interest in the application of IL-15 as a means to augment NK cell recovery after HSCT and also with NK cell adoptive immunotherapy [79,80]. The trans-presentation aspect of this cytokine presents some hurdles to induce optimal efficacy but efforts using conjugation with soluble receptor [81] or antibody complexes [82] appear to circumvent many of these limitations. While increased activity is indeed shown with IL-15 stimulation, prolonged exposure has been demonstrated to result in NK cell anergy in vitro [83] and in vivo [84], although this has recently been brought into question with clinical results following IL-15 exposure [85]. This brings up the question: what is the optimum way to activate NK cells prior to adoptive transfer and is there a way to prolong their activity and survival?

Inhibition of NK and T cells

Major hurdles for harnessing NK and T cells for immunotherapy are increasing infiltration into tumors, optimizing sustained antitumor responses, and avoiding suppression in the tumor microenvironment. Immune cells can become dysfunctional by three main pathways: exhaustion, anergy, and senescence, reviewed previously [86]. Immune cells are held in check by inhibitory receptors, which are advantageous for self-tolerance and could have a role in long-term persistence and memory response preservation [70,87,88]. However, many immunotherapies aim to unleash NK and T cells by blocking these inhibitory pathways. The best-studied inhibitory receptors are: CTLA-4, PD-1, TIM-3, LAG-3, and more recently, TIGIT. These are all found on NK and T cells, with the exception of conflicting reports about whether NK cells express CTLA-4 and PD-1 [89,90]. There is scant data on CTLA-4 on NK cells, but it has been reported in mouse NK cells following in vitro IL-2 stimulation and also within mouse tumors [91].

The possibility of NK cells expressing PD-1, and therefore being a key effector/responder in checkpoint blockade, has received a lot of attention in recent years. The concept that NK cells survive long enough to become exhausted given their high turnover rate and rapid responses is controversial in itself. PD-1 expression on NK cells was first reported in humans with multiple myeloma in 2010 [89]. However, the antibody used, CT-011 (pidilizumab), has since been shown to also bind Delta-like ligand 1, which is found on NK cells [92]. PD-1 expression on NK cells has been reported in mice and humans—after stimulation with IL-2 or IL-12, IL-15, and IL-18 plus glucocorticoids [93]; in infection settings [94]; and predominately in tumors (solid and hematologic) [95]. Summaries and commentaries on these findings have been published previously [96,97]. However, there were many studies that did not report PD-1 expression on NK cells [7,98]. Some possible reasons for this discrepancy are: (1) the difficulty in staining immune cell populations from tumor cells that have been treated with enzymes (which can remove markers such as CD3 necessary to delineate NK cells from T cells), (2) the lack of uniform analysis and interpretation of flow cytometry data—with the possibility of not excluding dead and dying cells which have been shown to bind anti-PD-1 antibodies nonspecifically [99], and (3) the confounding factors of inflammatory environments where increased expression of Fc Receptors could nonspecifically bind to detection antibodies and also where cells could undergo higher levels of trogocytosis (displaying membrane proteins from cells with which they have interacted). NK cells have been shown to acquire markers from other cells via trogocytosis, particularly when they are activated [100,101]. Additionally, there is a possibility that human and mouse NK cells are regulated differently and express different inhibitory receptors over time. Human NK cells can live much longer than mouse NK cells in vitro—15 weeks versus 7–10 days, so there is a possibility that there is suite of factors associated with survival, proliferation, and associated regulation that differ across the